Digital BW, The Print

Martin,

Very interesting and informative site. Thank you.

Alan Huntley

----- Original Message -----
From: "Martin Wesley" <mwesley250@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Monday, May 19, 2003 9:56 AM
Subject: [Digital BW] Digital, film, scanning comparisons


> Here is a site I ran across last night with a lot of digital to scanned
film comparision trials to for resolution, detail, etc. Also gives a close
up look at the differences in some film sizes and different scanning
resolutions. It may be of interest to some. Unfortunately color oriented
rather than B&W.

>
> http://www.clarkvision.com/imagedetail/index.html
>
> Martin Wesley

----- Original Message ----- 
From: "A. Huntley" <leicam6@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Tuesday, May 20, 2003 6:15 PM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> Martin,
>
> Very interesting and informative site. Thank you.
>
> Alan Huntley

Alan,

Glad someone found it of use. Unfortunately he (nor do any of the endless
threads here) does not address the basic issue of what happens when you try
to print in B&W. All of the files and examples he shows are presumably 3
channel RGB. I am unclined to agree with his 10 megapixal threshold being
the point where direct digital cature equals or exceeds 35mm color film but
what about B&W? Since the conversion to B&W drops 2/3 of the data the
implication is that to match 35mm B&W you would need 30 megapixal direct
capture.

Martin Wesley

http://www.borderless-photos.de/guests.html



>
> ----- Original Message -----
> From: "Martin Wesley" <mwesley250@...>
> To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Sent: Monday, May 19, 2003 9:56 AM
> Subject: [Digital BW] Digital, film, scanning comparisons
>
>
> > Here is a site I ran across last night with a lot of digital to scanned
> film comparision trials to for resolution, detail, etc. Also gives a close
> up look at the differences in some film sizes and different scanning
> resolutions. It may be of interest to some. Unfortunately color oriented
> rather than B&W.
> >
> > http://www.clarkvision.com/imagedetail/index.html
> >
> > Martin Wesley
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or
&amp;amp;quot;flames.&amp;amp;quot;
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

Martin writes:

> Since the conversion to B&W drops 2/3 of the
> data the implication is that to match 35mm B&W
> you would need 30 megapixal direct capture.

Agreed.  But you'll never get really good black and white from any RGB
imaging device.  The only way to get top quality black and white is with a
dedicated black and white imaging device.  A CCD could certainly do this,
but only if it were truly dedicated to the task, with no color filters, and
I don't see anyone producing that.

Similarly, you can't scan color film and convert it to B&W with results that
come anywhere near to true B&W film.

Anthony,

> Agreed.  But you'll never get really good black and white from any RGB
> imaging device.

Too much of a generalization.  You can get exceptional B&W from a scanner,
which most all use RGB sensors.  Also, scanning backs use RGB and give full
color data, and they convert very very well to grayscale.

> The only way to get top quality black and white is with a
> dedicated black and white imaging device.  A CCD could certainly do this,
> but only if it were truly dedicated to the task, with no color
> filters, and
> I don't see anyone producing that.

Well, if you're talking cameras, Kodak has, and if you're talking scanners,
the Leaf was.

> Similarly, you can't scan color film and convert it to B&W with
> results that
> come anywhere near to true B&W film.

To a degree, I agree...but it isn't as bad as you make it out to be.  Have
you actually done any experiments, or do you know of any on the web?

Austin

Martin-

I don't really think you are losing 2/3 of the info --just not using it.
Remember that the (Lab) Lightness or Luminosity Channel IS the B&W layer.
The rest is just color info dropped on top. (This is an oversimplification,
of course.)

The advantage of using the RGB before converting to B&W is in the channel
mixer.  That really amounts to the same as using yellow, orange, green, etc.
filters on a B&W film.

Seth


=Glad someone found it of use. Unfortunately he (nor do any of 
=the endless threads here) does not address the basic issue of 
=what happens when you try to print in B&W. All of the files 
=and examples he shows are presumably 3 channel RGB. I am 
=unclined to agree with his 10 megapixal threshold being the 
=point where direct digital cature equals or exceeds 35mm color 
=film but what about B&W? Since the conversion to B&W drops 2/3 
=of the data the implication is that to match 35mm B&W you 
=would need 30 megapixal direct capture.
=
=Martin Wesley
=

> Agreed.  But you'll never get really good black 
 > and white from any RGB imaging device.

There's no reason why that should be true.    Scanning black and 
white film should produce flawless grayscale images.  Why wouldn't 
it?   What is lost?   If you scan T-Max or Tri-X the shape of the 
curve should be the same for all three channels, with just a 
translation applied due to the film base color.  So pick the one 
with the best s/n.    

I agree that digital cameras are another story, because they are 
only giving you their best guess of what the grayscale scene looked 
like, based on reconstructing it from the separate elements in a 
Bayes array.   So a digital camera really just gives you a good 
guess.     Hopefully even this problem would go away if Foveon ever 
comes up with a >3MP version of their chip and straightens out some 
of the S/N differences in the channels.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Martin Wesley" 


> I am unclined to agree with his 10 megapixal threshold
> being the point where direct digital cature equals or 
> exceeds 35mm color film 


Why?   That would imply there's no difference for color film between 
a 2700 PPI scanner (11 MP) and a 4000 PPI film scanner (24 MP).   
But anyone who's compared the two for fine grain film shot with 
sharp optics and a tripod-mounted camera knows that's certainly not 
true.

> Too much of a generalization.  You can get
> exceptional B&W from a scanner, which most
> all use RGB sensors.

The scanner isn't the original imaging device.  It doesn't matter what
colors it examines, as long as it can recognize light and dark, because the
material being scanned is colorless.  Even a single scan in purple would do
just fine.

> Also, scanning backs use RGB and give full
> color data, and they convert very very well
> to grayscale.

Not as well as if they just scanned in black and white to begin with.

> Well, if you're talking cameras, Kodak has,
> and if you're talking scanners, the Leaf was.

Kodak's camera is the rare (unique?) exception to the rule.

> Have you actually done any experiments, or do
> you know of any on the web?

I've done both real and thought experiments.

It's easy enough to see the theory.  Start with any RGB image, and try to
reconstruct the original spectral distribution of the light that produced
it.  You'll find that this is impossible, as more than one distribution can
produce any give RGB signal.  Because of this, no RGB signal can be
converted to grayscale in a way that will emulate any arbitrary original B&W
image, because original B&W capture is a function of a continuous, often
irregular curve of spectral sensitivity, whereas an RGB image is just three
numbers.  A great deal of information is thrown away in RGB capture.  The
same is true in B&W capture, of course, but as long as the B&W is the
_original_ capture, it doesn't matter--you won't be trying to convert it to
grayscale if it already is grayscale.  (Note, however, that converting one
B&W image to another is even less plausible than converting an RGB image to
an emulation of an arbitrary B&W image.)

Seth writes:

> The advantage of using the RGB before converting
> to B&W is in the channel mixer.  That really amounts
> to the same as using yellow, orange, green, etc.
> filters on a B&W film.

Unfortunately, no, it's not the same thing.

Imagine a scene that is lit in some areas with monochromatic yellow light
(as from low-pressure sodium vapor or something), and in other areas with a
blend of red and green light that looks--at least to human eyes--to be the
same color of yellow as the monochromatic light.

If you look at this scene with your eyes, everything appears to be yellow.

If you photograph this scene with a color camera (digital or film, it
doesn't matter), the resulting images also render the entire scene as
yellow.

If you photograph this scene with black and white film and a narrow-band
monochromatic yellow filter, however, you get a different result.  The final
image will be bright in the areas lit by true, monochromatic yellow light,
and dark in the areas lit by a blend of red and green light.  Thus, you'll
see sharp contrasts in the result that were not present in real life (to
your eyes, that is) or to color film/digital cameras.

Given this, it should be obvious that there is no way to convert the RGB
image to any kind of grayscale image that will emulate the black and white
scene photographed through a yellow filter.  Why?  Because both the
red/green areas and the truly yellow areas came out to the same shade of
yellow in the color images--nothing you can do in post-production will allow
you to separate them into light and dark.  So no matter what manipulations
you attempt, you'll never get the same result that you would have gotten
with black and white and a yellow filter.

This holds not just for the use of filters with B&W film, but for the use of
B&W film in general.  Black and white film responds differently to light of
different colors than does color film or a color CCD.  Once an image is
converted to RGB or grayscale, virtually all the information on the original
spectral distribution of light in the scene is lost.  As a result, nothing
can convert one RGB or B&W rendering into another one that emulates capture
of the original image with a different spectral sensitivity curve, because
the spectral information required for this is gone.

This rule is independent of the capture method you use.  Digital or film,
the rule is the same.  The only way to get true black and white with any
other rendering of light than standard, human-eye rendering is to shoot B&W
from the beginning.  It can be black and white film, or a monochrome CCD,
but it has to be B&W to start with.

Peter writes:

> There's no reason why that should be true.

There's an excellent reason why it is true, which I've just explained in a
separate answer to Seth's post.

The only accurate B&W you can get from an RGB image is "perceptual" B&W that
precisely matches the spectral sensitive of the RGB capture device in the
first place.  All other B&W renderings are impossible.  You cannot get the
look of Tri-X from an RGB image, although you can try to simulate it by
cranking down the red channel a bit.  For real Tri-X, though, you need to
shoot Tri-X.

As a more extreme example, there is no way to get infrared from an RGB
image.  You can pretend by dropping the blue and green channels and
desaturating, but that's just pretend ... it's nothing like real infrared.
To get real infrared images, you must capture the image originally with a
device that can record infrared, such as IR film or a CCD with a filter that
transmits IR and only IR.

> Scanning black and white film should produce
> flawless grayscale images.

And it does.  But in that case you are not capturing the original image;
that has already been done for you.

> Why wouldn't it?   What is lost?

Nothing, in a scan.  Information is lost only when originally capturing the
image from real life.

No form of image capture today can capture information for the entire
visible spectrum for any image.  Color imaging reduces that entire spectrum
to three values; black and white imaging reduces the entire spectrum to one
value.  It's a one-way process.  You can do anything you want when making
that conversion, but once it's made, it's impossible to transform it into
the result of any other conversion, because the required original spectral
information is gone.

> If you scan T-Max or Tri-X the shape of the
> curve should be the same for all three channels,
> with just a translation applied due to the film
> base color.

It is.  And no adjustment for base color is required.

> I agree that digital cameras are another story ...

In this particular domain, it doesn't matter whether you use digital or
film.  The constraints of the conversion to RGB or grayscale are _identical_
for both types of image capture.  It doesn't have anything to do with
digital at all.

The only disadvantage digital photographers have is that they cannot shoot
straight black and white with their color digicams--whereas a film
photographer can replace color film with B&W film.

Anthony,

> > Too much of a generalization.  You can get
> > exceptional B&W from a scanner, which most
> > all use RGB sensors.
>
> The scanner isn't the original imaging device.

Don't do this.  Your comment was, and I quote:

"But you'll never get really good black and white from any RGB
imaging device."

It says ANY RGB imaging device.  As I said, your statement was too much of a
generalization, and whether it's the original imaging device or not, it IS
an imaging device.

> > Also, scanning backs use RGB and give full
> > color data, and they convert very very well
> > to grayscale.
>
> Not as well as if they just scanned in black and white to begin with.

That is not true as an overall statement.

> > Well, if you're talking cameras, Kodak has,
> > and if you're talking scanners, the Leaf was.
>
> Kodak's camera is the rare (unique?) exception to the rule.

But it WAS and IS available.

I don't have time to waste arguing with you.

Austin

----- Original Message ----- 

From: "Anthony Atkielski" <anthony@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 21, 2003 12:39 AM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> Martin writes:
>
> > Since the conversion to B&W drops 2/3 of the
> > data the implication is that to match 35mm B&W
> > you would need 30 megapixal direct capture.
>
> Agreed.  But you'll never get really good black and white from any RGB
> imaging device.  The only way to get top quality black and white is with a
> dedicated black and white imaging device.  A CCD could certainly do this,
> but only if it were truly dedicated to the task, with no color filters,
and
> I don't see anyone producing that.
>
> Similarly, you can't scan color film and convert it to B&W with results
that
> come anywhere near to true B&W film.
>
Anthony,

I have seen far too many fine B&W prints made from scans of color
transparencies to agree with you on that one!

My remarks were actually a question as to whether color digital cameras that
are a match for 35mm color film and transparencies would match 35mm B&W. I
am very interested to see some prints make with the newer 10+ mega pixel
cameras. If the prints are good, the numbers don't matter. Except the $$$
number of course. <G>

Martin Wesley

Austin writes:

> It says ANY RGB imaging device.

I usually think of original imaging devices when I speak of them at all.  A
projector is an imaging device, but few people think of projectors when they
hear the term.

Nevertheless, I'll be specific:  an ORIGINAL imaging device--the one that
translates the real world scene to some other representation.

> That is not true as an overall statement.

It cannot be otherwise, for reasons I've explained at length in other posts.

> I don't have time to waste arguing with you.

And yet here you are.  Hmm.

Martin writes:

> I have seen far too many fine B&W prints made
> from scans of color transparencies to agree
> with you on that one!

You can certainly get good B&W from color transparencies, but you can get
only ONE B&W that way.  I meant good in the sense of B&W as produced by true
B&W films.  You cannot duplicate the results of actual B&W films by
grayscaling RGB images.  You can get one type of B&W from RGB, with a couple
of minor variations, but you can't get the vast majority of B&W renderings
offered by dedicated B&W capture (film or digital).

The classic example is infrared (because that's easier for people to
understand), but it applies to B&W in the visible light range, too.

> My remarks were actually a question as to whether
> color digital cameras that are a match for 35mm
> color film and transparencies would match 35mm B&W.

Only to the extent that color film and transparencies are a match for 35mm
B&W.  The constraints are identical for both color digital and color film.

----- Original Message ----- 

From: "Anthony Atkielski" <anthony@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 21, 2003 9:39 AM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> Martin writes:
>
> > Since the conversion to B&W drops 2/3 of the
> > data the implication is that to match 35mm B&W
> > you would need 30 megapixal direct capture.
>
> Agreed.  But you'll never get really good black and white from
any RGB
> imaging device.  The only way to get top quality black and
white is with a
> dedicated black and white imaging device.  A CCD could
certainly do this,
> but only if it were truly dedicated to the task, with no color
filters, and
> I don't see anyone producing that.
>
> Similarly, you can't scan color film and convert it to B&W with
results that
> come anywhere near to true B&W film.

If we are on that route I suggest to get a scanner for B&W film
that is a true B&W scanner as well.
The Agfa Horizon Plus that I have is a 3 pass (3 filter) scanner
for RGB but it is a single pass (ND filter in front of the
sensor) for B&W scanning. 1200 ppi.  The results I get from my
Nikon 8000 and the Epson 3200 are much better though. On all 3 I
use wetmounting for the best results. If it was true that a
mosaic CCD or three line sensor can only contribute 1/3 of its
quality for a B&W image this difference wouldn't be so
significant. The Horizon is an older scanner but its construction
and optics belong to the best around. It just isn't so simple
that a mosaic CCD or CMOS can only contribute 1/3 of its data to
B&W.

I'm still using (as an amateur) colour negative and Polaroid B&W
film for everything and the scans are perfect. The digital takes
for art reproduction that I order from photographers are from
Sinar backs mainly. Excellent material for art prints that can't
be done that well with film>digital if it was only because one
avoids an extra cmyk>rgb translation in the process. Most art up
to A3 is scanned directly if possible.
Some is done on a wide format scanner. The rest of the work is
scanning and printing done for photographers. The colour quality
of the digital back images is better than the directly scanned
images. One wonders why but a thread on the colorsync list
revealed that scanners are still made to reproduce photo dyes
best (even the non film flatbeds) so real life pigments etc get
the wrong treatment. Digital backs are made for real life
colours.

The best process for art reproduction/archiving these days is
done with multi spectral takes where up to 7 narrow band filtered
CCD images are used for the final image. I'm sure one could brew
a nice B&W image of that colour image if needed, most likely a
much better one than if only one is done without a filter. Like
with all digital photography extra sampling adds.

Ernst

As I read this thread I cannot help but wonder if a dedicated b & w scanner
does exist somewhere, it would seem like a good fit for radiography (medical
& industrial)

 

I am also curious to know if I should buy a high quality scanner or jump to
fully digital. I have lots of MF black & white negs I'd like to digitize, so
I would like to know more about the quality of current MF scanners, what to
look for/avoid.

 

Tony

www.lightworx.ca <http://www.lightworx.ca/> 

-----Original Message-----
From: Anthony Atkielski [mailto:anthony@...] 
Sent: Wednesday, May 21, 2003 3:40 AM
To: DigitalBlackandWhiteThePrint@yahoogroups.com
Subject: Re: [Digital BW] Digital, film, scanning comparisons

 

Martin writes:

> Since the conversion to B&W drops 2/3 of the
> data the implication is that to match 35mm B&W
> you would need 30 megapixal direct capture.

Agreed.  But you'll never get really good black and white from any RGB
imaging device.  The only way to get top quality black and white is with a
dedicated black and white imaging device.  A CCD could certainly do this,
but only if it were truly dedicated to the task, with no color filters, and
I don't see anyone producing that.

Similarly, you can't scan color film and convert it to B&W with results that
come anywhere near to true B&W film.






Yahoo! Groups Sponsor


 
<http://rd.yahoo.com/M=251812.3170658.4537139.1261774/D=egroupweb/S=17050191
82:HM/A=1564416/R=0/*http:/www.netflix.com/Default?mqso=60164797&partid=3170
658> 


 
<http://us.adserver.yahoo.com/l?M=251812.3170658.4537139.1261774/D=egroupmai
l/S=:HM/A=1564416/rand=462135995> 


Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:

http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint

If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.

Please follow these basic guidelines:
- Include your full name with your message.
- Include the address of your website, if you have one.
- As threads develop, trim off excess portions of earlier messages to keep
them short.
- As the topic of a thread changes remember to change the subject header.
- Good manners are required at all time. No personal attacks or
&amp;amp;quot;flames.&amp;amp;quot;
- Complete your Yahoo profile.
- Before posting a question, search the message archives and the various
resources on the homepage. 




Your use of Yahoo! Groups is subject to the Yahoo!
<http://docs.yahoo.com/info/terms/>  Terms of Service. 



[Non-text portions of this message have been removed]

Ernst writes:

> If we are on that route I suggest to get a
> scanner for B&W film that is a true B&W scanner
> as well.

It shouldn't matter for scanners, since they are scanning material that
transmits all wavelengths of visible light equally.  However, a B&W scanner
might be very slightly better, all else being equal.  It would be faster,
too.

> If it was true that a mosaic CCD or three
> line sensor can only contribute 1/3 of its
> quality for a B&W image this difference
> wouldn't be so significant.

It's true for mosaic CCDs, but scanners don't use mosaics.  They scan all
three colors for every pixel.

> It just isn't so simple that a mosaic CCD or
> CMOS can only contribute 1/3 of its data to
> B&W.

You're confusing the original capture of the image from real life with
subsequent transformations.  It's the ORIGINAL capture that counts (I guess
I should have been clearer about that).  You don't lose anything with a RGB
scan of B&W film.

> The best process for art reproduction/archiving
> these days is done with multi spectral takes
> where up to 7 narrow band filtered CCD images
> are used for the final image. I'm sure one could brew
> a nice B&W image of that colour image if needed, most likely a
> much better one than if only one is done without
> a filter.

Sort of.  Having seven color values lets you grayscale with much more
flexibility than having three color values does.  However, direct B&W
capture records _all_ colors, not just three or seven, and so it works
better than anything else for B&W.

----- Original Message ----- 

From: "Seth Rossman" <seth@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 21, 2003 6:42 AM
Subject: RE: [Digital BW] Digital, film, scanning comparisons


> Martin-
>
> I don't really think you are losing 2/3 of the info --just not using it.
> Remember that the (Lab) Lightness or Luminosity Channel IS the B&W layer.
> The rest is just color info dropped on top. (This is an
oversimplification,
> of course.)

> The advantage of using the RGB before converting to B&W is in the channel
> mixer.  That really amounts to the same as using yellow, orange, green,
etc.
> filters on a B&W film.
>
>
Seth,

Well if I can't use it I guess I consider it lost. Your are right that by
pulling information from each of the three channels you can maximize the
result you get when you have to drop to one channel. It really isn't just a
megapixel game but the issue of resolution and bit depth. With RGB you have
in a sense a bit depth of 24 and with B&W 8. Tripling the ppi of a digital
camera would only help to some extent but perhaps not in the key area of
capturing a large number of tones. Maybe they need to just be 16-bit RGB.
That would be a real boost to direct digital B&W.

Martin

Tony writes:

> I am also curious to know if I should buy a high
> quality scanner or jump to fully digital.

MF film and a scanner will give you better quality for your money than will
digital, by a very handsome margin.

I looked at the same thing a while ago.  Going digital would have cost
$8000-$10000, and would have provided, at best, 2/3 the quality of 35mm
film.  I calculated that going MF would cost thousands less, and in fact,
even with a Hasselblad kit I still came out ahead--and that includes the
price of a MF scanner.  So I forgot about digital and went with MF, and I
have not regretted it.

> I have lots of MF black & white negs I'd like to
> digitize, so I would like to know more about the
> quality of current MF scanners, what to
> look for/avoid.

I can only speak for the Nikon LS-8000ED, but it's great!

I've already posted links to some B&W scans in a Portra 400BW thread, but if
you missed that, look at these

http://www.mxsmanic.com/stairs.jpg (small version of 6x6 scan at 4000 dpi on
LS-8000ED, 1/64 of original size)
http://www.mxsmanic.com/stairs1.jpg (tiny portion of above picture at
original size of scan)

Examination of Velvia slides with a microscope has revealed that the
LS-8000ED was pulling all the detail available on the slide.

----- Original Message ----- 

From: "Peter Nelson" <peter@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 21, 2003 7:14 AM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Martin Wesley"
>
>
> > I am unclined to agree with his 10 megapixal threshold
> > being the point where direct digital cature equals or
> > exceeds 35mm color film
>
>
> Why?   That would imply there's no difference for color film between
> a 2700 PPI scanner (11 MP) and a 4000 PPI film scanner (24 MP).
> But anyone who's compared the two for fine grain film shot with
> sharp optics and a tripod-mounted camera knows that's certainly not
> true.

Because I thought the information and examples he presented supported his
conclusion.

Martin

----- Original Message ----- 

From: "Anthony Atkielski" <anthony@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 21, 2003 11:06 AM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> Martin writes:
>
> > I have seen far too many fine B&W prints made
> > from scans of color transparencies to agree
> > with you on that one!
>
> You can certainly get good B&W from color transparencies, but you can get
> only ONE B&W that way.  I meant good in the sense of B&W as produced by
true
> B&W films.  You cannot duplicate the results of actual B&W films by
> grayscaling RGB images.  You can get one type of B&W from RGB, with a
couple
> of minor variations, but you can't get the vast majority of B&W renderings
> offered by dedicated B&W capture (film or digital).

Anthony,

At this point I am not sure how much digital B&W printing you have done but
you can get very different B&W renderings from color film scans or digital
color files depending upon how you blend the RGB channels in Photoshop. In
addition you can buy software that takes color files and applies curves to
them to simulate the response of various B&W films and color filters as
well.
>
> The classic example is infrared (because that's easier for people to
> understand), but it applies to B&W in the visible light range, too.

There has even been some software to simulate that!

I shoot B&W film myself and you are not incorrect in what you say, but it
really seems to have no significance in producing a fine print using digital
means.

Martin

Anthony,

> However, a 
> B&W scanner
> might be very slightly better, all else being equal.  It would be faster,
> too.

Why would a B&W scanner be faster?

Austin

> As I read this thread I cannot help but wonder if a dedicated b &
> w scanner
> does exist somewhere, it would seem like a good fit for
> radiography (medical
> & industrial)

Anton,

The Leafscan 45 and 35 both have dedicated B&W scanning modes.  They use a
single channel with a neutral density filter.  IMO, they produce superior
scans than just scanning B&W film with a typical CCD scanner that scans the
B&W film in RGB.

There also used to be dedicated B&W only flatbed scanners, and pre-press
scanners, as I designed a few of them.  That was over 10 years ago though,
and I haven't seen any since.

Austin

----- Original Message ----- 

From: "Anton Pickard" <lightworx@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 21, 2003 11:09 AM
Subject: RE: [Digital BW] Digital, film, scanning comparisons


> As I read this thread I cannot help but wonder if a dedicated b & w
scanner
> does exist somewhere, it would seem like a good fit for radiography
(medical
> & industrial)

Anton,

A drum scanner will give you a "monochrome" scan. Some of the older
multipass CCD scanners such as the Leaf also had neutral density
capability.>
>
> I am also curious to know if I should buy a high quality scanner or jump
to
> fully digital. I have lots of MF black & white negs I'd like to digitize,
so
> I would like to know more about the quality of current MF scanners, what
to
> look for/avoid.

I have owned the Polaroid 120SS and was very pleased with it. I have also
seen prints made from scans on the Nikon 8000 and Minolta Multi PRO which
were excellent. All of them have their supporters but I don't think you
could go drastically wrong with any of them.

If you have a significant number of MF negs you want to digitize then buying
a scanner will be more cost effective than having them scanned. Also gives
you a level of creative and quality control that is essential to getting the
most out of your negs.

Martin

Anthony,

> > It says ANY RGB imaging device.
>
> I usually think of original imaging devices when I speak of them
> at all.

Write what you mean.  I, nor can anyone else, read your damn mind.  It seems
more like a game to you.  You make a clearly wrong statement, someone points
out it's wrong, and why, then you claim what you said wasn't what you said
(and you conveniently remove what you initially said in your clipping).

> A
> projector is an imaging device, but few people think of
> projectors when they
> hear the term.

I don't care what other people think.  I are about what YOU said, and it was
wrong.

> > I don't have time to waste arguing with you.
>
> And yet here you are.  Hmm.

No, Anthony.  I am not arguing with you.  I am telling you like it is, and
you can choose to weasel around your incorrect statements trying to somehow
come up with some story as to why you weren't wrong, when you clearly were,
or you can simply admit that you were wrong, and move on.  Either way, it's
not arguing, at least on my part.

Austin

Martin,

> A drum scanner will give you a "monochrome" scan.

Er, how?  I don't know of any, but perhaps there are, that have a ND filter
channel.  PMT scanners use filters just like CCD scanners, it's just that
they use a beam splitter to separate out the channels, and they all three
read from the same beam.

Regards,

Austin

----- Original Message ----- 

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 21, 2003 12:10 PM
Subject: RE: [Digital BW] Digital, film, scanning comparisons


> Martin,
>
> > A drum scanner will give you a "monochrome" scan.
>
> Er, how?  I don't know of any, but perhaps there are, that have a ND
filter
> channel.  PMT scanners use filters just like CCD scanners, it's just that
> they use a beam splitter to separate out the channels, and they all three
> read from the same beam.

That's why I put the "monochrome" quotes. That does raise a question in my
mind though. With the drum scanners where in the light path are the color
filters? Are they between the PMT and the film or between the light sources
and the film? If the latter case, I would assume there is a theoretical
advantage in resolution to using a single color light source. If filtration
is after the light passes through the film than all three channels would be
equal.

Martin

Hi Martin,

> With the drum scanners where in the light path are the color
> filters? Are they between the PMT and the film or between the
> light sources
> and the film?

Light path of a drum scanner:

light source -> mirror -> drum -> film -> lense -> splitter -> filter(s) ->
PMT(s)

> If the latter case, I would assume there is a theoretical
> advantage in resolution to using a single color light source.

Single color light source, typically a halogen or something very bright and
focused.

> If
> filtration
> is after the light passes through the film than all three
> channels would be
> equal.

The filters will make them not equal though...  What do you mean by "equal",
and why is that important?

Austin

Martin writes:

> At this point I am not sure how much digital
> B&W printing you have done but you can get very
> different B&W renderings from color film scans
> or digital color files depending upon how you
> blend the RGB channels in Photoshop. In addition
> you can buy software that takes color files
> and applies curves to them to simulate the
> response of various B&W films and color filters as
> well.

"Simulate" is the operative word here.  You can make something look like
Tri-X or Tech Pan, but if you compare it with the same scene actually
photographed with these films, you may be surprised.  Almost all the
information required to reproduce the response of most B&W films is missing
from color images.

> There has even been some software to simulate that!

Compare the simulations to the real thing, and again you may be surprised.

> I shoot B&W film myself and you are not incorrect
> in what you say, but it really seems to have no
> significance in producing a fine print using digital
> means.

You can certainly produce fine prints; but if you want prints that look
exactly the way a certain film looks, you must capture the original scene
with that film.

Austin writes:

> Why would a B&W scanner be faster?

One pass instead of three, less processing of the data, and so on.

----- Original Message ----- 

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 21, 2003 12:52 PM
Subject: RE: [Digital BW] Digital, film, scanning comparisons


> Hi Martin,
>
> > With the drum scanners where in the light path are the color
> > filters? Are they between the PMT and the film or between the
> > light sources
> > and the film?
>
> Light path of a drum scanner:
>
> light source -> mirror -> drum -> film -> lense -> splitter ->
filter(s) ->
> PMT(s)

Austin,

Thanks for the info.
>
> > If the latter case, I would assume there is a theoretical
> > advantage in resolution to using a single color light source.
>
> Single color light source, typically a halogen or something very bright
and
> focused.
>
> > If
> > filtration
> > is after the light passes through the film than all three
> > channels would be
> > equal.
>
> The filters will make them not equal though...  What do you mean by
"equal",
> and why is that important?

My thought was that a light passing through the negative that is of high
frequency and very narrow in frequency range would result in higher
resolution in the scan since such a light would have less and more
consistant edge diffraction than say a white light source. In practical
terms this may not be significant since the quality of other elements in the
scanner might wipe out the benefit.

If the filters are after the splitter then the resolution of the three
channels should be the same if you are scanning a B&W neg, or am I missing
something such as PMT response to different colors.

Martin

Anthony,

> > Why would a B&W scanner be faster?
>
> One pass instead of three, less processing of the data, and so on.

1) you are assuming that a color scanner is three passes.  Not many, if any
made today, are.  Most all current scanners use a tri-linear sensor that
provides all three channels at once, so only one pass is needed.  Some
Nikons are an exception, as they have a tri-linear sensor, but with no color
filters.  The color filtration is done by the light source.  Their premise
failed, in that they expected to end up with the same scan time as a
tri-linear with filters, but because of banding issues due to calibration
errors in their firmware/hardware, these scanners are relegated to using
only one of the CCD rows.  That does not make it a three pass scanner, it
does do all three colors in the same pass, it's just that it will take three
exposures for each line.

2) processing of the data does not add any time to the scanning time (I've
designed scanners, and know what processing is done, and how it's done), as
the scanning time is far longer, and is the limiting factor.  Also, there
really is no "processing" that goes on in the scanner to the data that takes
any time.  The only thing that happens is the data goes from the A/D through
a LUT (Look Up Table), which is done in real time.  The only time any actual
processing is done is when you don't want the scanner to provide full
optical resolution data, and therefore it has to interpolate, but, this is
done "on the fly", and therefore adds insignificant time to the overall scan
time anyway.

3) and so on?  Would you be more specific?

Austin

Austin (or others)-

You seem to be in scanner mode so perhaps you could clarify something 
for me.  For as long as I can remember you have been saying things like 
this:

> The Leafscan 45 and 35 both have dedicated B&W scanning modes.  They 
> use a
> single channel with a neutral density filter.  IMO, they produce 
> superior
> scans than just scanning B&W film with a typical CCD scanner that 
> scans the
> B&W film in RGB.
>

but it has never been clear to me precisely why this is.

My scanner is a polaroid ss120.  It shines a white(ish?) light through 
my silver film (which I assume is fairly neutral but with varying 
density) and the intensity of resultant light is detected by a CCD 
after the light passes through a colour filter.

Why would the colour of the filter affect the relative intensities of 
the light hitting the CCD.  If my assumption that silver film is 
neutral is correct then aren't the intensities of the blue frequencies 
affected by passing through the film in a similar way to the red 
frequencies.  Why would it matter if the filter is neutral, red, green, 
blue, or whatever.

The only explanations I could come up with were:

1. film is not neutral (ok probably true but it seems to me the effect 
would be marginal),
2. the sensor's sensitivity varies with the wavelength of the light 
(ok, but since I am only interested in the intensities relative to each 
other why does that matter?)

Is one of these correct or, in your opinion, it is something else.

Thanks,

--
Kevin Gulstene

Hi Kevin,

> My scanner is a polaroid ss120.  It shines a white(ish?) light through
> my silver film (which I assume is fairly neutral but with varying
> density) and the intensity of resultant light is detected by a CCD
> after the light passes through a colour filter.

Passes through THREE different color filters...  One, red, one green and one
blue.  Resulting in three channels of color information.

> Why would the colour of the filter affect the relative intensities of
> the light hitting the CCD.

Right or wrong, that doesn't have a thing to do with the reason.  It's the
property of the CCD response/artifacting to different color lights.  Each of
the three lights gives different responses to the CCD and has different
artifacting.  Red, because it has the highest energy of the three, tends to
be the fuzziest, simply because of what are called bloom and smear.  Bloom
is basically saturation of the sensing element, smear is basically crosstalk
between adjacent sensing elements, because of the intensity (this is the
artifact that PMT scanners do NOT suffer from, as they scan one "spot" at a
time).  Blue is the next worst, then green is the best...but not always.

> Why would it matter if the filter is neutral, red, green,
> blue, or whatever.

CCD response is the primary reason.  Neutral density has the lowest
artifacting.

> The only explanations I could come up with were:
>
> 1. film is not neutral (ok probably true but it seems to me the effect
> would be marginal),

That is true, but as you suggest may be marginal.

> 2. the sensor's sensitivity varies with the wavelength of the light
> (ok, but since I am only interested in the intensities relative to each
> other why does that matter?)

Yes, and there is more to that, as I said, the artifacting varies as well.

> Is one of these correct or, in your opinion, it is something else.

Yes, and yes ;-)

Regards,

Austin

Austin-

Thanks for that.

>
>> Why would the colour of the filter affect the relative intensities of
>> the light hitting the CCD.
>
> Right or wrong, that doesn't have a thing to do with the reason.  It's 
> the
> property of the CCD response/artifacting to different color lights.  
> Each of
> the three lights gives different responses to the CCD and has different
> artifacting.  Red, because it has the highest energy of the three, 
> tends to
> be the fuzziest, simply because of what are called bloom and smear.  
> Bloom
> is basically saturation of the sensing element, smear is basically 
> crosstalk
> between adjacent sensing elements, because of the intensity (this is 
> the
> artifact that PMT scanners do NOT suffer from, as they scan one "spot" 
> at a
> time).  Blue is the next worst, then green is the best...but not 
> always.

I think I understand this but why is the CCD more sensitive to smear 
and bloom associated with red in the absence of blue and green.  That 
is, why would these artifacts appear when the light is filtered and not 
appear when it goes through a ND filter?

I am interested in this because I have recently started 4x5 and since 
my Umax powerlook 1100 is quite soft I'll soon be spending some more 
money on another  scanner.  I'm trying to appreciate how big a 
difference this makes in the overall scheme of things.

--
Kevin Gulstene

Question,

Why do you say red has " the highest energy of the three." I would 
expect that would depend on the spectrum of the light source and the 
bandwidth of the filters. Please elaborate.

Truman

Austin Franklin wrote:

>Hi Kevin,
>
>  
>
>>My scanner is a polaroid ss120.  It shines a white(ish?) light through
>>my silver film (which I assume is fairly neutral but with varying
>>density) and the intensity of resultant light is detected by a CCD
>>after the light passes through a colour filter.
>>    
>>
>
>Passes through THREE different color filters...  One, red, one green and one
>blue.  Resulting in three channels of color information.
>
>  
>
>>Why would the colour of the filter affect the relative intensities of
>>the light hitting the CCD.
>>    
>>
>
>Right or wrong, that doesn't have a thing to do with the reason.  It's the
>property of the CCD response/artifacting to different color lights.  Each of
>the three lights gives different responses to the CCD and has different
>artifacting.  Red, because it has the highest energy of the three, tends to
>be the fuzziest, simply because of what are called bloom and smear.  Bloom
>is basically saturation of the sensing element, smear is basically crosstalk
>between adjacent sensing elements, because of the intensity (this is the
>artifact that PMT scanners do NOT suffer from, as they scan one "spot" at a
>time).  Blue is the next worst, then green is the best...but not always.
>  
>



[Non-text portions of this message have been removed]

Hi Truman,

> Why do you say red has " the highest energy of the three." I would
> expect that would depend on the spectrum of the light source and the
> bandwidth of the filters. Please elaborate.

The light source is a white light, and has all colors, it is filtered with
"a" color filter, and only light of the color of that filter reaches that
sensor.  Red just simply has the highest energy of the three, as far as the
CCD "sees" it.  It has to do with the quantum efficiency of the CCD (how the
CCD reacts to different spectrums of light).  Red typically has up to an %80
quantum efficiency.  Blue and green only top out at about 60% right at their
peak sensitivity.  Red stays quite high for almost the full width of its
sensitivity.  Also, the spectral width of red is very very high, from 575nm
to 850+ nm.  Blue is 375nm - 550nm and green 450nm - 625nm.

Bring a color image into PS and thumb through the channels (CTL 1, 2 or 3
for RGB, and CTL ~ for all), and you'll see that red is easily the fuzziest,
blue and green are very image dependant as to which is better...

Regards,

Austin

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony 
Atkielski" <anthony@a...> wrote:
> Peter writes:
> 
> > There's no reason why that should be true.
> 
> There's an excellent reason why it is true, which I've just 
explained in a
> separate answer to Seth's post.
> 
> The only accurate B&W you can get from an RGB image 
> is "perceptual" B&W that precisely matches the spectral
> sensitive of the RGB capture device in the first place. 

Not if you're scanning black and white film.   The shape (N.B., the 
shape, not the translation) of the histogram will be the same in all 
3 channels of your scanning device.   All you need to pick the one 
that gives you the best S/N ratio.   That shape, furthermore, 
reflects the spectral sensitivity of the film, not the scanning 
device.   For instance, if the b+w film in question is pre-
panchromatic it will tail off on red subjects.  But it will do so for 
the R,G, and B scanning elements.

Quantum efficiently  has nothing to do with it.  Red is at the low end 
or the visible frequency spectrum. Normally the higher the frequency the 
wider the bandwidth of filters at that frequency, unless you use a local 
frequency reference to down convert the signal. Why is this not the case 
for red light which is a lower frequency than blue light? Different 
materials for CCD's have different frequency reposes. I guess I don't 
understand why "red" produces more energy output.

The energy of a photon is Planks constant X it frequency. A photon of 
blue has more energy than a photon of red. Different scanners use 
different light sources. I am still at a loss - not that I am disputing 
it - with your statement. I just don't understand.

Truman

Austin Franklin wrote:

>Hi Truman,
>
>  
>
>>Why do you say red has " the highest energy of the three." I would
>>expect that would depend on the spectrum of the light source and the
>>bandwidth of the filters. Please elaborate.
>>    
>>
>
>The light source is a white light, and has all colors, it is filtered with
>"a" color filter, and only light of the color of that filter reaches that
>sensor.  Red just simply has the highest energy of the three, as far as the
>CCD "sees" it.  It has to do with the quantum efficiency of the CCD (how the
>CCD reacts to different spectrums of light).  Red typically has up to an %80
>quantum efficiency.  Blue and green only top out at about 60% right at their
>peak sensitivity.  Red stays quite high for almost the full width of its
>sensitivity.  Also, the spectral width of red is very very high, from 575nm
>to 850+ nm.  Blue is 375nm - 550nm and green 450nm - 625nm.
>
>Bring a color image into PS and thumb through the channels (CTL 1, 2 or 3
>for RGB, and CTL ~ for all), and you'll see that red is easily the fuzziest,
>blue and green are very image dependant as to which is better...
>
>Regards,
>
>Austin
>
>
>
>
>Please visit the Group Homepage to check the Files, Bookmarks, Polls and other resources as they are often being updated. The page is at:
>
>http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
>If you wish to receive no emails or just a daily digest, or you wish to unsubscribe, please edit your Membership preferences by visiting this same page.
>
>Please follow these basic guidelines:
>- Include your full name with your message.
>- Include the address of your website, if you have one.
>- As threads develop, trim off excess portions of earlier messages to keep them short.
>- As the topic of a thread changes remember to change the subject header.
>- Good manners are required at all time. No personal attacks or "flames."
>- Complete your Yahoo profile.
>- Before posting a question, search the message archives and the various resources on the homepage. 
>
>
> 
>
>Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ 
>
>
>
>  
>



[Non-text portions of this message have been removed]

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Kevin 
Gulstene <kevin@d...> wrote:
> Austin-
> 
> Thanks for that.
> 
> >
> >> Why would the colour of the filter affect the relative 
intensities of
> >> the light hitting the CCD.
> >
> > Right or wrong, that doesn't have a thing to do with the 
reason.  It's 
> > the
> > property of the CCD response/artifacting to different color 
lights.  
> > Each of
> > the three lights gives different responses to the CCD and 
has different
> > artifacting.  Red, because it has the highest energy of the 
three, 
> > tends to
> > be the fuzziest, simply because of what are called bloom and 
smear.  
> > Bloom
> > is basically saturation of the sensing element, smear is 
basically 
> > crosstalk
> > between adjacent sensing elements, because of the 
intensity (this is 
> > the
> > artifact that PMT scanners do NOT suffer from, as they scan 
one "spot" 
> > at a
> > time).  Blue is the next worst, then green is the best...but not 
> > always.
> 
> I think I understand this but why is the CCD more sensitive to 
smear 
> and bloom associated with red in the absence of blue and 
green.  That 
> is, why would these artifacts appear when the light is filtered 
and not 
> appear when it goes through a ND filter?
> 
> I am interested in this because I have recently started 4x5 and 
since 
> my Umax powerlook 1100 is quite soft I'll soon be spending 
some more 
> money on another  scanner.  I'm trying to appreciate how big a 
> difference this makes in the overall scheme of things.
> 
> --
> Kevin Gulstene

I don't buy into Austin's argument.  If you scan with white light
passing thru the film and illuminating the CCD you have the
artifacts, fuzziness, blooming whatever of the entire visible
spectrum all added together.  You have the worst case.  If
there are some frequencies (color) that have a better response
you won't take advantage of it.  On the other hand if you scan
in RGB you at least have the possibility of picking which
third of the visible spectrum gives best results.  And even if
you just average the three values you ought to get better
results than just one value from a white light scan  -- like a
builtin multi-pass scan.

Kevin, as to the practical question about a 4x5 scanner, I've
used an Epson 1600 for 4x5 negs to print on an 1160 and
I can easily get as much quality out of the scan as I could
possibly get out of the printer.  Recently I purchased a bigger
printer, a 7500, and I just got an Epson 3200.  I expect them
to match very well but haven't done any definitive tests yet.

Roy
www.harrington.com

Truman,

> Quantum efficiently  has nothing to do with it.

Well, yes it does.  Quantum efficiency is the ratio  of the number of
photogenerated electrons captured by a sensor to the number of photons
incident upon the sensor during a period of time.  What that basically means
is that red, because of it's %80 quantum efficiency over a VERY broad
spectrum will fill the well much faster than the other two colors, which
means red will smear and bloom first, given the same amount of light energy
of each color.

> Red is at the low end
> or the visible frequency spectrum. Normally the higher the frequency the
> wider the bandwidth of filters at that frequency, unless you use a local
> frequency reference to down convert the signal. Why is this not the case
> for red light which is a lower frequency than blue light? Different
> materials for CCD's have different frequency reposes. I guess I don't
> understand why "red" produces more energy output.

Because the CCD is most sensitive to it, which is what quantum efficiency
is!

> The energy of a photon is Planks constant X it frequency. A photon of
> blue has more energy than a photon of red.

That doesn't matter.  What matters is how the CCD sees it, which is quantum
efficiency...

> Different scanners use
> different light sources. I am still at a loss - not that I am disputing
> it - with your statement. I just don't understand.

Light sources have nothing to do with it.  The light is filtered by either
the RG or B filter, and the CCD reacts to that color light only.  I don't
understand why you think that matters?  All the light sources cover the
spectrum that is being detected.

Austin

Roy,

> I don't buy into Austin's argument.
> If you scan with white light
> passing thru the film and illuminating the CCD you have the
> artifacts, fuzziness, blooming whatever of the entire visible
> spectrum all added together.

The neutral density filter reduces the entire spectrum to an even level.
There is no bloom and smear with the ND filter, and the response is uniform
and not peaky, as it is with blue and green.

> You have the worst case.

I disagree.  You have a best case, as the spectrum is now even.  It's easy
to see that the scans come out superior using the ND filter to scanning with
RGB, this isn't arguable.  Comparing individual channels, the green is
typically the closest, and sometimes the blue, but the ND scan is sharper,
and has better tonality.

> If
> there are some frequencies (color) that have a better response
> you won't take advantage of it.

Why do you believe you want to do that?

> On the other hand if you scan
> in RGB you at least have the possibility of picking which
> third of the visible spectrum gives best results.  And even if
> you just average the three values you ought to get better
> results than just one value from a white light scan  -- like a
> builtin multi-pass scan.

Well, it doesn't work that way.  Why don't you simply take a look at a scan
of a B&W negative done in RGB and thumb through the channels as I suggested
in PS.  You will clearly see that the red channel is the fuzziest.  This is
just a very well known issue with CCD sensors.

Austin

Anthony,

>"Simulate" is the operative word here.  You can make something 
>look like Tri-X or Tech Pan, but if you compare it with the 
>same scene actually photographed with these films, you may be 
>surprised.  Almost all the information required to reproduce the 
>response of most B&W films is missing from color images.

>You can certainly produce fine prints; but if you want prints 
>that look exactly the way a certain film looks, you must 
>capture the original scene with that film.
 
I've been following this thread with great interest.  It's a subject
that has concerned me for a long time, as I've hinted at in past posts
about my fear of losing the look and feel of particular films.  You
have finally brought to the table a credible explanation for what I
have only suspected.

I don't have a high end digi cam, but do have a 4mpx model with a
decent lens, and so have been able to try making BW prints from a
few images.  I even tried using the demo version of the software that
emulates different films, and it's quite good.  The few images I
worked up were sharp, well rendered, and made prints that were
basically "good" in many ways, yet failed to fully satisfy by not
"pressing my button" in the way my scans of Tri-X negs have.

I'm not giving up hope for an all-digital workflow.  Perhaps better
cameras, software, and my own improving skills will make it possible
someday.  In the meantime I'll save my shekels and shoot film.  6x7
Tri-X negs made with Pentax 67 lenses are hard to beat.

Thanks very much for the great input.

Regards,
Clayton


Info on black and white digital printing at    
http://www.cjcom.net/digiprnarts.htm

Austin-

I have a couple of questions remaining but it seems likely that this 
thread is a little esoteric -- my bad.  If you don't mind I'll send 
them to you directly.

--
Kevin

Hi Kevin,

By all means.

Regards,

Austin

> -----Original Message-----
> From: Kevin Gulstene [mailto:kevin@...]
> Sent: Thursday, May 22, 2003 1:52 AM
> To: DigitalBlackandWhiteThePrint@yahoogroups.com
> Subject: Re: [Digital BW] Why is ND B&W scan better -- was Digital,
> film, scanning compar
> 
> 
> Austin-
> 
> I have a couple of questions remaining but it seems likely that this 
> thread is a little esoteric -- my bad.  If you don't mind I'll send 
> them to you directly.
> 
> --
> Kevin

Peter writes:

> Not if you're scanning black and white film.

If you are scanning black and white, you get the spectral sensitivity of the
original B&W capture.  No real difference, except that even less further
manipulation of the image is possible.

> That shape, furthermore, reflects the spectral
> sensitivity of the film, not the scanning
> device.

True for both RGB and B&W capture.

Clayton writes:

> It's a subject that has concerned me for a
> long time, as I've hinted at in past posts
> about my fear of losing the look and feel
> of particular films.

I worry about it, too.  There is a very widespread but mistaken belief that
the images produced by any B&W film can be duplicated by some magic
manipulation of a color image in Photoshop.  That just isn't possible.  In
fact, you cannot even duplicate the images produced by a different color
film or a different electronic sensor in Photoshop.  This is just one of the
many misunderstandings that seem to plaque digital photographers, who
typically know something about photography but nothing about information
theory, even though information theory now intrudes so ubiquitously into
digital photography that it cannot be ignored.

> The few images I worked up were sharp, well
> rendered, and made prints that were basically
> "good" in many ways, yet failed to fully satisfy
> by not "pressing my button" in the way my scans
> of Tri-X negs have.

The converted B&W images will never exactly match Tri-X or any other B&W
film.

> I'm not giving up hope for an all-digital workflow.

Imaging workflows are never all-digital; that's another widespread
misconception of photographers.  The image-capture phase is always analog,
and so is the printing or display phase.  Only the part in between can be
digitized.  If you take photographs with a so-called "digital" camera, you
are not capturing images digitally; you are capturing images with an analog
electronic sensor, the signals from which are then sampled to produce
digital output.  Likewise, when you print on an ink-jet printer (or display
on a monitor, or print on any other device), you are converting digital data
back to analog signals.

The ultimate results of any imaging workflow will always be a function of
the quality of the analog components of the system at each end.  Most of the
alleged quality of digital photography comes not from the fact that digital
cameras produce digital output, but from the fact that they use electronic
sensors that have a number of inherent advantages over film.

A corollary of this is that no digital system can ever exceed the quality of
the best analog system--because any system that interacts with the real
world must contain at least a few analog components at the interface points.

> Perhaps better cameras, software, and my own
> improving skills will make it possible
> someday.

No, it will never be possible, as it is a restriction imposed by information
theory; like the obstacle of diffraction limits in lenses, it cannot be
surmounted.

> In the meantime I'll save my shekels and shoot film.  6x7
> Tri-X negs made with Pentax 67 lenses are hard to beat.

I like Tri-X better in MF, because it has the same great look in the way it
handles light, but it shows much less grain.  Also, the MF flavor of Tri-X
has a bit more shadow detail.

> I worry about it, too.  There is a very widespread but 
> mistaken belief that the images produced by any B&W film can 
> be duplicated by some magic manipulation of a color image in 
> Photoshop.

There are other problems with this approach as well. If you use the
channel mixer and make large adjustments in the relationships between
colors you often end up with grainy images, loss of detail and halos
around objects in the transition areas between colors.

Here's an example:
http://daniel.staver.no/img/channelmixer1.jpg

The picture is a high quality jpeg from a Minolta Dimage 7i (which I
sold again last year in favor of my old analogue cameras btw).

Picture 1: Here I've used channel mixer to tone down the blue and
brighten red. I like contrasty images, so for me this is a look I'd
often like in my pictures.

Picture 2: Here I've made a mask of the flower (a really poor mask, but
it's just an example), converted to grayscale and applied curve
adjustment to the image with the mask in place to simulate adjustments
similar to what I did with the channel mixer. 

Picture 3: The original image. 

The result in the second image is much better than the first since
detail from all three channels is combined, and adjustments are then
made while the channel mixer simply discards information from channels
you don't use which results in exaggerated grain and loss of detail.

I notice the same effects on 16bit scans I make with my FS4000 and then
convert to BW using channel mixer, so the problem is not limited to
JPEG's from digital cameras.

So all technicalities aside, these problems have made me wary of doing
big tonal adjustments with the channel mixer, and I've mostly switched
from trying to make good BW from color film and now try to make it right
with BW film and filters when I record the image instead.

--
Daniel Staver
http://daniel.staver.no

Anthony,

> I worry about it, too.  There is a very widespread but mistaken
> belief that
> the images produced by any B&W film can be duplicated by some magic
> manipulation of a color image in Photoshop.  That just isn't possible.

Well, it is possible to come up with very good images converting color to
B&W, and what "qualities" and/or "characteristics" someone else may deem
"duplicate", may not be the ones YOU deem, but that doesn't make their
belief incorrect.

Another issue is, why is duplicate important at all?  Why not take it for
what it is, and if you like it, so be it, and if you don't, so be it.  No
doubt it's different, but does it give you the results you want is what
should be what is important, not trying to "duplicate" something.

> > The few images I worked up were sharp, well
> > rendered, and made prints that were basically
> > "good" in many ways, yet failed to fully satisfy
> > by not "pressing my button" in the way my scans
> > of Tri-X negs have.
>
> The converted B&W images will never exactly match Tri-X or any other B&W
> film.

Of course not, but it certainly may give someone an image that they are very
happy with, and has all the qualities that they want in an image.  If you
want Tri-X, shoot Tri-X, but not everyone wants the look of Tri-X.  Also,
Tri-X is simply not Tri-X.  It is very dependant on the development, and
therefore one persons Tri-X will not be the same as another's Tri-X.

Austin

Austin writes:

> Well, it is possible to come up with very good
> images converting color to B&W, and what "qualities"
> and/or "characteristics" someone else may deem
> "duplicate", may not be the ones YOU deem, but
> that doesn't make their belief incorrect.

And vice versa.  Sure, you can convert color to B&W with interesting
results, but that doesn't help if they are not the results that you want.

> Another issue is, why is duplicate important
> at all?

The reason photographers shoot B&W film in the first place is that they wish
to obtain a certain "look" that is unique to B&W film (and sometimes to a
_specific_ B&W film).  I don't see any reason why they should be expected to
sacrifice all of these flexibility and choice in favor of some color-to-B&W
half-solution.

> Why not take it for what it is, and if you like it,
> so be it, and if you don't, so be it.

Why give up what photographers already have?  There's no _advantage_ to
converting from color to various ersatz B&W renderings.  What's wrong with
continuing to shoot B&W in the first place?

> ... does it give you the results you want is what
> should be what is important, not trying to "duplicate"
> something.

If the results I want are Tri-X, then it must _duplicate_ Tri-X.  I'm not
interested in "almost as good."  And I know that many other photographers
feel the same way, which is why they continue to put real B&W film into
their cameras.

> Of course not, but it certainly may give someone
> an image that they are very happy with ...

Then again, it may not.  And in that case, they must shoot B&W to begin
with, or they won't be happy.  Why does that seem to be a problem?

> If you want Tri-X, shoot Tri-X, but not everyone
> wants the look of Tri-X.

Those who don't are not obligated to shoot the film, nor would I presume to
suggest that they do so.

> Also, Tri-X is simply not Tri-X.  It is very dependant
> on the development, and therefore one persons Tri-X will
> not be the same as another's Tri-X.

All the more reason to shoot Tri-X to begin with.  That might be difficult
to simulate from color.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Austin 
Franklin" <darkroom@i...> wrote:
> Roy,
> The neutral density filter reduces the entire spectrum to
> an even level. There is no bloom and smear with the ND 
> filter, and the response is uniform and not peaky, as 
> it is with blue and green.

Whoa!   A neutral density filter does NOT "even out" the spectrum of 
anything.   That's why it's called "neutral" - it passes all 
wavelengths equally.  If something came it peaky blue or green it 
comes out peaky blue or green.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony 
Atkielski" <anthony@a...> wrote:
> Clayton writes:
> 
> > It's a subject that has concerned me for a
> > long time, as I've hinted at in past posts
> > about my fear of losing the look and feel
> > of particular films.
> 
> I worry about it, too.  There is a very widespread but mistaken 
belief that
> the images produced by any B&W film can be duplicated by some magic
> manipulation of a color image in Photoshop.

Why should this be so hard?   Every black and white film has a 
characteristic response curve, and that curve can be characterized 
for R, G, and B (because even modern "panchromatic" films are not 
perfectly panchromatic).    So if you start off with a wide-latitude 
color film, like Portra, what limitations do you bump up against 
trying to simulate a given B+W film? 

It seems like all you have to do is compute a scalar from your 
source (color film) response curve to your target (B+W) response 
curve for each intensity level.   When you're done you have a vector 
of scalar values and that vector is the curve you use for your 
conversion.    What would NOT be described by that curve?   It would 
tell you how much shadow detail you get, how highlights look, how 
smooth or even midtones are, etc.

Peter,

> > Roy,
> > The neutral density filter reduces the entire spectrum to
> > an even level. There is no bloom and smear with the ND
> > filter, and the response is uniform and not peaky, as
> > it is with blue and green.
>
> Whoa!

Whoa your self!

> A neutral density filter does NOT "even out" the spectrum of
> anything.   That's why it's called "neutral" - it passes all
> wavelengths equally.

Yes, that's what I said.  I know how a ND filter works, thanks.  It PRESENTS
(what ever word you want to use) an even spectrum to the CCD, instead of
having colored filters which present basically a band pass in a bell curve
shape to the CCD.

> If something came it peaky blue or green it
> comes out peaky blue or green.

You don't understand.  The peaky blue/green is the CCD response to that
specific color.  With an ND filter, the CCD response is not peaky to any
colors.

Austin

Peter writes:

> Why should this be so hard?

I've explained this several times.  Apparently I lack the necessary talent
to make the explanation understandable.

> Every black and white film has a characteristic
> response curve, and that curve can be characterized
> for R, G, and B ...

No, it cannot.  The curve is just that: a curve.  The RGB values are just
three numbers--equivalent to three straight lines of different heights in
the spectrum.  All the rest of the curve between those three lines is lost
when a picture is taken in color.  Because of this, the original curve of
the original scene CANNOT be recreated from the RGB values.  And since the
original curve of the original scene MUST be recreated in order to proper
render a scene in B&W based on the sensitivity curves of a black and white
film, it is impossible to recreate the look of a scene capture in one way
with the look of a scene captured in another way.

The rule is much more general than this, in fact.  It is impossible to
translate any scene represented as a single set of monochrome values or as a
set of R,G,B values into any other scene represented in the same way,
because you MUST HAVE the original curves of the original scene in order to
effect the transformation, and you cannot get these from just one number, or
even from just three numbers.

Additionally, the rule applies indifferently to color film and B&W film, and
to color electronic sensors and B&W electronic sensors.

> So if you start off with a wide-latitude
> color film, like Portra, what limitations
> do you bump up against trying to simulate
> a given B+W film?

See above.  Portra produces a triplet of RGB values based on its own unique
sensitivity curve.  Once the film has done this, the original spectral
characteristics of the original scene are lost.  To create a rendering that
matches what some other film would have produced, you must have the original
spectrum of the original scene; three numbers won't do.

> It seems like all you have to do is compute
> a scalar from your source (color film) response
> curve to your target (B+W) response curve for
> each intensity level.

But you no longer have each intensity level.  You have only three levels.
Everything else in between is lost.

I've already provided an example.  Let me try again.

You have a special film (or a special filtered CCD) that is sensitive only
to the yellow light of low-pressure sodium-vapor lamps.  You take a picture
of a scene lit with these lamps in some areas, and lit with a blend of red
and green light in other areas.  You then take the same picture using an
ordinary color film or color CCD.

What will you get?  On the special yellow-sensitive film, the areas lit by
LPS lamps will appear bright, and all other areas will appear dark.  On the
color film, however, all areas will appear equally bright.

There is _no way_ to convert the RGB image from the color film in a way that
will duplicate the results you got from the yellow-sensitive film, because
the important information on the spectral distribution of the original light
is gone.

> What would NOT be described by that curve?

See above.

This rule applies even to color-to-color transformations.  You cannot
convert Portra to make it look just like Velvia, because part of the
spectral information on the original scene is lost and cannot be
reconstructed to make the transformation.

Anthony,

> > Every black and white film has a characteristic
> > response curve, and that curve can be characterized
> > for R, G, and B ...
>
> No, it cannot.  The curve is just that: a curve.  The RGB values are just
> three numbers--equivalent to three straight lines of different heights in
> the spectrum.

Huh?  And yes, it can be characterized, to some degree of accuracy.  To
characterize it for RGB, you would use THREE curves, one for each color, not
just three numbers!

> All the rest of the curve between those three lines is lost
> when a picture is taken in color.

I don't believe you know what you're talking about.  The three colors (which
are actually CMY), have overlapping regions.  They are not brickwall
cutoffs, they taper off on the ends.  They in fact DO encompass the entire
spectrum that B&W film does as well.  The spectral response of Portra 400VC
is 375nm to 680nm.  The spectral response of Tri-X is 400nm to 650nm.  So,
in fact, Portra has a higher spectral response.

If the spectral response of your color film encompasses that of the
grayscale (which it does), and the film response is deterministic (which it
is), and the conversion to RGB is also deterministic (which it is, and vice
versa), then it can be characterized.

Like you said, this is not mystical, it is purely good science, and
obviously an understanding of how this really works without simply guessing.
The operative words being good, understanding, and not guessing.

Austin

Austin writes:

> I don't believe you know what you're talking about.

Read my example in my post to Peter with respect to yellow-sensitive film,
and tell me how you'd duplicate the response of this film in a conversion
from RGB.

If RGB were sufficient for transformations from any curve to any other
curve, then it would be impossible for us to see yellow by mixing red and
green.  Since, however, we _do_ see yellow by mixing red and green, clearly
there is information lost in the conversion to RGB.  Once a yellow surface
is photographed in RGB, we can never know if it was really yellow, or just a
blend or red and green.  This proves that the conversion involves an
irreversible loss of information, and so it is impossible to reconstruct the
curve of any other film or sensor from RGB alone.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Austin 

> > If something came it peaky blue or green it
> > comes out peaky blue or green.
> 
> You don't understand.  The peaky blue/green is the
> CCD response to that> specific color.  With an ND filter,
> the CCD response is not peaky to any colors.

If that's what you meant then why bother with any filter
at all?

Not that I agree with the whole of his argument, but I belive what Anthony 
is trying to explain is this:

I give you an RGB value, say.. 0,0,120 .. nice and medium blue.

Can you draw the original spectral input to the imaging site?  Was it a 
very bright blue at the bottom end of the film's blue response or a 
moderately bright blue just above that?

Read:  the process of binning is not reversible.

It seems to me that we need to clear up the language that we are using. 
There are a lot of technical terms being thrown around in non-technical 
ways, only confusing the issue.  We would also do well to get rid of the 
argumentative language and personal jabs.

-Jon

Anthony,

> > I don't believe you know what you're talking about.
>
> Read my example in my post to Peter with respect to yellow-sensitive film,
> and tell me how you'd duplicate the response of this film in a conversion
> from RGB.

I've not got time to waste with this.  I KNOW that was I said is correct and
based on fact.  You really need to go look at the spectral response curves
for color and B&W films, and you will see what I said is true.

> If RGB were sufficient for transformations from any curve to any other
> curve...

No, I did not say that, and in true Anthony form, you are putting words in
my mouth to somehow try to convince your self that you aren't wrong.  I said
that the spectral response of Portra is sufficient to encompass Tri-X, which
is what the issue is, period, and my statement is purely fact, and not
disputable.

Austin

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony 
Atkielski" <anthony@a...> wrote:
> Peter writes:
> 
> > Why should this be so hard?
> 
> I've explained this several times.  Apparently I lack the 
necessary talent
> to make the explanation understandable.
> 
> > Every black and white film has a characteristic
> > response curve, and that curve can be characterized
> > for R, G, and B ...
> 
> No, it cannot.  The curve is just that: a curve.  The RGB values 
are just
> three numbers--equivalent to three straight lines of different 
heights in
> the spectrum.  All the rest of the curve between those three lines 
is lost
> when a picture is taken in color. 

The spectral response of both the film and the filters used in CCD 
Bayes patterns overlap (as do the photopigments in the human eye) so 
I'm not sure what you mean by "in between".  Are you saying there 
are wavelengths of light that fall between the response wavelengths 
of the color film (or CCD element filters) that the black and white 
picks up?   Could you give a specific example?


> Because of this, the original curve of the original scene 
> CANNOT be recreated from the RGB values.  And since the
> original curve of the original scene MUST be recreated 
> in order to proper render a scene in B&W based on the
> sensitivity curves of a black and white film, 

I don't know what you mean by the "curve of the original scene".  
All you need to care about is the curve(s) of the recording medium.  
That's provided by knowing the response curve of the color film in 
each channel.   I don't understand what information you think is 
missing.   You keep talking about data "in between" the R,G, and B 
values but those sensitivy curves overlap.

> > > If something came it peaky blue or green it
> > > comes out peaky blue or green.
> >
> > You don't understand.  The peaky blue/green is the
> > CCD response to that> specific color.  With an ND filter,
> > the CCD response is not peaky to any colors.
>
> If that's what you meant then why bother with any filter
> at all?

Peter,

Because it will saturate the CCD.  Using stronger light can give more even
illumination, and it can be diffused better.

Austin

Peter writes:

> The spectral response of both the film and
> the filters used in CCD Bayes patterns overlap
> (as do the photopigments in the human eye) so
> I'm not sure what you mean by "in between".

I simplified the illustration.  The fact that they overlap does not change
the principle, or the constraints.

My example of the yellow-sensitive film illustrates this.

If RGB were sufficient to represent an entire spectrum from the original
scene, then none of the color systems used by human beings would work,
because they all depend on the inability of an RGB system to distinguish
certain spectral characteristics.  There is no way to distinguish between
monochromatic yellow and a mixture of red and green for example, but in real
life there _is_ a difference ... it's just that the information required to
detect that difference is missing from RGB.

> Are you saying there are wavelengths of light that
> fall between the response wavelengths of the color
> film (or CCD element filters) that the black and white
> picks up?

No, I'm saying that three numbers do not tell you which intensities at which
frequencies produced the RGB values.  Each number is a summation of many
different frequencies of light, and all distinction between them is lost.
But since you need to preserve this distinction in order to reconstruct the
curves you need to translate from one colorspace to another, the net result
is that RGB cannot be used to convert from one color or B&W film or sensor
to another.

> Could you give a specific example?

I already did.  I'll copy and paste it here:

You have a special film (or a special filtered CCD) that is sensitive only
to the yellow light of low-pressure sodium-vapor lamps.  You take a picture
of a scene that is lit with these lamps in some areas, and lit with a blend
of red and green light in other areas.  You then take the same picture using
an ordinary color film or color CCD.

What will you get?  On the special yellow-sensitive film, the areas lit by
LPS lamps will appear bright, and all other areas will appear dark.  On the
color film, however, all areas will appear equally bright.

There is _no way_ to convert the RGB image from the color film in a way that
will duplicate the results you got from the yellow-sensitive film, because
the important information on the spectral distribution of the original light
is gone.

> I don't know what you mean by the "curve of the original scene".

If you pass the light from a spot in the original scene through a prism,
you'll see the spectral distribution of the light from that scene.  That's
what you need in order to convert from one colorspace to another.  Just an
isolated red, green, or blue value won't do.

> All you need to care about is the curve(s) of
> the recording medium.

The curve of the recording medium won't do you any good unless you have the
curve of the original scene as well.

> I don't understand what information you think is
> missing.

The spectral distribution of the light in the original scene.

> You keep talking about data "in between" the R,G, and B
> values but those sensitivy curves overlap.

They overlap, but that isn't enough.  If that were enough, then there would
be virtually no such thing as color blindness (which typically involves only
a slight different in sensitivity curves), and no RGB color system would be
capable of producing the entire rainbow of colors for human vision.

Austin writes:

> I've not got time to waste with this.  I KNOW that
> was I said is correct and based on fact.

In that case, you should be able to easily answer my question.  How would
you reconstruct the light and dark rendering of the yellow-sensitive film
from an RGB image?  Remember, RGB cannot distinguish between yellow light
and a blend of red and green light, so the light from the original scene
will seem equally bright everywhere.  In the yellow-sensitive film, however,
only the truly yellow-lit areas will seem bright.  There is _no way_ to
transform the RGB in a way that will produce the results obtained with the
yellow-sensitive film.  It simply cannot be done.

Feel free to prove me wrong, if you can.

> You really need to go look at the spectral
> response curves for color and B&W films, and
> you will see what I said is true.

I have.  None of them are regular enough to be reproduced from three data
points.  If all of them were points on regular ellipses, it could be done,
but they aren't, so it's not possible.

> I said that the spectral response of Portra is
> sufficient to encompass Tri-X ...

If you mean that it will respond to the same frequencies, that may be true
(I'd have to look).  But since it does not have the same response curves, an
RGB image from Portra cannot be converted into a B&W image from Tri-X.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony 
Atkielski" <anthony@a...> wrote:


> If RGB were sufficient for transformations from any 
> curve to any other curve, then it would be impossible
> for us to see yellow by mixing red and green. 

I'm sorry, but this is a FASCINATING statement!   Could you 
elaborate? 

We see yellow because the photopigments for red and green in the 
retinal cones both happen to overlap their sensitivity in the range 
of wavelengths we call 'yellow', between about 540 and 560 nm.  


> Since, however, we _do_ see yellow by mixing red and green, clearly
> there is information lost in the conversion to RGB. 

Clearly there is information lost in your understanding of the 
physiology of color perception.


> Once a yellow surface is photographed in RGB, we can never
> know if it was really yellow, or just a blend or red and
> green. 

If a blend of red and green produces the perception of yellow it IS 
yellow.   If you take a photo of a scene with an RGB monitor in it 
with a yellow screen you know that the yellow is created from red 
and green phosphors.  Are you saying that black and white film will 
record this differently from an identical yellow created by say 
mixing cadmium yellow paint? 

The only way what you're saying could occur is if the sensing 
elements (whether film dyes, photopigments, or CCD filters) were 
NONoverlapping.   Imagine an RGB sensor that ONLY saw 640 nm, 500 nm 
and 440 nm.   If you had something in your scene at 560 nm you would 
see yellow, amd TMax film would record a shade of gray, but the RGB 
sensor would see black.

What you're probably thinking of is metamerism.  That happens when 
you have a light source illuminating a pigment or dye whose 
reflectance spectrum has peaks or valleys with respect to the 
emission spectrum of the light source.    But that's not a function 
of the recording medium.  Using the above numbers, if I have LEDs of 
440, 500, and 650 nm illuminating a color patch that only reflected 
at 560 nm it would look black.  It would also photograph black with 
all film and with all digicams.

Anthony,

I worry that you seem to be hooked on one particular artificial
interpretation of the world - that produced by Tri-X film, and seem to
denigrate other artificial interpretations. Others may like some of the
other interpretations, whether produced in Photoshop or some other chemical
mix in a plastic film. I don't see why one interpretation is any more valid
than any other - apart from the one fact that you like it.

Bob Frost.

----- Original Message ----- 
From: "Anthony Atkielski" <anthony@...>
>
> > It's a subject that has concerned me for a
> > long time, as I've hinted at in past posts
> > about my fear of losing the look and feel
> > of particular films.
>
> I worry about it, too.  There is a very widespread but mistaken belief
that

> the images produced by any B&W film can be duplicated by some magic
> manipulation of a color image in Photoshop.  That just isn't possible.  In
> fact, you cannot even duplicate the images produced by a different color
> film or a different electronic sensor in Photoshop.

----- Original Message ----- 

From: "Anthony Atkielski" <anthony@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, May 22, 2003 6:16 PM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> The rule is much more general than this, in fact.  It is
impossible to
> translate any scene represented as a single set of monochrome
values or as a
> set of R,G,B values into any other scene represented in the
same way,
> because you MUST HAVE the original curves of the original scene
in order to
> effect the transformation, and you cannot get these from just
one number, or
> even from just three numbers.
>
> Additionally, the rule applies indifferently to color film and
B&W film, and
> to color electronic sensors and B&W electronic sensors.

etc etc.

Anthony,

It is all very correct. In practice however the narrow band
filters are not that narrow, our eyes and brain are based on tri
stimuli as well (but a bit more complicated), and a B&W picture
made with a narrow band yellow filter looks weird anyway. So that
means that much can be done afterwards on an RGB image and we are
not selective enough to see the difference. If it has to be
exotic then digital can do as much or more but not exactly the
same. So that digital replication of the Tri X curve will satisfy
a lot of us, to some it means an extra taste next to the original
and to you it never will be the same. And for me it means that I
never get that chrome yellow or deep red in the painting exactly
reproduced in the print. In science you can't get away with that,
it could mean that the big bang was twice as long ago.

Ernst

Peter writes:

> We see yellow because the photopigments for
> red and green in the retinal cones both happen
> to overlap their sensitivity in the range
> of wavelengths we call 'yellow', between
> about 540 and 560 nm.

Correct.

> Clearly there is information lost in your
> understanding of the physiology of color perception.

No, my understand is correct.

> If a blend of red and green produces the perception
> of yellow it IS yellow.

To human eyes, yes ... but to other types of image capture devices, no.

I suppose I should return to my example of the yellow-sensitive film, the
one that everyone persists in ignoring.

Truly yellow light, from a physics standpoint, falls within a narrow band of
wavelengths.  It is distinct from the blend of red and green light that
human beings perceive as yellow.  However, human beings cannot tell the
difference between this "real" yellow and the "perceptual" yellow made by
mixing red and green.  Similarly, all color display and printing systems for
human beings are designed around this defect of human vision, and so they
cannot make this distinction, either.

However, it's certainly possible to develop an imaging device that is
sensitive only to spectral yellow, and does not respond to red or green.
Such a device, when presented with a scene that contains both
spectral-yellow-lit areas and areas lit in red and green, will see only the
spectral yellow areas as being lit.

Now, the problem arises when an RGB-based capture system records this scene.
The system will see all of the scene as being equally lit, because it cannot
distinguish between spectral yellow and a blend of red and green.  The RGB
scene thus recorded will therefore show all areas of the scene lit equally.

Given this, it should be obvious why the RGB scene cannot be transformed
into a correct representation of the scene as it is recorded by the
yellow-sensitive device described above.  The distinction between spectral
yellow and blended red and green, which would have been recorded by the
yellow-sensitive device, disappears when the scene is recorded in RGB.
Thereafter, no amount of manipulation can change this.

This illustration is a very simple example, but the principle applies
equally in much more complex situations.  And this is why one RGB
representation cannot generally be transformed into another, nor can it be
transformed into B&W representations with complete accuracy.  Too much
information is gone.

> If you take a photo of a scene with an RGB monitor
> in it with a yellow screen you know that the yellow
> is created from red and green phosphors.

Yes.  But if there is a yellow box next to the monitor, you do not know
whether the box is truly yellow (that is, reflecting light at the frequency
of spectral yellow) or just appears to be that way because it reflects red
and green equally.

> Are you saying that black and white film will
> record this differently from an identical yellow
> created by say mixing cadmium yellow paint?

YES!!  (If cadmium yellow is a spectral yellow, and I think it is.)

B&W film--as well as color film and CCDs--records light differently based on
its frequency, not based on how it might be represented in RGB.  If the
sensitivity peaks at spectral yellow frequencies but is lower and unequal at
red and green frequencies, the B&W film will render different "types" of
yellow with different luminosities.

This is what gives different films and different imaging devices their
distinctive palettes.  And this is also why no accurate conversion between
palettes is possible using simple RGB information alone.

> The only way what you're saying could occur is
> if the sensing elements (whether film dyes, photopigments,
> or CCD filters) were NONoverlapping.

No, it happens even for overlapping sensitivities, as I have explained for
the umpteenth time above.

> Imagine an RGB sensor that ONLY saw 640 nm, 500 nm
> and 440 nm.   If you had something in your scene
> at 560 nm you would see yellow, amd TMax film would
> record a shade of gray, but the RGB sensor would see
> black.

Correct.  But this principle also holds when the RGB sensitivity is a group
of curves, instead of monofrequency sensitivity as in your example.

> What you're probably thinking of is metamerism.

No, what I'm thinking of is what I've described here in detail, over, and
over, and over.  Isn't there anyone who understands this?

> Using the above numbers, if I have LEDs of 440,
> 500, and 650 nm illuminating a color patch that
> only reflected at 560 nm it would look black.  It
> would also photograph black with all film and with all digicams.

Yes.  But I'm not talking about illumination, I'm talking about capture.

Bob writes:

> I worry that you seem to be hooked on one particular
> artificial interpretation of the world - that produced
> by Tri-X film, and seem to denigrate other artificial
> interpretations.

Not at all.  I'm just trying to make clear that you cannot transform one
interpretation into another using mere RGB or grayscale data.

Ernst writes:

> In practice however the narrow band filters are
> not that narrow ...

They don't have to be.  The problem exists no matter how broad the filtering
is.

> ... our eyes and brain are based on tri stimuli
> as well (but a bit more complicated) ...

That's why we cannot tell the difference between yellow and a mix of red and
green.  An alien species might wonder why we insisted that both were the
same, when they could easily see the difference.

> So that digital replication of the Tri X curve
> will satisfy a lot of us ...

It might.  But it is very important to understand that Tri-X cannot be
_duplicated_ by any manipulation of RGB.  This is one of the fundamental
notions that seem to escape many photographers, and as a result they reach
all sorts of wildly inaccurate conclusions about what digital photography
can or cannot do.

C'mon guys, give it a rest...

Roger

  -----Original Message-----
  From: Anthony Atkielski [mailto:anthony@...]
  Sent: Thursday, May 22, 2003 12:53 PM
  To: DigitalBlackandWhiteThePrint@yahoogroups.com
  Subject: Re: [Digital BW] Digital, film, scanning comparisons


  Ernst writes:

  > In practice however the narrow band filters are
  > not that narrow ...

  They don't have to be.  The problem exists no matter how broad the
filtering
  is.

  > ... our eyes and brain are based on tri stimuli
  > as well (but a bit more complicated) ...

  That's why we cannot tell the difference between yellow and a mix of red
and
  green.  An alien species might wonder why we insisted that both were the
  same, when they could easily see the difference.

  > So that digital replication of the Tri X curve
  > will satisfy a lot of us ...

  It might.  But it is very important to understand that Tri-X cannot be
  _duplicated_ by any manipulation of RGB.  This is one of the fundamental
  notions that seem to escape many photographers, and as a result they reach
  all sorts of wildly inaccurate conclusions about what digital photography
  can or cannot do.



        Yahoo! Groups Sponsor
              ADVERTISEMENT




  Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:

  http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint

  If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.

  Please follow these basic guidelines:
  - Include your full name with your message.
  - Include the address of your website, if you have one.
  - As threads develop, trim off excess portions of earlier messages to keep
them short.
  - As the topic of a thread changes remember to change the subject header.
  - Good manners are required at all time. No personal attacks or flames
  - Complete your Yahoo profile.
  - Before posting a question, search the message archives and the various
resources on the homepage.




  Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.


[Non-text portions of this message have been removed]

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony 
Atkielski" <anthony@a...> wrote:

> However, it's certainly possible to develop an imaging
> device that is sensitive only to spectral yellow, 
> and does not respond to red or green. Such a device, 
> when presented with a scene that contains both
> spectral-yellow-lit areas and areas lit in red and 
> green, will see only the spectral yellow areas as being
> lit.

But that's just as artificial as the single-wavelength LED example I 
used.   You have not demonstrated this has any bearing on the real 
world problem of mapping color film to black and white.    I asked 
you before to identify some wavelength that would "fall between" 
sensitivity of some color-triplet system (color film or digicam) 
that would get recorded on b+w film.





> > Are you saying that black and white film will
> > record this differently from an identical yellow
> > created by say mixing cadmium yellow paint?
> 
> YES!!  (If cadmium yellow is a spectral yellow, and I think it is.)
> 
> B&W film--as well as color film and CCDs--records light
> differently based on its frequency, not based on how it
> might be represented in RGB.  If the sensitivity peaks
> at spectral yellow frequencies but is lower and unequal
> at red and green frequencies, the B&W film will render 
> different "types" of yellow with different luminosities.

Give a specific example of a wavelength that would fall betweeen the 
peaks in the manner you describe so we can test it.    I think 
you're describing something purely conjectural that has no bearing 
on the real world, but I'll be happy to be proven wrong.

Various colors do produce the same grayscale value.  But, using color film,
which records at least as wide a spectral response as B&W film does, can be
converted it to grayscale to give a the full range of grayscale values.  The
spectral response of both films is quite well documented (Provia and Tri-X
in the case of my other post), and it is therefore quite deterministic to
produce a mapping from the information on the color film, as produced by
it's spectral response, to the same spectral response that B&W film does.

What you can not do is the reserve...you can't have a grayscale value and
turn it into a color value, as said above, it could be any of a number of
color values that are equal to that grayscale value.  But, this in no way
mitigates the ability to closely replicate the response of B&W film from the
information on color film.

Austin

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Roger L 
Sopher" <rlsopher@c...> wrote:
> C'mon guys, give it a rest...

Why?  It's not off-topic.

This IS relevant to printing in general and black and white printing 
in particular.   The relationship between the reflectance spectra of 
pigments and colors we perceive is important to problems like 
metamerism, profile-matching of monitors and printers, and how to 
use CMYK (and its variants) to produce even grayscales.

Anthony raises some interesting questions.   Assuming that a Y 
pigment in an inkjet printer produces an actual yellow (say, 560 nm) 
what does it mean to "match" this to a yellow on a monitor which is 
really comprised of a red and a green?   If you go to my web page 
where I have enlarged  "grayscale" images showing the actual dots of 
ink laid down by a 2200 you can see all the different ways 
the "same" color is created.   How might these methods prove 
DIFFERENTLY sensitive to ambient lighting and how might that inform 
our choice of drivers or inks?

My undergraduate major was the neurophysiology of visual 
processing.  (imaging sticking single-cell recording devices in 
cat's brains).   So this is a very interesting topic, as well.  

Shortly after I joined the group some months back I complained that 
one difference between darkroom printers I knew and digital ones is 
that many darkroom printers I knew were genuinely interested in the 
chemistry and physics of their craft, whereas too many digital 
printers "black-boxed" everything.   When I have questions about the 
chemistry of developing and printing or the physical properties of 
emulsions and papers I can always find some old darkroom geek who 
knows the answer.   But when I have similar questions about the 
physical and chemical properties of inks and inkjet papers it's hard 
to find people here who know or even care.   Your complaint 
reinforces that stereotype.

Peter,
> Shortly after I joined the group some months back I complained that 
> one difference between darkroom printers I knew and digital ones is 
> that many darkroom printers I knew were genuinely interested in the 
> chemistry and physics of their craft, whereas too many digital 
> printers "black-boxed" everything.   When I have questions about 
the 
> chemistry of developing and printing or the physical properties of 
> emulsions and papers I can always find some old darkroom geek who 
> knows the answer.   But when I have similar questions about the 
> physical and chemical properties of inks and inkjet papers it's 
hard 
> to find people here who know or even care.   Your complaint 
> reinforces that stereotype.

Who cares what the physical properties of these materials are from a 
practical point of view. The properties are basically immutible and 
there is little prgamatic value to be gained by this knowledge. In 
other words, IMO, you are getting to low level in your knowledge 
quest as an end user to provide useful results.

Now if you are thinking about manufacturing a product, that's a 
different issue. 

But why not worry about making photographs instead of all this 
ancillary stuff?

Don Bryant

Anthony,

> But it is very important to understand that Tri-X cannot be
> _duplicated_ by any manipulation of RGB.

You are misunderstanding something, obviously.  Light is nothing but a
frequency and intensity.  Color is nothing but a response to a particular
frequency.  Show me anything on the spectral response curve of Tri-X that
can not be represented in RGB.  You can't.

An RGB sensor is completely capable of a spectral response that encompasses
the entire spectral response range of Tri-X.  The most recent CCD sensor I
designed with has a spectral response curve of 380-530nm for blue, 420-600nm
for green and 580-850+nm for red.  That gives an overall spectral response
of 380-850+ nm.  Tri-X has a spectral response of 400nm to just past 650nm.
Smaller than the CCD.

Austin

Peter writes:

> But that's just as artificial as the single-wavelength
> LED example I used.

It is a simplified illustration.

> You have not demonstrated this has any bearing
> on the real world problem of mapping color film
> to black and white.

I've tried to explain this in as many ways as I can.  Since I'm not making
any progress, I'll stop here.  Believe whatever you wish to believe; it will
affect your results, not mine.

Austin writes:

> ... it is therefore quite deterministic to
> produce a mapping from the information on the
> color film, as produced by it's spectral response,
> to the same spectral response that B&W film does.

Absolutely ... but this presumes that you have the original complete
spectrum to work from, not just RGB values.  Once the color film has
converted this broad spectral response to three RGB values, a conversion to
B&W based on this correlation between curves is no longer possible.

Anthony,

> > ... it is therefore quite deterministic to
> > produce a mapping from the information on the
> > color film, as produced by it's spectral response,
> > to the same spectral response that B&W film does.
>
> Absolutely ... but this presumes that you have the original complete
> spectrum to work from, not just RGB values.

Once again, RGB values from a CCD imaging device encompass the entire
spectral range that Tri-X encompasses, and therefore you DO have the
original complete spectrum to work from, with the "original" spectrum being
defined as the spectrum that Tri-X can record.  Good grief.

> Once the color film has
> converted this broad spectral response to three RGB values, a
> conversion to
> B&W based on this correlation between curves is no longer possible.

And, as I've said, is simply wrong.  You somehow believe that the spectrum
that RGB represents is smaller than the spectrum that Tri-X responds to, and
that is your mistake.

What you are confusing is the limiting of spectrum by differing color
spaces, and in the case of Tri-X, you will find it has a very limited
spectral response, and therefore RGB will encompass it and do exactly what
you claim it won't, reproduce it in it's entirety.

Austin

Hi Peter,

For the simple reason that your (plural) "discussion" has declined to the
level of oh yeah I'm right and you are wrong. One can only state the same
arguments so many times before it becomes obvious neither party is willing
to be convinced of the others position.

Just my opinion, I'm not the email police and and I can always click the X
on Outlook (at the danger of getting repetitive motion injury...)

My own scientific background isn't too shabby and possibly rivals your own.
Just for the record I am also an old "dark room geek...."

Roger

Roger L. Sopher, MD
Professor Emeritus of Pathology

  -----Original Message-----
  From: Peter Nelson [mailto:peter@...]
  Sent: Thursday, May 22, 2003 2:31 PM
  To: DigitalBlackandWhiteThePrint@yahoogroups.com
  Subject: Re: [Digital BW] Digital, film, scanning comparisons


  --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Roger L
  Sopher" <rlsopher@c...> wrote:
  > C'mon guys, give it a rest...

  Why?  It's not off-topic.

  This IS relevant to printing in general and black and white printing
  in particular.   The relationship between the reflectance spectra of
  pigments and colors we perceive is important to problems like
  metamerism, profile-matching of monitors and printers, and how to
  use CMYK (and its variants) to produce even grayscales.

  Anthony raises some interesting questions.   Assuming that a Y
  pigment in an inkjet printer produces an actual yellow (say, 560 nm)
  what does it mean to "match" this to a yellow on a monitor which is
  really comprised of a red and a green?   If you go to my web page
  where I have enlarged  "grayscale" images showing the actual dots of
  ink laid down by a 2200 you can see all the different ways
  the "same" color is created.   How might these methods prove
  DIFFERENTLY sensitive to ambient lighting and how might that inform
  our choice of drivers or inks?

  My undergraduate major was the neurophysiology of visual
  processing.  (imaging sticking single-cell recording devices in
  cat's brains).   So this is a very interesting topic, as well.

  Shortly after I joined the group some months back I complained that
  one difference between darkroom printers I knew and digital ones is
  that many darkroom printers I knew were genuinely interested in the
  chemistry and physics of their craft, whereas too many digital
  printers "black-boxed" everything.   When I have questions about the
  chemistry of developing and printing or the physical properties of
  emulsions and papers I can always find some old darkroom geek who
  knows the answer.   But when I have similar questions about the
  physical and chemical properties of inks and inkjet papers it's hard
  to find people here who know or even care.   Your complaint
  reinforces that stereotype.



        Yahoo! Groups Sponsor

                Select
MakeAcuraAudiBMWCadillacChevroletChryslerDodgeFordGMCHondaInfinitiJaguarJeep

        Please visit the Group Homepage to check the Files, Bookmarks, Polls
and other resources as they are often being updated. The page is at:

        http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint

        If you wish to receive no emails or just a daily digest, or you wish
to unsubscribe, please edit your Membership preferences by visiting this
same page.

        Please follow these basic guidelines:
        - Include your full name with your message.
        - Include the address of your website, if you have one.
        - As threads develop, trim off excess portions of earlier messages
to keep them short.
        - As the topic of a thread changes remember to change the subject
header.
        - Good manners are required at all time. No personal attacks or
flames
        - Complete your Yahoo profile.
        - Before posting a question, search the message archives and the
various resources on the homepage.




        Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.



[Non-text portions of this message have been removed]

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "donbga" 

> Who cares what the physical properties of these materials
> are from a practical point of view.

I had a practical need for that information a few months ago here, 
when I wanted to know more about the chemical and physical 
properties of inkjet paper and inks so I could formulate an inkjet 
receptive surface on true artist's canvas. (you can't paint 
on "inkjet canvas 'paper'")

My comment was about the cultural difference between the two 
groups.  I was struck by a lack of deep-down technical curiosity 
about the underlying chemical and physical processes involved in 
inkjet printing.   Granted the majority  of darkroom printers don't 
have it eaither, but a significant minority do.  Any active darkroom 
forum will have a few people who can asnwer questions about 
developer chemistry, for instance, but inkjet forums seldom have 
anyone who knows anything about inkjet chemistry.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony 
Atkielski" <anthony@a...> wrote:
> Peter writes:
> 
> > But that's just as artificial as the single-wavelength
> > LED example I used.
> 
> It is a simplified illustration.
> 
> > You have not demonstrated this has any bearing
> > on the real world problem of mapping color film
> > to black and white.
> 
> I've tried to explain this in as many ways as I can. 

All I asked for was a real-world empirical example, not a 
gedankenexperiment.  Why can't you name a specific wavelength that 
is picked up with a certain density on some black and white film, 
but "falls in between" two colors on some color film in such a way 
that if converted to b+w it would look different?

The key to this answer and most of this discussion is the filters. Any 
filter for electromagnetic energy - microwave, optical or Xray have 
characteristics. These filter characteristics come in two flavors, 
amplitude response and phase response. A black and white sensor captures 
only the amplitue, i.e., equivalent to the energy energy, of the wave 
that hits it at a specific frequency. The response is not flat - as we 
all know nor should it necessarly be or else we would not use filters 
when shooting B&W film. When you shoot in color you have three amplitude 
responses to contend with - one for each filter measures the amplitude 
at each frequency. It's the area in the fall off of these filters where 
the problem arises.

While you would have to know quite a few specifics about the color 
filters used to answer such a question, in general the calibration to 
ture a color negative into a "TriX" image would be quite an undertaking.

The issues of dynamic range also comes to play in the shadow and 
highlight areas. As long as you are in the linear range of the sensor ( 
or film ) you can do all the manipulation you want. When you get into 
the nonlinear range, e.g., the shoulder and toe, all bets are off and 
calibration of color to black and white that maintains the fidelidity 
you get from a black and white sensor is most likely not possible.

Truman

Peter Nelson wrote:

>All I asked for was a real-world empirical example, not a 
>gedankenexperiment.  Why can't you name a specific wavelength that 
>is picked up with a certain density on some black and white film, 
>but "falls in between" two colors on some color film in such a way 
>that if converted to b+w it would look different?
>
>
>  
>

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Austin 


> Once again, RGB values from a CCD imaging device
> encompass the entire spectral range that Tri-X encompasses,
> and therefore you DO have the original complete spectrum
> to work from, with the "original" spectrum being
> defined as the spectrum that Tri-X can record.  Good
> grief.

He MAY have a point but he hasn't provided any empirical evidence of 
it.  He's also doing a lousy job explaining his argument so I'll try 
to paraphrase.
   
Say you have a CCD with RGB sensors whose PEAK response is 610nm, 
500 nm and 430 nm.  They all overlap but those are their peaks.  Say 
you have two "yellow" objects - X comprised of a mix of red (610 nm) 
and green (500 nm) and Y - true yellow - 560 nm. 

X will produce a strong signal in the R and G channels, so when 
converted to B+W it will be "bright".   Y will still produce a 
signal in the R and G channels but it will be down the skirts of 
both filters so that signal will be weaker.  Thus when converted to 
grayscale the two yellows will have different intensities. 

But since b+w film has more-or-less uniform spectral response the 
two yellows would look the same on Tri-X.  

Did I explain that right, Anthony? 

Here's the flaw, as I see it.  For X to have the same intensity in 
the original scene as pure yellow the R and G components would each 
have to be half as bright as the pure yellow.  Think of it this way -
 suppose you have a 10mW yellow LED.   What wattage would an R and  
G LED have to be to produce the same brightness of yellow?  5mW 
each.  (pretending linear response, equal efficiency, etc)  So in 
the X and Y example, above the R and G would each have to be dimmer 
to start with, so it all evens out.

So I agree with you - I think Anthony has to come up with real-world 
evidence that what he's saying is true.

> All I asked for was a real-world empirical example, not a
> gedankenexperiment.  Why can't you name a specific wavelength that
> is picked up with a certain density on some black and white film,
> but "falls in between" two colors on some color film in such a way
> that if converted to b+w it would look different?

Peter,

There isn't one, if you are talking Provia 400 vs Tri-X, as I've shown the
entire spectral response of Tri-X fits within the spectral response of
Provia 400.  So, any frequency that Tri-X responds to, Provia will as well.

Austin

Truman,

> When you shoot in color you have three amplitude
> responses to contend with - one for each filter measures the amplitude
> at each frequency. It's the area in the fall off of these filters where
> the problem arises.

Specifically, what problem?  Yes, the quantum efficiency of the CCD is less
in this range, but the information is still "seen".

> While you would have to know quite a few specifics about the color
> filters used to answer such a question, in general the calibration to
> ture a color negative into a "TriX" image would be quite an undertaking.

And I disagree, but you can prove me wrong by agreeing to pay me to
undertake the algorithm, and if I can't do it, you don't one me a thing ;-)

> The issues of dynamic range also comes to play in the shadow and
> highlight areas. As long as you are in the linear range of the sensor (
> or film ) you can do all the manipulation you want. When you get into
> the nonlinear range, e.g., the shoulder and toe, all bets are off and
> calibration of color to black and white that maintains the fidelidity
> you get from a black and white sensor is most likely not possible.

And again, I disagree.  As long as it's deterministic, it can be reproduced.
What you (or Anthony) will end up fussing about is the accuracy (having to
do with the dynamic range as you point out, and having to do with fidelity)
of the reproduction...but I'm not convinced that this is really an issue,
but it is something to look for, no doubt.  But, I still believe that within
the VISIBLE grayscale space (let's even say 8 bit space, which is larger
than our vision can discern in grayscale anyway), this is simply not an
issue and can be done.

Austin

Peter writes:

> Did I explain that right, Anthony? 

No, you got most of it wrong.

Anthony Atkielski wrote:

> Similarly, all color display and printing systems for
> human beings are designed around this defect of human vision, and so 
> they
> cannot make this distinction, either

It's not a bug.  It's a feature.

M.

In any sensor when one is operating outside the linear dynamic range, 
the sensor is responding nonlinearly to the stimuli. I for example you 
are at the at a signal energy that when broken into three frequency 
components each component will be in the non-linear portion of the 
dynamic range then by recording the energy on the three sensors. This 
will result in different answers for the total energy on the sensor. In 
a photograph it will most likely result in a loss of detail in one color 
channel or the other and result in distortion in the photograph just the 
same way as one gets distortion in an audio system, an C or X band 
imaging radar when one is trying to operate outside the dynamic range of 
the sensor.

Truman

It is important to stay in the linear range.

Austin Franklin wrote:

>Truman,
>  
>
>
>  
>
>>The issues of dynamic range also comes to play in the shadow and
>>highlight areas. As long as you are in the linear range of the sensor (
>>or film ) you can do all the manipulation you want. When you get into
>>the nonlinear range, e.g., the shoulder and toe, all bets are off and
>>calibration of color to black and white that maintains the fidelidity
>>you get from a black and white sensor is most likely not possible.
>>    
>>
>
>And again, I disagree.  As long as it's deterministic, it can be reproduced.
>What you (or Anthony) will end up fussing about is the accuracy (having to
>do with the dynamic range as you point out, and having to do with fidelity)
>of the reproduction...but I'm not convinced that this is really an issue,
>but it is something to look for, no doubt.  But, I still believe that within
>the VISIBLE grayscale space (let's even say 8 bit space, which is larger
>than our vision can discern in grayscale anyway), this is simply not an
>issue and can be done.
>
>Austin
>  
>



[Non-text portions of this message have been removed]

donbga wrote (snip):
"Who cares what the physical properties of these materials
are from a practical point of view".

and  Peter Nelson replied  (snip) :
"I (am)struck by a lack of deep-down technical curiosity 
about the underlying chemical and physical processes involved
in inkjet printing (among inkjet printers)".

In simpler times, some (like DaVinci) wore both the 'major artist' 
and the 'major scientist' coats, but it hasn't been done in modern 
times (since Goethe). Artists today make images, not scientific 
explanations; (the converse limitation is true for scientists). 
An image-maker's goal is to produce the deepest and richest image 
he/she can, and to do this he/she uses whatever means are at hand 
and deemed suitable. As Don (dongba) says, "from a practical point 
of view, who cares".

Peter, your analysis/contention here seems superficial. Show me 
a major photographer who understands his/her physical materials 
in the same way a chemist or physicist does -- you can't. 
A number of the best (chemical process) printers (e.g. P. Strand,
P. Caponigro, A. Stieglitz), who had/have a deep intuitive grasp 
of how to use their materials, have made clear that they 
do not employ a truly scientific understanding of the stuff they 
work with -- and source material on this point is available, and you 
could check it out. Further, painters and sculptors are just like photographers: they use what's at hand and what they consider 
suitable, just like Don says; they care about how their image works,
and they don't rely on formal scientific understanding to do this.

Curiosity takes more shapes than you are allowing for. You need 
to escape from your personal presumptions and look at what the major 
practioners actually do.  

And from my standpoint, I'll join the others who've said "Enough 
already" to the extended discussion of the physics of 'digital, 
film, and scanning'. It's not really germane to making deep and 
rich printed images.

John           

----- Original Message ----- 
  From: Peter Nelson 
  To: DigitalBlackandWhiteThePrint@yahoogroups.com 
  Sent: Thursday, May 22, 2003 6:19 PM
  Subject: Re: [Digital BW] Digital, film, scanning comparisons


  --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "donbga" 

  > Who cares what the physical properties of these materials
  > are from a practical point of view.

  I had a practical need for that information a few months ago here, 
  when I wanted to know more about the chemical and physical 
  properties of inkjet paper and inks so I could formulate an inkjet 
  receptive surface on true artist's canvas. (you can't paint 
  on "inkjet canvas 'paper'")

  My comment was about the cultural difference between the two 
  groups.  I was struck by a lack of deep-down technical curiosity 
  about the underlying chemical and physical processes involved in 
  inkjet printing.   Granted the majority  of darkroom printers don't 
  have it eaither, but a significant minority do.  Any active darkroom 
  forum will have a few people who can asnwer questions about 
  developer chemistry, for instance, but inkjet forums seldom have 
  anyone who knows anything about inkjet chemistry.




        Yahoo! Groups Sponsor 
       
       

  Please visit the Group Homepage to check the Files, Bookmarks, Polls and other resources as they are often being updated. The page is at:

  http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint

  If you wish to receive no emails or just a daily digest, or you wish to unsubscribe, please edit your Membership preferences by visiting this same page.

  Please follow these basic guidelines:
  - Include your full name with your message.
  - Include the address of your website, if you have one.
  - As threads develop, trim off excess portions of earlier messages to keep them short.
  - As the topic of a thread changes remember to change the subject header.
  - Good manners are required at all time. No personal attacks or flames
  - Complete your Yahoo profile.
  - Before posting a question, search the message archives and the various resources on the homepage. 




  Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service. 



[Non-text portions of this message have been removed]

Truman,

> In any sensor when one is operating outside the linear dynamic range,
> the sensor is responding nonlinearly to the stimuli.

I don't see how linearity is an issue, as long as it's repeatable and can be
characterized.

> I for example you
> are at the at a signal energy that when broken into three frequency
> components each component will be in the non-linear portion of the
> dynamic range then by recording the energy on the three sensors. This
> will result in different answers for the total energy on the sensor.

But isn't the answer the same given a consistent input?

> In
> a photograph it will most likely result in a loss of detail in one color
> channel or the other and result in distortion in the photograph just the
> same way as one gets distortion in an audio system, an C or X band
> imaging radar when one is trying to operate outside the dynamic range of
> the sensor.

But we're not talking about working outside the dynamic range of the sensor.

> It is important to stay in the linear range.

I disagree.  I deal with non-linear signal processing all the time.  I don't
see this as an issue.

Regards,

Austin

Anthony,

Aside from the fact that this thread has degenerated into a "I'm right,
you're wrong." situation and all parties should give it and the list a rest,
you are essentially saying that 3 X 4 = 12 is not the same as 2 X 6 = 12 nor
the same as 4 X 3 = 12. While they are not the same, 12 = 12 = 12 or the end
result is the same regardless of how we got there. In the end the final
print is everything and the path chosen is immaterial. Don't worry about the
technical details at this level. Have some of your B&W negs scanned and make
some digital prints. Practice is much more important than theory.

Martin Wesley

http://www.borderless-photos.de/guests.html



----- Original Message ----- 
From: "Anthony Atkielski" <anthony@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, May 22, 2003 1:49 PM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> Austin writes:
>
> > ... it is therefore quite deterministic to
> > produce a mapping from the information on the
> > color film, as produced by it's spectral response,
> > to the same spectral response that B&W film does.
>
> Absolutely ... but this presumes that you have the original complete
> spectrum to work from, not just RGB values.  Once the color film has
> converted this broad spectral response to three RGB values, a conversion
to
> B&W based on this correlation between curves is no longer possible.
>
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

----- Original Message ----- 

From: "Peter Nelson" <peter@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, May 22, 2003 3:19 PM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "donbga"
>
> > Who cares what the physical properties of these materials
> > are from a practical point of view.
>
> I had a practical need for that information a few months ago here,
> when I wanted to know more about the chemical and physical
> properties of inkjet paper and inks so I could formulate an inkjet
> receptive surface on true artist's canvas. (you can't paint
> on "inkjet canvas 'paper'")
>
> My comment was about the cultural difference between the two
> groups.  I was struck by a lack of deep-down technical curiosity
> about the underlying chemical and physical processes involved in
> inkjet printing.   Granted the majority  of darkroom printers don't
> have it eaither, but a significant minority do.  Any active darkroom
> forum will have a few people who can asnwer questions about
> developer chemistry, for instance, but inkjet forums seldom have
> anyone who knows anything about inkjet chemistry.

Peter,

I think that in to a large extent the type of info you are looking for is
being closely guarded by the manufacturers which makes it tough to discuss
the differences an effects. It would be like asking Ilford what their
gelatin formula is for their papers. They just aren't going to publish that
info.

I would love to know more about ink receptor coatings, sizings and
brighteners but I don't think we will get it.

Martin Wesley

Hello Don,

>Who cares what the physical properties of these materials are from a 
>practical point of view...But why not worry about making photographs 
>instead of all this ancillary stuff?

I do.  I think this is an important subject and have been following
the thread with keen interest.  I am especially interested because it
may ultimately affect my decision on whether to spend several thousand
dollars on digi equipment.  I also agree with Peter that it is on
topic and that we should know more about our craft. 

One of the problems in the debate aspect of this thread, as I see it,
is that some of the participants have so far refused to acknowledge
some of Anthony's important points.  Anthony never said we can't get
good BW prints from color capture.  Of course we can - it's been done,
and as Austin points out, there is plenty enough information there to
do it and even to emulate other films, spectral response and all that.
  Anthony has never once disputed that.

What he IS saying is that we cannot exactly duplicate what the BW film
can do.  There is a difference between emulate and duplicate.  Whether
the difference is discernable to the photographer is another matter. 
Both I and another photographer have reported that we have so far
failed in our efforts to get a color->bw result that satisfies us, and
I think that fact is an important part of the topic.  In fact, it's
really what the topic is about.  Anthony's information is the first
plausible explanation for a question I've asked on and off for months
and gotten almost zero response to.

It has never really been about Tri-X, just that it has been the film
used in the discussion as an example, partly because of my input and
because it is such a distinctive and widely used film.  I have used
Tri-X for so many years that all my intuitive artistic sensibilities
are attuned to it, to the point that I have, so far, not been able to
emulate it to my satisaction.  I continue to hope that the technology
and my own skills and knowledge will someday give me satisfactory
results, as I'd love to get away from dependence on film.

I appreciate very much what Anthony has brought to this and I'm sorry
he has become so frustrated at the obtuseness of some of the other
participants.  His explanations have been crystal clear.

I do agree, however, that the usefulness of this thread has run its
course and we ought to all hang it up and head for the showers.    

Regards,
Clayton


Info on black and white digital printing at    
http://www.cjcom.net/digiprnarts.htm

Truman,

>...in general the calibration to turn a color negative into 
>a "TriX" image would be quite an undertaking.

>When you get into the nonlinear range, e.g., the shoulder and 
>toe, all bets are off and calibration of color to black and 
>white that maintains the fidelidity you get from a black and 
>white sensor is most likely not possible.

...not to mention the part that the film's grain structure plays in
the equation.  Thanks very much for a very sensible reply.


Regards,
Clayton


Info on black and white digital printing at    
http://www.cjcom.net/digiprnarts.htm

Clayton,

> One of the problems in the debate aspect of this thread, as I see it,
> is that some of the participants have so far refused to acknowledge
> some of Anthony's important points.

You assume they are important.  I do not believe they are.  I would
acknowledge them if I thought they were.  They have also not been shown to
have any importance/substance, which is what I take issue with.

> What he IS saying is that we cannot exactly duplicate what the BW film
> can do.

What does "exactly" mean?  This is the problem I have with Anthony.  He uses
words that are amorphous, and when it is taken to task for his statements,
he believes he can then move the meaning around (to quote "It was a
simplified illustration" etc.).  I, for one, get tired of it.  Of course,
you can't "exactly" duplicate Tri-X with anything but Tri-X, that isn't in
dispute...but you CAN to a very very very high degree, replicate it.

> Whether
> the difference is discernable to the photographer is another matter.

No, that should be the ONLY matter.

> Both I and another photographer have reported that we have so far
> failed in our efforts to get a color->bw result that satisfies us, and
> I think that fact is an important part of the topic.  In fact, it's
> really what the topic is about.

Agreed.  I don't know what you have done, so I certainly can't comment on
it.

> Anthony's information is the first
> plausible explanation for a question I've asked on and off for months
> and gotten almost zero response to.

Er, what information?  He hasn't said anything of any substance that I've
seen.  Specifically, what is it you believe has some substance that he's
said?

> It has never really been about Tri-X, just that it has been the film
> used in the discussion as an example, partly because of my input and
> because it is such a distinctive and widely used film.

I'm fine with that, but it certainly helps to have a baseline to discuss, as
in my comparison of the Provia vs Tri-X spectral sensitivity.

> I have used
> Tri-X for so many years that all my intuitive artistic sensibilities
> are attuned to it...

Same here...it is my favorite film.

> ...to the point that I have, so far, not been able to
> emulate it to my satisaction.

Yes, but that's the problem.  Emulate what?  To what development?  To what
exposure?  What YOU do with Tri-X, I don't know, and it very may be
substantially different than what I do with it.  I only know what I do with
it.  It's a moving target.  Given that, Anthony can go on adnauseum saying
"see, I said you can't emulate it".

> I continue to hope that the technology
> and my own skills and knowledge will someday give me satisfactory
> results, as I'd love to get away from dependence on film.

Whoa, who said anything about not using film?  I am specifically discussing
scanning of color film vs scanning of Tri-X and being able to convert the
scanned color film into grayscale that can't be distinguished from Tri-X.
That is an entirely different problem than using a digital capture device
and trying to match Tri-X.

> I appreciate very much what Anthony has brought to this and I'm sorry
> he has become so frustrated at the obtuseness of some of the other
> participants.

Interesting.  I see it entirely differently.  It's that I have become
frustrated with HIS obtuseness and inability to understand basic concepts
that I believe he is missing.

> His explanations have been crystal clear.

It's not his explanations I've seen as clear.  So far, he hasn't answered my
question about frequency, and until he does, his "theory" is clearly
unsubstantiated, and that's clear to me.

I'm happy to listen to what you believe is clear, and in fact I'd appreciate
it...because if you do believe there is any substance there, and I've
overlooked it, I'd like to see it.

Austin

> Truman,
>
> >...in general the calibration to turn a color negative into
> >a "TriX" image would be quite an undertaking.
>
> >When you get into the nonlinear range, e.g., the shoulder and
> >toe, all bets are off and calibration of color to black and
> >white that maintains the fidelidity you get from a black and
> >white sensor is most likely not possible.
>
> ...not to mention the part that the film's grain structure plays in
> the equation.  Thanks very much for a very sensible reply.
>
>
> Regards,
> Clayton

Clayton,

You are right about the grain structure as far as emulation goes, but I am
specifically addressing the tonality issue, which I strongly disagree with
Truman's assessment.  I still see no reason why there is any problem at all
in the shoulder and/or toe.  So what if it's non-linear, it's still
deterministic and given that, still mappable.  I do it all the time, so do
you too, if you save tonal curve adjustment settings.

Austin

John,

I am not sure I can agree that we have no scientist/philosopher/artists with
us today. Eliot Porter leaps to mind as a modern scientist/artist. However,
I think you have largely hit the nail on the head regarding this thread.

I personally believe that it is theoretically possible to scan a piece of
color film, positive or negative, and manipulate the scan so that tonally
and texturally it would be taken for a shot made with Tri-X developed in
some specific manner. The science and technology is there if someone wants
to do it. Intuitively my response is why bother. If you like the Tri-X look,
then shoot Tri-X and scan it.

I realize that there is a time value to shooting slides or color film and
having them commercially processed but if you really want a certain look
that can be obtained directly, then my feeling is that you should pursue
that route. If you are not strongly wedded to that look, then by all means
experiment and try different workflows to find what satisfies you.

Clayton's question is pertinent in trying to decide on whether to jump to
digital capture or not. I think the same reasoning applies. If you really
like the look of a certain film and process, it would seem better to stick
with what you know and do best. Look at digital in this case as an expansion
of your tools not necessarily a replacement.

So for me the whole long thread over whether you can or you can't do the
conversion, doesn't seem of great practical importance to the issue of
making a fine "Tri-X" print.

Sometimes it is of value to look at the science or detailed technology and
theory of what is going on to improve our understanding of the process. I am
not sure that is the case here.

Overall the "old pros" need to be tolerant of the new comers' questions and
concerns even if they have heard it a thousand times before and the new
comers' need to be tolerant of the more experienced members' need to discuss
things at length and in technical detail.

Martin Wesley

http://www.borderless-photos.de/guests.html



----- Original Message ----- 
From: "John/Julie Gittins" <jgittins2@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, May 22, 2003 6:38 PM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> donbga wrote (snip):
> "Who cares what the physical properties of these materials
> are from a practical point of view".
>
> and  Peter Nelson replied  (snip) :
> "I (am)struck by a lack of deep-down technical curiosity
> about the underlying chemical and physical processes involved
> in inkjet printing (among inkjet printers)".
>
> In simpler times, some (like DaVinci) wore both the 'major artist'
> and the 'major scientist' coats, but it hasn't been done in modern
> times (since Goethe). Artists today make images, not scientific
> explanations; (the converse limitation is true for scientists).
> An image-maker's goal is to produce the deepest and richest image
> he/she can, and to do this he/she uses whatever means are at hand
> and deemed suitable. As Don (dongba) says, "from a practical point
> of view, who cares".
>
> Peter, your analysis/contention here seems superficial. Show me
> a major photographer who understands his/her physical materials
> in the same way a chemist or physicist does -- you can't.
> A number of the best (chemical process) printers (e.g. P. Strand,
> P. Caponigro, A. Stieglitz), who had/have a deep intuitive grasp
> of how to use their materials, have made clear that they
> do not employ a truly scientific understanding of the stuff they
> work with -- and source material on this point is available, and you
> could check it out. Further, painters and sculptors are just like
photographers: they use what's at hand and what they consider
> suitable, just like Don says; they care about how their image works,
> and they don't rely on formal scientific understanding to do this.
>
> Curiosity takes more shapes than you are allowing for. You need
> to escape from your personal presumptions and look at what the major
> practioners actually do.
>
> And from my standpoint, I'll join the others who've said "Enough
> already" to the extended discussion of the physics of 'digital,
> film, and scanning'. It's not really germane to making deep and
> rich printed images.
>
> John
>
> ----- Original Message ----- 
>   From: Peter Nelson
>   To: DigitalBlackandWhiteThePrint@yahoogroups.com
>   Sent: Thursday, May 22, 2003 6:19 PM
>   Subject: Re: [Digital BW] Digital, film, scanning comparisons
>
>
>   --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "donbga"
>
>   > Who cares what the physical properties of these materials
>   > are from a practical point of view.
>
>   I had a practical need for that information a few months ago here,
>   when I wanted to know more about the chemical and physical
>   properties of inkjet paper and inks so I could formulate an inkjet
>   receptive surface on true artist's canvas. (you can't paint
>   on "inkjet canvas 'paper'")
>
>   My comment was about the cultural difference between the two
>   groups.  I was struck by a lack of deep-down technical curiosity
>   about the underlying chemical and physical processes involved in
>   inkjet printing.   Granted the majority  of darkroom printers don't
>   have it eaither, but a significant minority do.  Any active darkroom
>   forum will have a few people who can asnwer questions about
>   developer chemistry, for instance, but inkjet forums seldom have
>   anyone who knows anything about inkjet chemistry.
>
>
>
>
>         Yahoo! Groups Sponsor
>
>
>
>   Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
>   http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
>   If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
>   Please follow these basic guidelines:
>   - Include your full name with your message.
>   - Include the address of your website, if you have one.
>   - As threads develop, trim off excess portions of earlier messages to
keep them short.
>   - As the topic of a thread changes remember to change the subject
header.
>   - Good manners are required at all time. No personal attacks or flames
>   - Complete your Yahoo profile.
>   - Before posting a question, search the message archives and the various
resources on the homepage.
>
>
>
>
>   Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
>
>
>
> [Non-text portions of this message have been removed]
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Clayton 
Jones" <cj@c...> wrote:

.....
> 
> What he IS saying is that we cannot exactly duplicate what the 
BW film
> can do.  There is a difference between emulate and duplicate.  
Whether
> the difference is discernable to the photographer is another 
matter. 
> Both I and another photographer have reported that we have so 
far
> failed in our efforts to get a color->bw result that satisfies us, 
and
> I think that fact is an important part of the topic.  In fact, it's
> really what the topic is about.  Anthony's information is the first
> plausible explanation for a question I've asked on and off for 
months
> and gotten almost zero response to.
...
> I appreciate very much what Anthony has brought to this and 
I'm sorry
> he has become so frustrated at the obtuseness of some of the 
other
> participants.  His explanations have been crystal clear.
> 
> Regards,
> Clayton
> 

Clayton, 

I think you are right on in your assessment of the
situation.   Anthony's explanations seemed pretty clear as
well but the nagging thought is -- does it make any difference
in the real world photography?  Notch filters are pretty rare
in ordinary photography so I decided to look at the response
curves of real black&white filters.  B+W is kind enough to
show response curves for many of their filters.

The question is: if you take a B&W picture with a filter and
B&W film, can you get the same grayscale response by
taking the picture in color and then channel mixing the RGB.
My first reaction is that you can do pretty well -- like make a
blue sky darker by not using much of the Blue channel.

But on considering what filters I do use -- several yellows,
oranges, reds and greens -- I have a lot of options on
what spectral response I want for a particular image.  Then
looking at the response curves for the various filters it
evident that the frequencies selected have fairly sharp cut-offs.
Some of the filters that I choose between are closer than
50nm apart and quite steep (0-80% in less than 50nm).

There's no way that channel mixing from RGB can come
close to that shape of response curve and selectivity.  So
at this point I'm inclined to think that Anthony not only has a
theoretical point but an issue that can in some cases has
a very real effect on an image.

Roy

Roy writes:

> Anthony's explanations seemed pretty clear as
> well but the nagging thought is -- does it
> make any difference in the real world photography?

That is for each photographer to decide.  Some photographers consider
certain types of prints to be absolutely preferable above all others; other
photographers don't much care about the printing process, as long as it more
or less accurately reflects the original image.

Like others here, I've not been able to get the B&W images I'd really like
to see through any manipulation of RGB color images.  The only way to get
what I want is to shoot black and white film in the first place.  It does
not have to be conventional B&W film--it could be chromogenic or even a
dedicated B&W electronic sensor (if such things were easily to be had), but
it has to be B&W from the start in order to get some of the unique looks of
B&W photography.  This isn't ever going to change, so photographers keenly
interested in the subtleties of B&W are going to have to continue to shoot
dedicated black and white.  Maybe one day they'll shoot B&W digitally, but
even then, they'll have to do it with a camera dedicated to B&W, and not a
color digital camera.

> There's no way that channel mixing from RGB can
> come close to that shape of response curve and
> selectivity.  So at this point I'm inclined to think
> that Anthony not only has a theoretical point but
> an issue that can in some cases has a very real
> effect on an image.

Exactly.  Some people might not care, but others might find it very
troublesome, particularly those who are deeply interested in the very best
of B&W photography.

> There's no way that channel mixing from RGB can come
> close to that shape of response curve and selectivity.  So
> at this point I'm inclined to think that Anthony not only has a
> theoretical point but an issue that can in some cases has
> a very real effect on an image.
>
> Roy

Roy,

YOU may be very well right about what you postulate, but that isn't what
Anthony said.  He said that you can't "duplicate" Tri-X by using color
information and mapping the tonal curves.  He didn't say you can't
"duplicate" Tri-X when used with colored filters...etc.  That's possibly a
different problem.  That changes the response curve for Tri-X, which he
never said was what he was doing.

Now, on to your theory.  Why do you believe that adjusting the tonal
response with a colored filter can't be duplicated?  What is the frequency
range represented by RGB, and what frequency do you believe RGB doesn't
cover?

Austin

Thanks all for this great discussion. As regards scanning neutral density
images (B & W) on a color scanner, would it make a difference if  instead of
changing the sensors, the light source was another color? I presume the best
solution would be a sensor array that was not sensitive to color at all but
just light intensity.

 

Tony

-----Original Message-----
From: Austin Franklin [mailto:darkroom@...] 
Sent: Thursday, May 22, 2003 12:15 AM
To: DigitalBlackandWhiteThePrint@yahoogroups.com
Subject: RE: [Digital BW] Why is ND B&W scan better -- was Digital, film,
scanning comparisons

 

Truman,

> Quantum efficiently  has nothing to do with it.

Well, yes it does.  Quantum efficiency is the ratio  of the number of
photogenerated electrons captured by a sensor to the number of photons
incident upon the sensor during a period of time.  What that basically means
is that red, because of it's %80 quantum efficiency over a VERY broad
spectrum will fill the well much faster than the other two colors, which
means red will smear and bloom first, given the same amount of light energy
of each color.

> Red is at the low end
> or the visible frequency spectrum. Normally the higher the frequency the
> wider the bandwidth of filters at that frequency, unless you use a local
> frequency reference to down convert the signal. Why is this not the case
> for red light which is a lower frequency than blue light? Different
> materials for CCD's have different frequency reposes. I guess I don't
> understand why "red" produces more energy output.

Because the CCD is most sensitive to it, which is what quantum efficiency
is!

> The energy of a photon is Planks constant X it frequency. A photon of
> blue has more energy than a photon of red.

That doesn't matter.  What matters is how the CCD sees it, which is quantum
efficiency...

> Different scanners use
> different light sources. I am still at a loss - not that I am disputing
> it - with your statement. I just don't understand.

Light sources have nothing to do with it.  The light is filtered by either
the RG or B filter, and the CCD reacts to that color light only.  I don't
understand why you think that matters?  All the light sources cover the
spectrum that is being detected.

Austin






Yahoo! Groups Sponsor


 
<http://rd.yahoo.com/M=251812.3170658.4537139.1261774/D=egroupweb/S=17050191
82:HM/A=1564415/R=0/*http:/www.netflix.com/Default?mqso=60164784&partid=3170
658> 


 
<http://us.adserver.yahoo.com/l?M=251812.3170658.4537139.1261774/D=egroupmai
l/S=:HM/A=1564415/rand=695069525> 


Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:

http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint

If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.

Please follow these basic guidelines:
- Include your full name with your message.
- Include the address of your website, if you have one.
- As threads develop, trim off excess portions of earlier messages to keep
them short.
- As the topic of a thread changes remember to change the subject header.
- Good manners are required at all time. No personal attacks or
&amp;amp;quot;flames.&amp;amp;quot;
- Complete your Yahoo profile.
- Before posting a question, search the message archives and the various
resources on the homepage. 




Your use of Yahoo! Groups is subject to the Yahoo!
<http://docs.yahoo.com/info/terms/>  Terms of Service. 



[Non-text portions of this message have been removed]

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Austin 
Franklin" <darkroom@i...> wrote:
> > There's no way that channel mixing from RGB can come
> > close to that shape of response curve and selectivity.  So
> > at this point I'm inclined to think that Anthony not only has a
> > theoretical point but an issue that can in some cases has
> > a very real effect on an image.
> >
> > Roy
> 
> Roy,
> 
> YOU may be very well right about what you postulate, but that 
isn't what
> Anthony said.  He said that you can't "duplicate" Tri-X by using 
color
> information and mapping the tonal curves.  He didn't say you 
can't
> "duplicate" Tri-X when used with colored filters...etc.  That's 
possibly a
> different problem.  That changes the response curve for Tri-X, 
which he
> never said was what he was doing.
> 
> Now, on to your theory.  Why do you believe that adjusting the 
tonal
> response with a colored filter can't be duplicated?  What is the 
frequency
> range represented by RGB, and what frequency do you believe 
RGB doesn't
> cover?
> 
> Austin

Austin,

There isn't any frequency that's not covered -- the issue is the
SHAPE of the frequency response curve.  Using a colored
filter at picture taking time, you can very selectively choose a
response curve based on the film and the colored filter.
However if you take the picture with color film you've reduced
the spectrum information to just 3 values RGB and the
response curves for those 3 values are fixed.  With channel
mixing you only get to pick percentages of those fixed
response curves -- NOT simulate or duplicate the SHAPE
of an arbitrary colored filter response.

The whole concept is very simple.  In the real world visible
light is a continuous spectrum of light frequencies with
arbitrary amplitudes for each frequency.  Conversion to RGB
is reducing an essentially infinite number of possibilities to
just 3 numbers.  The 3 numbers work as well as they do
because they were specifically designed with what the human
eyeball can distinguish.

Roy

Hi Roy,

> There isn't any frequency that's not covered -- the issue is the
> SHAPE of the frequency response curve.  Using a colored
> filter at picture taking time, you can very selectively choose a
> response curve based on the film and the colored filter.

That curve is made up of frequency vs energy, and what is modified are only
frequency and energy, since that is all there is.

> However if you take the picture with color film you've reduced
> the spectrum information to just 3 values RGB and the
> response curves for those 3 values are fixed.

Again, what frequencies are not represented by the 3 RGB values that are
available on Tri-X?  The response curve can in fact be modified (mapped), so
I'm not clear why you believe it's fixed...

> With channel
> mixing you only get to pick percentages of those fixed
> response curves -- NOT simulate or duplicate the SHAPE
> of an arbitrary colored filter response.

I don't care about channel mixer.  You CAN using a tonal curve adjustment
tool, remap the color values to be any response curve you want, within
reasonable bounds of course.

> The whole concept is very simple.  In the real world visible
> light is a continuous spectrum of light frequencies with
> arbitrary amplitudes for each frequency.  Conversion to RGB
> is reducing an essentially infinite number of possibilities to
> just 3 numbers.

I disagree.  You are reducing a frequency and amplitude of a 2D visible area
to three numbers when you convert it to RGB, and to one number when you
convert it to B&W.  The frequency and amplitude are only two numbers, and
correspond to that one point.  Again, what frequency/amplitude that is
visible to Tri-X can not be represented by the three RGB values?  It's a
simple question.

If you can't demonstrate what I ask above, then your belief, and Anthony's,
is simply wrong.

Austin

Hi Tony,

> Thanks all for this great discussion. As regards scanning neutral density
> images (B & W) on a color scanner,

Actually, I was talking about scanning using a ND filter over the CCD.
Typically, film scanners scan B&W film using RGB filters...

> would it make a difference if
> instead of
> changing the sensors, the light source was another color?

I believe the Nikon scanners try varying the light source for the three
different colors (RGB), and I guess if they wanted to, they could just scan
B&W using white light...if that's what you're getting at. I don't believe
they work that way now though...unfortunately.

> I
> presume the best
> solution would be a sensor array that was not sensitive to color
> at all but
> just light intensity.

It's the colored filters that make the CCD sensitive to a particular
spectral range, without a colored filter (or simply using a ND filter), it
is responsive to the entire spectral range it is sensitive to, which is far
wider than the RGB range.

Regards,

Austin

Austin,

 

Thanks, I confess I am somewhat new to all this (digital) and I am trying to
figure out the best path to get the best B & W images using digital means.
Since joining the group last week, I am feeling somewhat overwhelmed at the
learning curve.

 

I have been trying to follow the "digital,film scanning comparisons" thread
as best I can as I get up to speed on all this before I make some decisions
on which way to approach all this. In the meantime, I will read all the
group archive articles and try to make sense of it all.

 

About 10 yrs ago I was doing a lot of MF & LF B&W work and would like to
explore the best ways and means to getting as close to what I am used to in
silver printing via digital. I slipped away from photography to pursue
"multimedia" and am now turning back to it.

 

Any starting points you (or anyone else) can recommend as I begin this
journey would be welcome. The extent of my digital involvement thus far has
been with a Nikon 950 and an Epson 1270 printer. But I have a lot of MF&LF
negs I'd like to work with going forward.

 

So any help is most welcome,

 

Regards,

Tony

 

 

www.lightworx.ca

-----Original Message-----
From: Austin Franklin [mailto:darkroom@...] 
Sent: Friday, May 23, 2003 1:03 PM
To: DigitalBlackandWhiteThePrint@yahoogroups.com
Subject: RE: [Digital BW] Why is ND B&W scan better -- was Digital, film,
scanning comparisons

 

Hi Tony,

> Thanks all for this great discussion. As regards scanning neutral density
> images (B & W) on a color scanner,

Actually, I was talking about scanning using a ND filter over the CCD.
Typically, film scanners scan B&W film using RGB filters...

> would it make a difference if
> instead of
> changing the sensors, the light source was another color?

I believe the Nikon scanners try varying the light source for the three
different colors (RGB), and I guess if they wanted to, they could just scan
B&W using white light...if that's what you're getting at. I don't believe
they work that way now though...unfortunately.

> I
> presume the best
> solution would be a sensor array that was not sensitive to color
> at all but
> just light intensity.

It's the colored filters that make the CCD sensitive to a particular
spectral range, without a colored filter (or simply using a ND filter), it
is responsive to the entire spectral range it is sensitive to, which is far
wider than the RGB range.

Regards,

Austin






Yahoo! Groups Sponsor


 
<http://rd.yahoo.com/M=251812.3170658.4537139.1261774/D=egroupweb/S=17050191
82:HM/A=1564415/R=0/*http:/www.netflix.com/Default?mqso=60164784&partid=3170
658> 


 
<http://us.adserver.yahoo.com/l?M=251812.3170658.4537139.1261774/D=egroupmai
l/S=:HM/A=1564415/rand=933448972> 


Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:

http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint

If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.

Please follow these basic guidelines:
- Include your full name with your message.
- Include the address of your website, if you have one.
- As threads develop, trim off excess portions of earlier messages to keep
them short.
- As the topic of a thread changes remember to change the subject header.
- Good manners are required at all time. No personal attacks or flames
- Complete your Yahoo profile.
- Before posting a question, search the message archives and the various
resources on the homepage. 




Your use of Yahoo! Groups is subject to the Yahoo!
<http://docs.yahoo.com/info/terms/>  Terms of Service. 



[Non-text portions of this message have been removed]

Hello Roy,

>I think you are right on in your assessment of the
>situation.   Anthony's explanations seemed pretty clear as
>well but the nagging thought is -- does it make any difference
>in the real world photography?  

I think only to the degree that this subject matters to the 
photographer.  Even with my 4mpx P&S I've taken some good images and
turned them into small BW prints that look pretty nice.  But no matter
what I've tried (including that neat BW conversion plugin), they still
fall short of what I'd like to see.  For several years (during the
80's) I tried most of the new films that came out and always came back
to good old Tri-X.  I just love the look of it.  So for someone who is
so attuned to a particular film the subject may be more important. 
Digital is growing, and the possibility of leaving film behind grows
more tempting every year.  Upon what will this decision be based?

For months I have simultaneously been tempted by the new cameras
(especially the new 10d) and nagged by doubts about whether I could be
happy using them.  I've brought up the question 2 or 3 times over the
past year, with nobody really weighing in on it.  I was delighted to
read Anthony's ideas because they finally gave a plausible explantion
for what I've been experiencing.

>The question is: if you take a B&W picture with a filter and
>B&W film, can you get the same grayscale response by
>taking the picture in color and then channel mixing the RGB.
>My first reaction is that you can do pretty well 

Yes, amazingly well.


>There's no way that channel mixing from RGB can come
>close to that shape of response curve and selectivity.  So
>at this point I'm inclined to think that Anthony not only has a
>theoretical point but an issue that can in some cases has
>a very real effect on an image.

I agree, and it may not matter to some.  But just having the idea to
chew on at least affords some insight into the process.  As soon as I
read "Anthony's Theorem" my "Ah,Ha!" lights flashed and suddenly my
long time question was clarified.  I still have to make my own
decision, but now at least I have a handle on it.   To me that's the
most important thing from this thread: I can see the question more
clearly.  All the theoretical debate completely missed the point.   

Thanks, Roy.

Regards,
Clayton


Info on black and white digital printing at    
http://www.cjcom.net/digiprnarts.htm

Hello Tony,

>trying to figure out the best path to get the best B & W 
>images using digital means.

>Any starting points you (or anyone else) can recommend as I begin 
>this journey would be welcome. The extent of my digital involvement 
>thus far has been with a Nikon 950 and an Epson 1270 printer. But I 
>have a lot of MF&LF negs I'd like to work with going forward.

Choosing a particular technical path is an expensive decision and one
not easily reversed.  You already have a 1270 and a body of work in
your negs.  I know from experience that you will learn an enormous
amount by spending several months scanning and printing those.  That
way you'll make a much more informed final decision.  There are many
pathways and each have their advocates.  It's hard to make a choice
based only on what others say.

I'm getting remarkably good scans of my 6x7 and 4x5 negs on an Epson
1640SU 1600dpi flatbed with tranny adapter (35mm is not acceptable). 
This level of scanner can be had for a relatively modest amount and
will get you started.  If you decide to continue shooting film you can
invest in a film scanner, or if you decide to go to digital capture
then you won't have overspent on the scanner.  

Regards,
Clayton


Info on black and white digital printing at    
http://www.cjcom.net/digiprnarts.htm

Peter,

Well said: I have found this discussion very interesting. Far more
interesting than the repetitive posts on whether paper x is better or worse
than paper y with inks a, b, or c in person z's opinion.

Bob Frost.

----- Original Message ----- 

From: "Peter Nelson" <peter@...>


>
> Shortly after I joined the group some months back I complained that
> one difference between darkroom printers I knew and digital ones is
> that many darkroom printers I knew were genuinely interested in the
> chemistry and physics of their craft, whereas too many digital
> printers "black-boxed" everything.   When I have questions about the
> chemistry of developing and printing or the physical properties of
> emulsions and papers I can always find some old darkroom geek who
> knows the answer.   But when I have similar questions about the
> physical and chemical properties of inks and inkjet papers it's hard
> to find people here who know or even care.   Your complaint
> reinforces that stereotype.

Clayton,

 

Many thanks for your suggestion, as you say a lot of $$$ can be spent
choosing a particular path and your recommendation on the flatbed really
sits well with me as I move forward, and is something I had frankly not
considered, thinking that a MF film scanner was what I needed but at this
point I cannot justify the price tag. I also realize I also need to get out
there and see first hand what the technology is capable of.

 

My standards for quality are quite high, as I assume everyone else's here is
as well, but it is also a subjective call unless we both happen to be
looking at the same print at the same time and can agree on what a well
executed image is.

 

I live near Toronto, so being able to see good examples of well done inkjet
images should not be too difficult.

You have given me another perspective on all this and it is most welcome.

 

Regards,

Tony

-----Original Message-----
From: Clayton Jones [mailto:cj@...] 
Sent: Friday, May 23, 2003 5:03 PM
To: DigitalBlackandWhiteThePrint@yahoogroups.com
Subject: Re: [Digital BW] Why is ND B&W scan better -- was Digital, film,
scanning compar

 

Hello Tony,

>trying to figure out the best path to get the best B & W 
>images using digital means.

>Any starting points you (or anyone else) can recommend as I begin 
>this journey would be welcome. The extent of my digital involvement 
>thus far has been with a Nikon 950 and an Epson 1270 printer. But I 
>have a lot of MF&LF negs I'd like to work with going forward.

Choosing a particular technical path is an expensive decision and one
not easily reversed.  You already have a 1270 and a body of work in
your negs.  I know from experience that you will learn an enormous
amount by spending several months scanning and printing those.  That
way you'll make a much more informed final decision.  There are many
pathways and each have their advocates.  It's hard to make a choice
based only on what others say.

I'm getting remarkably good scans of my 6x7 and 4x5 negs on an Epson
1640SU 1600dpi flatbed with tranny adapter (35mm is not acceptable). 
This level of scanner can be had for a relatively modest amount and
will get you started.  If you decide to continue shooting film you can
invest in a film scanner, or if you decide to go to digital capture
then you won't have overspent on the scanner.  

Regards,
Clayton


Info on black and white digital printing at    
http://www.cjcom.net/digiprnarts.htm






Yahoo! Groups Sponsor




 
<http://rd.yahoo.com/M=247865.3269369.4566997.1261774/D=egroupweb/S=17050191
82:HM/A=1482387/R=0/*http:/ads.x10.com/?bHlhaG9vaG0xLmRhd=1053724679%3eM=247
865.3269369.4566997.1261774/D=egroupweb/S=1705019182:HM/A=1482387/R=1=105372
4679%3eM=247865.3269369.4566997.1261774/D=egroupweb/S=1705019182:HM/A=148238
7/R=2> 


 
<http://us.adserver.yahoo.com/l?M=247865.3269369.4566997.1261774/D=egroupmai
l/S=:HM/A=1482387/rand=541975552> 


Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:

http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint

If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.

Please follow these basic guidelines:
- Include your full name with your message.
- Include the address of your website, if you have one.
- As threads develop, trim off excess portions of earlier messages to keep
them short.
- As the topic of a thread changes remember to change the subject header.
- Good manners are required at all time. No personal attacks or flames
- Complete your Yahoo profile.
- Before posting a question, search the message archives and the various
resources on the homepage. 




Your use of Yahoo! Groups is subject to the Yahoo!
<http://docs.yahoo.com/info/terms/>  Terms of Service. 



[Non-text portions of this message have been removed]

Martin,

I tried and tried to get some sort of acceptable B&W prints from D30 image
files that I felt would benefit from a grayscale presentation. I used
Channel Mixer, I used the B&W conversion tools available via Pixel Genius's
Photokit, I tried the B&W Pro conversion tool from Imaging Factory, I tried
Fred Miranda's conversion actions, etc. Not one sang for me. Not one even
compared to my digital quads from 4x5/8x10 neg scans; let alone compared to
finely crafted silver print. I finally concluded that either it just can't
be done or I don't currently possess the skill level to make it happen.
Perhaps you're right that I 30 megapixel direct capture could, at least,
come close to film.

BTW, I'm sorry I started the whole film vs. digital thread...that was
certainly not my intention. I shoot film and digital and enjoy the print
results from both. Just for interest sake, I did experiment with
downsampling my 35mm Tango drum scan and generated several prints from the
film scan and digital. The film scan result was nice, but, for the subject
at hand, I thought the digital print was more pleasing to look at. Was the
digital print "better?" Who cares...I liked what I saw. This was color
though..........

Sorry, to bring this back on topic...I agree with other comments that I've
seen posted here...I don't believe we should think of digitally produced B&W
to be better than silver prints created via traditional means, or
vice-versa. They are different "animals" and have their own inherent beauty.
I have labored in the wet darkroom for 25+ years producing nothing but B&W;
mostly from LF negs. For the past few years, I've been working through the
computer producing quadtones via MIS VM inks and Paul's curves. Don't have a
wet darkroom set up. But...I miss it! Sure, I have much more control over
the process working in PS. I'd like to think that I've produced a couple of
nice prints...at least one that I can think of just could not have been
produced in the wet darkroom. I have A4 size images hanging on my living
room wall that I also have the identical image (approx 11x14) in silver. The
quad prints have that almost platinum look to them where the image appears
to be part of the paper...the near whites/grays/blacks in the selenium-toned
silver print simply sing. Both presentations are very nice, but totally
different.

Personally...to borrow a few words from the great Edward Weston...I don't
care if an image in printed on a bath mat as long as it's a good print and,
more importantly, has something to say. I see a lot of technically great
photographs that contain absolutely NO heart! But, that's a subject probably
best left alone.

Thanks for the camaraderie,
Alan Huntley

----- Original Message -----
From: "Martin Wesley" <mwesley250@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Tuesday, May 20, 2003 9:01 PM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> ----- Original Message -----
> From: "A. Huntley" <leicam6@...>
> To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Sent: Tuesday, May 20, 2003 6:15 PM
> Subject: Re: [Digital BW] Digital, film, scanning comparisons
>
>
> > Martin,
> >
> > Very interesting and informative site. Thank you.
> >
> > Alan Huntley
>
> Alan,
>
> Glad someone found it of use. Unfortunately he (nor do any of the endless
> threads here) does not address the basic issue of what happens when you
try
> to print in B&W. All of the files and examples he shows are presumably 3
> channel RGB. I am unclined to agree with his 10 megapixal threshold being
> the point where direct digital cature equals or exceeds 35mm color film
but

> what about B&W? Since the conversion to B&W drops 2/3 of the data the
> implication is that to match 35mm B&W you would need 30 megapixal direct
> capture.
>
> Martin Wesley

Alan,

> I tried and tried to get some sort of acceptable B&W prints from D30 image
> files that I felt would benefit from a grayscale presentation. I used
> Channel Mixer, I used the B&W conversion tools available via
> Pixel Genius's
> Photokit, I tried the B&W Pro conversion tool from Imaging
> Factory, I tried
> Fred Miranda's conversion actions, etc. Not one sang for me.

Well, Jerry (God bless his soul!) pretty much used the D-30 and D-60
exclusively as far as I know, and his images were stupendous...AND...they
were B&W:

http://www.westernechoes.com/

So, I'm not quite sure why you weren't getting prints that "sing", but the
ones I saw from Jerry very much so did.

Regards,

Austin

----- Original Message ----- 

From: "A. Huntley" <leicam6@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Saturday, May 24, 2003 1:45 PM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> Martin,
>
> I tried and tried to get some sort of acceptable B&W prints from D30 image
> files that I felt would benefit from a grayscale presentation. I used
> Channel Mixer, I used the B&W conversion tools available via Pixel
Genius's
> Photokit, I tried the B&W Pro conversion tool from Imaging Factory, I
tried
> Fred Miranda's conversion actions, etc. Not one sang for me. Not one even
> compared to my digital quads from 4x5/8x10 neg scans; let alone compared
to
> finely crafted silver print.

Alan,

I have never had a chance to play around with a better quality digital for
B&W. The wife has a Canon G2 (3.1 mega pixel) for snap shots and my attempts
to get a nice print were somewhat successful if I printed very small, 3x4. I
have a handful of 13X18 B&W prints from other people done with SLR type
cameras that are excellent. So I have to say it can be done.

That said, the feel is very different from prints made from scans of 8x10
and 4x5 negs. I think it is simply a matter of how much fine detail is
recorded and can be gotten onto a piece of paper. Prints from 35mm are much
different from prints from LF too but I like work from both formats.

> I finally concluded that either it just can't
> be done or I don't currently possess the skill level to make it happen.
> Perhaps you're right that I 30 megapixel direct capture could, at least,
> come close to film.

As Austin said, Jerry Olson did some incredible B&W stuff using montages of
digital images that were a blend of scanned film and digital camera files.
Excellent print quality. I do know he was looking to move up to the D60 and
was on a quest for as much image data as possible.
>
> BTW, I'm sorry I started the whole film vs. digital thread...that was
> certainly not my intention.

I don't feel that it is off topic at all and is very pertinent as long as we
are discussing it in terms of how it relates to making a final B&W print.
What happened, (and not for the first time either!) was that it drifted off
into comparing film to digital with little of no mention of output and color
at that, and degenerated into the old "Yes it is. No it isn't" refrain!

> I shoot film and digital and enjoy the print
> results from both. Just for interest sake, I did experiment with
> downsampling my 35mm Tango drum scan and generated several prints from the
> film scan and digital. The film scan result was nice, but, for the subject
> at hand, I thought the digital print was more pleasing to look at. Was the
> digital print "better?" Who cares...I liked what I saw. This was color
> though..........
>
> Sorry, to bring this back on topic...I agree with other comments that I've
> seen posted here...I don't believe we should think of digitally produced
B&W
> to be better than silver prints created via traditional means, or
> vice-versa. They are different "animals" and have their own inherent
beauty.
> I have labored in the wet darkroom for 25+ years producing nothing but
B&W;
> mostly from LF negs. For the past few years, I've been working through the
> computer producing quadtones via MIS VM inks and Paul's curves. Don't have
a
> wet darkroom set up. But...I miss it! Sure, I have much more control over
> the process working in PS. I'd like to think that I've produced a couple
of
> nice prints...at least one that I can think of just could not have been
> produced in the wet darkroom. I have A4 size images hanging on my living
> room wall that I also have the identical image (approx 11x14) in silver.
The
> quad prints have that almost platinum look to them where the image appears
> to be part of the paper...the near whites/grays/blacks in the
selenium-toned
> silver print simply sing. Both presentations are very nice, but totally
> different.

I agree with you. I have done the same thing comparing silver and inkjet
prints from the same negative. They are just different. Both renditions have
image specific strengths and weaknesses to my eye.

I do think that the last 2 or 3 years struggling with Photoshop and digital
B&W printing has made me a better printer. In the process of learning all
this I have become much more sensitive to subtle tonal refinements and
simply look much more closely at prints than I ever did before.
>
> Personally...to borrow a few words from the great Edward Weston...I don't
> care if an image in printed on a bath mat as long as it's a good print
and,
> more importantly, has something to say. I see a lot of technically great
> photographs that contain absolutely NO heart! But, that's a subject
probably
> best left alone.

Oh we have run that thread several times and that can be profitable if
inflammatory! The list should be about art too and not just the technical
issues.

Martin Wesley

http://www.borderless-photos.de/guests.html

Hi Austin,

Agreed. I know that Jerry used the D30/D60 for B&W work and, yes, his images
were/are fantastic! All I said (or meant) was that I personally have not
been successful at converting/printing one of my D30 files that produces a
B&W print that I like. Therefore, I've stuck with good 'ole Tri-X (mostly
LF) and scanning.

Alan Huntley

----- Original Message -----
From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Saturday, May 24, 2003 3:20 PM
Subject: RE: [Digital BW] Digital, film, scanning comparisons


> Alan,
>
> > I tried and tried to get some sort of acceptable B&W prints from D30
image
> > files that I felt would benefit from a grayscale presentation. I used
> > Channel Mixer, I used the B&W conversion tools available via
> > Pixel Genius's
> > Photokit, I tried the B&W Pro conversion tool from Imaging
> > Factory, I tried
> > Fred Miranda's conversion actions, etc. Not one sang for me.
>
> Well, Jerry (God bless his soul!) pretty much used the D-30 and D-60
> exclusively as far as I know, and his images were stupendous...AND...they
> were B&W:
>
> http://www.westernechoes.com/
>
> So, I'm not quite sure why you weren't getting prints that "sing", but the
> ones I saw from Jerry very much so did.
>
> Regards,
>
> Austin
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

Martin writes:

> I have never had a chance to play around with a better quality digital for
> B&W. The wife has a Canon G2 (3.1 mega pixel) for snap shots and my
attempts
> to get a nice print were somewhat successful if I printed very small, 3x4.
I
> have a handful of 13X18 B&W prints from other people done with SLR type
> cameras that are excellent. So I have to say it can be done.

Hmm...you've got me thinking now...I've never tried making small prints from
my digitial files. Now that you mention it I think Jerry Olsen remarked once
that he was generating 5x7's from his D30 files. I'll have to experiment
some and see what materializes. Thanks.

> That said, the feel is very different from prints made from scans of 8x10
> and 4x5 negs. I think it is simply a matter of how much fine detail is
> recorded and can be gotten onto a piece of paper. Prints from 35mm are
much
> different from prints from LF too but I like work from both formats.

Agreed. On a slightly off-topic comment...back when I produced only contact
prints from my 8x10 negs (I could not afford the cost of enlarging equipment
for this size neg) I always felt that "something" was at work during image
capture due to the viewing of your subject at actual print size. To this
day, I continue to marvel at the projected image on an 8x10 groundglass.

> As Austin said, Jerry Olson did some incredible B&W stuff using montages
of
> digital images that were a blend of scanned film and digital camera files.
> Excellent print quality. I do know he was looking to move up to the D60
and
> was on a quest for as much image data as possible.

Agreed. I love Jerry's images.

Alan Huntley

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Clayton Jones" 
<cj@c...> wrote:


> One of the problems in the debate aspect of this thread, 
> as I see it, is that some of the participants have so 
> far refused to acknowledge some of Anthony's important
> points. 

I think that's because Anthony has done such a god-awful job of 
articulating what he thinks the problem is.   We're dealing with very 
straightforward physical properties here.   Notice that in all my 
examples I use specific wavelengths and actual physical properties of 
specific light-emitting or absobing materials. 

. . . deletia . . . 
  

> Both I and another photographer have reported that we
> have so far failed in our efforts to get a color->bw 
> result that satisfies us, and I think that fact is an 
> important part of the topic.

I disagree, because it's not a "fact" it's a PERCEPTION.  (I suppose 
it's a fact that you have that perception, i.e., it's a fact that you 
are dissatisfied).   For it to be a fact that's useful to this 
discussion someone needs to specify what exactly is missing when you 
go from color to black and white, and where in the process it went 
missing.   Anthony said that informationed "between" the R,G, and B 
peaks was getting lost or attenuated and I have repeatedly asked 
Anthony to identify a wavelength of light where this problem occurs.

Photographic composition is an art.  But there is (or should be) 
no 'art' to getting a good scan and print - it's pure science and 
engineering.

----- Original Message ----- 

From: "A. Huntley" <leicam6@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Monday, May 26, 2003 10:45 AM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


(snip earlier)

> > That said, the feel is very different from prints made from scans of
8x10
> > and 4x5 negs. I think it is simply a matter of how much fine detail is
> > recorded and can be gotten onto a piece of paper. Prints from 35mm are
> much
> > different from prints from LF too but I like work from both formats.
>
> Agreed. On a slightly off-topic comment...back when I produced only
contact
> prints from my 8x10 negs (I could not afford the cost of enlarging
equipment
> for this size neg) I always felt that "something" was at work during image
> capture due to the viewing of your subject at actual print size. To this
> day, I continue to marvel at the projected image on an 8x10 ground glass.

Alan,

I remember Morley Baer letting me look through his old 8x10 and it was a
real eye opener in comparison to the puny 4x5 I was working with. I never
went to 8x10 due to the enlarger barrier but now with digital the economics
shift and 8x10 lends itself to less expensive scanning solutions than 4x5.

Martin

Martin writes:

> Alan,
>
> I remember Morley Baer letting me look through his old 8x10 and it was a
> real eye opener in comparison to the puny 4x5 I was working with. I never
> went to 8x10 due to the enlarger barrier but now with digital the
economics
> shift and 8x10 lends itself to less expensive scanning solutions than 4x5.
>
> Martin

I'm a little jealous and envious now...I absolutely love Morley's work and
have at least one video where he is interviewed. He seemed like an
unpretentious man...one of those rare photographers that is both GREAT and
approachable. I would have loved to do a workshop with him or, at least,
have met him.

Did you know that Brett Weston referred to the 4x5 as his miniature format
camera? I know that he used MF during his later years, but quite a few of us
seem to migrate downward in format as we age. I'm trying to talk my son into
lugging the 8x10, but he's not biting yet...smart boy!

Once I started using the 8x10 I pretty much gave up on 4x5 for about 10
years. To my eye, B&W contact prints are simply stunning and the 8x10 is
just big enough to see the image well. Though, my late friend Fred Picker
did some really lovely 4x5 contacts of snow/ice/water details of brooks in
Vermont. However, as you said with digital the economics of using 8x10 is
looking really good. The cost of film is way up there now--I used to buy 50
sheet boxes of Tri-X for about $50US--but, processing can be done with a few
cheap plastic trays, and probably just about any flatbed that can handle
this size film would be fine. My Epson 1680 does an acceptable job. Most of
the B&W quads I have produced from 8x10 have been roughly A4 size. I guess
the only real downside to the economics involved is that you'll want a LF
printer! <g> Epson 7000 anyone?

Alan Huntley

Hi Peter,

> I have repeatedly asked 
> Anthony to identify a wavelength of light where this problem occurs.

As have I, and the request gets completely ignored.

Regards,

Austin

Peter writes:

> Anthony said that informationed "between" the R,G,
> and B peaks was getting lost or attenuated and I
> have repeatedly asked Anthony to identify a wavelength
> of light where this problem occurs.

Measure the spectral distribution of a scene (with a spectrometer, for
example).  Then take a picture of that scene, in color (digital or film).
Now try to reconstruct the original spectral distribution curve using only
the three data points provided by the RGB values in the color image.  You
won't be able to do it, and you'll discover very quickly just how much
information is gone.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony 
Atkielski" <anthony@a...> wrote:
> Peter writes:
> 
> > Anthony said that informationed "between" the R,G,
> > and B peaks was getting lost or attenuated and I
> > have repeatedly asked Anthony to identify a wavelength
> > of light where this problem occurs.
> 
> Measure the spectral distribution of a scene (with a spectrometer, 
for
> example).  Then take a picture of that scene, in color (digital or 
film).
> Now try to reconstruct the original spectral distribution curve 
using only
> the three data points provided by the RGB values in the color 
image.  You
> won't be able to do it, and you'll discover very quickly just how 
much
> information is gone.

That's what you SAY, but you haven't provided a specific example.  
We've repeatedly asked you to give us objective support for what 
you're saying - but all you offer is gedanken experiments and 
hypothetical situations.  If what you're saying is true you ought to 
be able to provide concrete evidence of it.

Remember, your EYE doesn't reconstruct the spectral distribution of 
a scene either because it also relies on just 3 overlapping 
photopigments.   

If what you're saying is true it ought to be a piece of cake to 
identify a color that comes out with a noticably different density 
compared to some reference color when scanned from black and white 
versus color film.

Hi Peter,

> If what you're saying is true it ought to be a piece of cake to
> identify a color that comes out with a noticeably different density
> compared to some reference color when scanned from black and white
> versus color film.

And...keep in mind that grossly different colors produce the exact same
grayscale response!

Austin

Peter writes:

> That's what you SAY, but you haven't provided a
> specific example.

That's what the paragraph you backquoted was:

AA> Measure the spectral distribution of a scene (with a
AA> spectrometer, for example).  Then take a picture of
AA> that scene, in color (digital or film).  Now try to reconstruct
AA> the original spectral distribution curve using only
AA> the three data points provided by the RGB values in
AA> the color image.  You won't be able to do it, and you'll
AA> discover very quickly just how much information is gone.

See?

> We've repeatedly asked you to give us objective
> support for what you're saying ...

Who is "we"?  As far as I can tell, you are speaking only for yourself.
Several people here understand what I'm saying, even if you do not.  Perhaps
they can explain it better than I can, as I've given many different examples
and I seem to be making no progress.

> ... but all you offer is gedanken experiments and
> hypothetical situations.

Just like Einstein.

> If what you're saying is true you ought to
> be able to provide concrete evidence of it.

If you understand what I'm saying, you have all the evidence you need on
that basis alone.  Kind of like algebra, as opposed to arithmetic.

> Remember, your EYE doesn't reconstruct the spectral
> distribution of a scene either because it also
> relies on just 3 overlapping photopigments.

It doesn't have to.  It doesn't perform any conversions to other
colorspaces.

If you think about color blindness (specifically, the anomalous
trichromacies, not achromatopsia or deuteranopia or anything like that),
that might help to understand the difficulties involved.

> If what you're saying is true it ought to be a
> piece of cake to identify a color that comes out
> with a noticably different density compared to some
> reference color when scanned from black and white
> versus color film.

No, that does not follow.

Hi Anthony,

> > We've repeatedly asked you to give us objective
> > support for what you're saying ...
>
> Who is "we"?

I'm the other part of "we"...

> > If what you're saying is true you ought to
> > be able to provide concrete evidence of it.
>
> If you understand what I'm saying, you have all the evidence you need on
> that basis alone.  Kind of like algebra, as opposed to arithmetic.
>
> > If what you're saying is true it ought to be a
> > piece of cake to identify a color that comes out
> > with a noticeably different density compared to some
> > reference color when scanned from black and white
> > versus color film.
>
> No, that does not follow.

Noticeably different density is not relevant as far as I am concerned, as
that can be corrected as long as it is deterministic.  What you need to show
to show that this won't work is a frequency that IS distinctly represented
on B&W film, and is NOT distinctly represented on color.

Austin

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony 
Atkielski" <anthony@a...> wrote:
> Peter writes:
> 
> > That's what you SAY, but you haven't provided a
> > specific example.
> 
> That's what the paragraph you backquoted was:
> 
> AA> Measure the spectral distribution of a scene (with a
> AA> spectrometer, for example).  Then take a picture of
> AA> that scene, in color (digital or film).  Now try to reconstruct
> AA> the original spectral distribution curve using only
> AA> the three data points provided by the RGB values in
> AA> the color image.  You won't be able to do it, and you'll
> AA> discover very quickly just how much information is gone.
> 
> See?


You really don't know what "empirical" means, do you?


> If you understand what I'm saying, you have all the 
> evidence you need on that basis alone. 

"Evidence" is not what someone hypothesizes - evidence has to exist 
empirically.  You haven't provided ANY evidence.

Austin also asked you for a clear, real world example, so I'm not 
the only one.    

Why don't you find a common everyday object - some consumer product 
any of us can buy at the local drugstore if we want to duplicate 
your results - say, a can of Pepsi, a Kodak film box, Head and 
Shoulders shampoo, etc.  Shoot it at noon on a sunny day with Portra 
160 and TMax (or whatever films you want).  Scan them both in and 
convert the color image to grayscale and show us where you think it 
goes wrong.   Then we can repeat your experiment with the same 
subject, lighting, and film, and see if we can't do better.

Go to my website where I posted scans of various test images if you 
want to see examples of empirical evidence.   I don't say "pretend 
you set your printer to 'black only' use EEM paper, 2880 DPI..." 
etc.    I DO it and post the results.  That's what you need to do.

Peter writes:

> You really don't know what "empirical"
> means, do you?

Yes, I do.

Can you compose an entire post without insulting me in some way?

> "Evidence" is not what someone hypothesizes -
> evidence has to exist empirically.

No, it does not.  For example, it is sufficient to say that x = a+b,
therefore b = x-a without the need to "empirically" provide "evidence" for
this by inserting actual numbers into these equations.

Demonstrating the principle I'm discussing here by inserting actual numbers
would be extraordinarily time-consuming.

> Austin also asked you for a clear, real world
> example, so I'm not the only one.

Austin is having the same trouble understanding it, which surprises me, as
he does not seem unintelligent.

Hi Anthony,

> Austin is having the same trouble understanding it, which surprises me, as
> he does not seem unintelligent.

I do understand EXACTLY what you are saying, AND I simply disagree with it.

Austin

Austin writes:

> I do understand EXACTLY what you are saying, AND
> I simply disagree with it.

So how would you reconstruct a continuous, irregular curve of spectral
energy distribution from just three data points?

It's possible if the curve is a portion of, say, an ellipse.  But the real
world curves aren't, so you cannot reconstruct them from three data points.
And because of this, you cannot use RGB values to reconstruct the
information you would need to accurate simulate the spectral response of any
arbitrary monochrome or RGB sensor or film.  The required information just
isn't there, period.

Hi Anthony,

> > I do understand EXACTLY what you are saying, AND
> > I simply disagree with it.
>
> So how would you reconstruct a continuous, irregular curve of spectral
> energy distribution from just three data points?
>
> It's possible if the curve is a portion of, say, an ellipse.  But the real
> world curves aren't, so you cannot reconstruct them from three
> data points.
> And because of this, you cannot use RGB values to reconstruct the
> information you would need to accurate simulate the spectral
> response of any
> arbitrary monochrome or RGB sensor or film.  The required information just
> isn't there, period.

I understand that is what you are saying, and again, I simply disagree.  I
don't know what it is that is missing in yours and or my understanding, but
I simply don't see that it can't be done.  I have all the information I
need, the frequency and the intensity.  Both the color and B&W films
response is deterministic to the frequency and intensity...so I believe I
have the information necessary to map one to the other.  The converse is, of
course, not true, you can't go from B&W to color.

Regards,

Austin

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony 
Atkielski" <anthony@a...> wrote:
> Austin writes:
> 
> > I do understand EXACTLY what you are saying, AND
> > I simply disagree with it.
> 
> So how would you reconstruct a continuous, irregular curve of 
spectral
> energy distribution from just three data points?
> 
> It's possible if the curve is a portion of, say, an ellipse.  But 
the real
> world curves aren't, so you cannot reconstruct them from three data 
points.
> And because of this, you cannot use RGB values to reconstruct the
> information you would need to accurate simulate the spectral 
response of any
> arbitrary monochrome or RGB sensor or film.  The required 
information just
> isn't there, period.


But the lack of that precision may not have any practical effect.  
The reason why you need to provide empirical evidence is that you 
need to show that what you're talking about has any relevance in the 
real world.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Austin 
Franklin" <darkroom@i...> wrote:
> Hi Anthony,
> 
> > > I do understand EXACTLY what you are saying, AND
> > > I simply disagree with it.
> >
> > So how would you reconstruct a continuous, irregular curve of 
spectral
> > energy distribution from just three data points?
> >
> > It's possible if the curve is a portion of, say, an ellipse.  But 
the real
> > world curves aren't, so you cannot reconstruct them from three
> > data points.
> > And because of this, you cannot use RGB values to reconstruct the
> > information you would need to accurate simulate the spectral
> > response of any
> > arbitrary monochrome or RGB sensor or film.  The required 
information just
> > isn't there, period.
> 
> I understand that is what you are saying, and again, I simply 
disagree.  I
> don't know what it is that is missing in yours and or my 
understanding, but
> I simply don't see that it can't be done.  I have all the 
information I
> need, the frequency and the intensity.  Both the color and B&W films
> response is deterministic to the frequency and intensity...so I 
believe I
> have the information necessary to map one to the other.  The 
converse is, of
> course, not true, you can't go from B&W to color.

I agree with Austin.   Tell us precisely what information is 
missing.  Don't speak vaguely about the "curves" because we know what 
the response curve is of the color dyes used in the film.   So we 
know exactly how much to compensate the density by for any color.   
Or put another way:  because we know the shape of the film's response 
curves there is only one unique point on the spectrum that will 
produce a given density in all three dyes.  So what's missing? 

You say that other people here understand you, but I don't see them 
jumping in to clarify what you're saying.

Ah!

> It's possible if the curve is a portion of, say, an ellipse.  But
> the real
> world curves aren't, so you cannot reconstruct them from three data
> points.

Ah!  (again).  I think I see the confusion.  You can construct an ellipse
from three points (actually an arc), if the three points are X,Y
coordinates.  These three data values are not two dimensional.

To do the conversion from RGB to grayscale, you are not reconstructing a
curve with the three data points, you are reconstructing only a SINGLE point
in the XY space from the three data points.  Frequency is one vector and
intensity is the other.  That's it.  No curve, no ellipse etc.  Just one XY
point from three RGB values.  You do this for EVERY different RGB value that
is valid for the film type you are "simulating", and you have reconstructed
the grayscale response curve for that film.

Austin

> I agree with Austin.   Tell us precisely what information is
 > missing.  Don't speak vaguely about the "curves" because we know what
 > the response curve is of the color dyes used in the film.   So we
 > know exactly how much to compensate the density by for any color.
 > Or put another way:  because we know the shape of the film's response
 > curves there is only one unique point on the spectrum that will
 > produce a given density in all three dyes.  So what's missing?
 >
 > You say that other people here understand you, but I don't see them
 > jumping in to clarify what you're saying.

I think I need to weigh in again.  Perhaps no one read my previous post. 
*I* understand Anthony, at least on this point:

It is a basic mathematical fact that the process of binning is *not* 
reversible under almost all circumstances.

Let's say that we have a spectral input of amplitude A(F), where F is the 
frequency.  Let's take two simple response functions (the same applies to 
three, but I the extra lines clutter things up), r(F) and g(F).

If we take two binning functions to collect r and g...
R = integral over all F of (A(F)*r(F))
G = integral over all F of (A(F)*g(F))

Look at those last two equations.  If I give you the result of the 
integral (the R value, for instance) and the r(F) response function of the 
filter/CCD combination, can you tell me what the input A(F) is?

Answer: for all but the most trivial A(F) and r(F), no.  For an arbitrary 
A(F), never.

More simply:  if you know the exact spectral response of, say, the red 
element of a CCD sensor, and I tell you I got a red value of 130, can you 
tell me the original spectral input?

Answer: no.

-- 
Jon Dubovsky ( entropy@... )

I think it's probably simpler than that. Given a pure color -light of a 
single wavelength, say from a laser. Can the values of the R, B, G 
sensors be used to estimate the response of a B&W film to that same 
laser? Most likely not.

However, given a complex scene illuminated by white light, I would 
expect you could come close.

Truman

Jon Dubovsky wrote:

> >
>I think I need to weigh in again.  Perhaps no one read my previous post. 
>*I* understand Anthony, at least on this point:
>
>It is a basic mathematical fact that the process of binning is *not* 
>reversible under almost all circumstances.
>
>Let's say that we have a spectral input of amplitude A(F), where F is the 
>frequency.  Let's take two simple response functions (the same applies to 
>three, but I the extra lines clutter things up), r(F) and g(F).
>
>If we take two binning functions to collect r and g...
>R = integral over all F of (A(F)*r(F))
>G = integral over all F of (A(F)*g(F))
>
>Look at those last two equations.  If I give you the result of the 
>integral (the R value, for instance) and the r(F) response function of the 
>filter/CCD combination, can you tell me what the input A(F) is?
>
>Answer: for all but the most trivial A(F) and r(F), no.  For an arbitrary 
>A(F), never.
>
>More simply:  if you know the exact spectral response of, say, the red 
>element of a CCD sensor, and I tell you I got a red value of 130, can you 
>tell me the original spectral input?
>
>Answer: no.
>
>  
>

> I think it's probably simpler than that. Given a pure color -light of a
 > single wavelength, say from a laser. Can the values of the R, B, G
 > sensors be used to estimate the response of a B&W film to that same
 > laser? Most likely not.
 >
 > However, given a complex scene illuminated by white light, I would
 > expect you could come close.
 >
 > Truman

Nay, it is almost exactly the opposite.  The more constrained the input 
function A(F), the easier it is to guess it given the bin output 
functions.  The more complex (unconstrained) the input function (i.e. the 
reflection from a complicated scene, which has a very unpredictable 
spectral response) the less likely it is your guess will be right.

Can we put this matter to bed?  If need be, we can generate example 
functions and pose the challenge.

-- 
Jon Dubovsky ( entropy@... )

> More simply:  if you know the exact spectral response of, say, the red
> element of a CCD sensor, and I tell you I got a red value of 130, can you
> tell me the original spectral input?
>
> Answer: no.

Jon,

I don't understand the relevance of that.  You do NOT have only the red
element (unless it's only red, that is), you have three values.  Also, I do
not see how knowing the exact spectral response of the CCD is required to do
this.  It IS one method, but, IMO, a poor one at best, as it adds another
variable that is unnecessary.

What you do is simply create a "map" for the RGB values to grayscale values
using a LUT.  The LUT is generated by taking a full spectrum image with both
the B&W film, exposed and developed as you like and the color film, scanning
the two films on the same scanner, and "matching" the graytones to the RGB
tones at the same point in the image.  Of course there is going to be some
error, but I believe the error is insignificant.

I do not believe this needs to be calculated, it can be, if the mapping is
deterministic, what is called, characterized.  If scanner differences are an
issue, you simply characterize this for the different scanner, following the
same procedure.  If you do it once, you can do it a thousand times, for any
combination of films.

Austin

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Peter 
Nelson" <peter@s...> wrote:
> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, 
"Austin 
> Franklin" <darkroom@i...> wrote:
> > Hi Anthony,
> > 
> > > > I do understand EXACTLY what you are saying, AND
> > > > I simply disagree with it.
> > >
> > > So how would you reconstruct a continuous, irregular curve 
of 
> spectral
> > > energy distribution from just three data points?
> > >
> > > It's possible if the curve is a portion of, say, an ellipse.  But 
> the real
> > > world curves aren't, so you cannot reconstruct them from 
three
> > > data points.
> > > And because of this, you cannot use RGB values to 
reconstruct the
> > > information you would need to accurate simulate the 
spectral
> > > response of any
> > > arbitrary monochrome or RGB sensor or film.  The required 
> information just
> > > isn't there, period.
> > 
> > I understand that is what you are saying, and again, I simply 
> disagree.  I
> > don't know what it is that is missing in yours and or my 
> understanding, but
> > I simply don't see that it can't be done.  I have all the 
> information I
> > need, the frequency and the intensity.  Both the color and 
B&W films
> > response is deterministic to the frequency and intensity...so I 
> believe I
> > have the information necessary to map one to the other.  The 
> converse is, of
> > course, not true, you can't go from B&W to color.
> 
> I agree with Austin.   Tell us precisely what information is 
> missing.  Don't speak vaguely about the "curves" because we 
know what 
> the response curve is of the color dyes used in the film.   So we 
> know exactly how much to compensate the density by for any 
color.   
> Or put another way:  because we know the shape of the film's 
response 
> curves there is only one unique point on the spectrum that will 
> produce a given density in all three dyes.  So what's missing?

I suppose this is totally beaten to death, but...

The part I think you are missing is that given a particular
triplet of RGB densities you are NOT trying to go back to a
"unique point on the spectrum".   The light in a real scene that
exposed that point in the film is NOT one unique frequency.
Every single point is composed of the entire  visible spectrum.
You'll have an amplitude (probably different) for every single
frequency.   The three density numbers can't possibly tell
what all the amplitudes of all the frequencies are.  Sure it
tells you a fair amount about the general shape, and it's close 
enough for the human color perception.  But B&W film with a 
colored filter can be much more selective about what 
frequencies to be sensitve to and what ones not to be.  How
dark each piece of file gets is based on integrating over the
entire visible spectrum of the energy * filter * film sensitivity.
One RGB triplett just doesn't have enough information to
do that integration.

For the same RGB values you can and mostly will have 
different overall spectra and if you had taken the image 
with B&W and filter you could very likely have differentiation
in B&W that was lost if you went via RGB.

Roy

> 
> You say that other people here understand you, but I don't see 
them 
> jumping in to clarify what you're saying.

> I don't understand the relevance of that.  You do NOT have only the red
 > element (unless it's only red, that is), you have three values.  Also,
 > I do not see how knowing the exact spectral response of the CCD is
 > required to do this.  It IS one method, but, IMO, a poor one at best,
 > as it adds another variable that is unnecessary.
 >
 > What you do is simply create a "map" for the RGB values to grayscale
 > values using a LUT.  The LUT is generated by taking a full spectrum
 > image with both the B&W film, exposed and developed as you like and
 > the color film, scanning the two films on the same scanner, and
 > "matching" the graytones to the RGB tones at the same point in the
 > image.  Of course there is going to be some error, but I believe
 > the error is insignificant.
 >
 > I do not believe this needs to be calculated, it can be, if the
 > mapping is deterministic, what is called, characterized.  If scanner
 > differences are an issue, you simply characterize this for the
 > different scanner, following the same procedure.  If you do it once,
 > you can do it a thousand times, for any combination of films.

Except in the most trivial of cases, given the output of any number of 
binning functions, it is a mathematical impossibility to convert those 
into the outputs of a different set of binning functions.

Why?  The aggregation of A(F) into any number of outputs fewer than all of 
F results in a masking of A(F).  After that, it is impossible to 
reconstruct A(F).

Mathematically:  Beyond trivial response functions (i.e. an impulse 
function), multiple A(F)'s can produce the same outputs.  I.e. an output 
does *not* map to a unique input.

Finis.

-- 
Jon Dubovsky ( entropy@... )

Austin writes:

> I have all the information I need, the frequency
> and the intensity.

"The" frequency?  But for every pixel in the original scene, there are
multiple frequencies of light, each with its own intensity.  Indeed, in most
cases, there is an _infinite number_ of different frequencies, each with its
own intensity.  You can plot this on a graph, with the x axis set to
frequency and the y axis set to intensity.  The graph produces a continuous,
wiggly line.  And this line is different for every pixel in the original
scene.

Now, when you take a color picture, this line is collapsed into just three
numbers: red, green, and blue.  To do this, a one-way function is applied to
the original spectrum.  The red, green, and blue values are determined by
integrating the areas under three separate curves superimposed on the
original spectrum.  However, there are many different distributions of
energy under these curves that will produce any given red, green, or blue
value.  The conversion is many-to-one.  As a result it is irreversible.
Once you've recorded your three numbers, you can never reproduce the
spectral distribution that is responsible for creating those numbers,
because there exists an infinite number of different distributions that will
produce any given set of numbers.

In other words, once you've captured the image in color with an RGB capture
method, you can never again recreate the original spectral distribution of
the scene.

And this is where the problem arises.  Because, in order to accurately
duplicate the results that would be obtained in photographing the scene with
a _different_ capture device (color or black and white, it doesn't matter),
you _must_ have the original spectral information for the scene.  But you
can't get that now, because you lost it when you performed your one-way
conversion to RGB.  So the duplication is impossible.  In fact, even a
decent simulation may be impossible.  You'll never be able to simulate the
effects of a narrow-band color filter, for example; you can't even come
close.

Just think of the color yellow.  Yellow in RGB is represented by roughly
equal R and G values, and a low B value.  But for any given triplet of RGB
values perceived as yellow, there exists an _infinity_ of original spectral
distributions that can produce that triplet--and you have no way of knowing
which one of these distributions produced it in the original image.  And
even though all these distributions produce an identical result in your RGB
capture, to another device with a different spectral sensitivity (such as
B&W film behind a narrow-band color filter, or even B&W film by itself),
they may _not_ produce identical results; they may, in fact, produce
dramatically different results.  And there is no way for you to know which
results they might have produced with other capture methods.  As a result,
you cannot duplicate or simulate the results that would be obtained with
those other methods using only your RGB information.

> The converse is, of course, not true, you can't
> go from B&W to color.

Actually, that is only a specific and obvious instance of a much more
general problem.  You can't go from color to color, either (which is why you
cannot accurately simulate Velvia with a scan of Provia).  You can't
simulate the results from anything that would normally be a function of the
original spectral distribution, because you no longer have that
distribution, and you can't recreate it from a simple trio of numbers.

Peter writes:

> But the lack of that precision may not have any
> practical effect.

It doesn't require much precision.

It's very easy to demonstrate.  Take a picture in black and white using a
narrow-band color filter.  Those are not difficult to find.  Then take a
picture of the same scene in color without a filter.  Then try to convert
the color image to B&W in a way that duplicates the same shades of gray you
see in the B&W image taken with the filter.  You'll find that it cannot be
done.

> The reason why you need to provide empirical
> evidence is that you need to show that what you're
> talking about has any relevance in the real world.

Only to you.  The real-world relevance is self-evident to me and to several
other people.  The fact that you do not see it doesn't mean that it isn't
there.  See the example above, which is about as real-world as anything ever
gets.

Peter writes:

> I agree with Austin.   Tell us precisely what
> information is missing.

I have, again and again.  I've just tried again with Austin, although I'm
not optimistic.

> Don't speak vaguely about the "curves" because
> we know what the response curve is of the color
> dyes used in the film.

There is nothing vague about curves.

> So we know exactly how much to compensate the
> density by for any color.

You cannot compensate in this way.  In fact, I'm not even sure what you are
talking about, since it has nothing to do with the constraints I'm
describing.  This does not bode well.

> Or put another way:  because we know the shape
> of the film's response curves there is only one
> unique point on the spectrum that will produce
> a given density in all three dyes.

Wrong.  There are multiple distributions of light energy in the original
scene that will produce identical results in the dyes.  If that were not the
case, it would be impossible to produce "yellow" by mixing red and green
light.  That only works because many different spectral distributions are
recorded identically by RGB sensors, such as film (or CCDs, or the human
eye).

If you are convinced that every unique distribution of light frequencies in
the original scene produces a unique RGB result, then explain why mixing red
and green light produces exactly the same result as yellow light, even
though they have completely different spectral distributions.

> So what's missing?

Just about everything.

> You say that other people here understand you,
> but I don't see them jumping in to clarify what
> you're saying.

They may find it difficult to explain it in a better way, or maybe they lack
my patience.

Austin writes:

> To do the conversion from RGB to grayscale, you
> are not reconstructing a curve with the three data
> points ...

Yes, you are.  The response of a B&W sensor to the original scene is a
function of the complete curve present in the original scene.  To simulate
that response, you must have the original curve.  Therefore, to get that
response from an RGB image, you must have a way to accurately reconstruct
that curve from your RGB values.  Every possible original curve, in other
words, must have a unique RGB representation.  If there are any
duplications, then it will not be possible to construct the B&W image from
RGB data alone.  And unfortunately, in practice, there are many
duplications.  Each RGB value can represent an infinite number of different
original curves.  And since these original curves do not all produce
identical results on sensors with different spectral sensitivities from the
sensor that produced your RGB data, you cannot reproduce the B&W results
from your RGB information.

> Frequency is one vector and
> intensity is the other.  That's it.

But the original scene didn't contain just one frequency.  It contained an
entire spectrum of frequencies, for _each point_ in the image.  And you've
lost those in your conversion of the original light in the scene to a
triplet of RGB values.

> You do this for EVERY different RGB value that
> is valid for the film type you are "simulating",
> and you have reconstructed the grayscale response
> curve for that film.

You can't.  See above.

You've got it!  That's exactly the problem, and that's why you can never
simulate the response of one film or sensor using data from another film or
sensor.  You can't get a Tri-X look from a digital RGB sensor, and you can't
get a Velvia look from a Provia scan, and so on.

----- Original Message -----
From: "Jon Dubovsky" <entropy@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 28, 2003 04:27
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> > I agree with Austin.   Tell us precisely what information is
>  > missing.  Don't speak vaguely about the "curves" because we know what
>  > the response curve is of the color dyes used in the film.   So we
>  > know exactly how much to compensate the density by for any color.
>  > Or put another way:  because we know the shape of the film's response
>  > curves there is only one unique point on the spectrum that will
>  > produce a given density in all three dyes.  So what's missing?
>  >
>  > You say that other people here understand you, but I don't see them
>  > jumping in to clarify what you're saying.
>
> I think I need to weigh in again.  Perhaps no one read my previous post.
> *I* understand Anthony, at least on this point:
>
> It is a basic mathematical fact that the process of binning is *not*
> reversible under almost all circumstances.
>
> Let's say that we have a spectral input of amplitude A(F), where F is the
> frequency.  Let's take two simple response functions (the same applies to
> three, but I the extra lines clutter things up), r(F) and g(F).
>
> If we take two binning functions to collect r and g...
> R = integral over all F of (A(F)*r(F))
> G = integral over all F of (A(F)*g(F))
>
> Look at those last two equations.  If I give you the result of the
> integral (the R value, for instance) and the r(F) response function of the
> filter/CCD combination, can you tell me what the input A(F) is?
>
> Answer: for all but the most trivial A(F) and r(F), no.  For an arbitrary
> A(F), never.
>
> More simply:  if you know the exact spectral response of, say, the red
> element of a CCD sensor, and I tell you I got a red value of 130, can you
> tell me the original spectral input?
>
> Answer: no.
>
> --
> Jon Dubovsky ( entropy@... )
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

Truman writes:

> However, given a complex scene illuminated by
> white light, I would expect you could come close.

Sometimes.  But there are important real-world cases that just don't work.
You cannot simulate the results of shooting B&W through a filter by using
any transformation of RGB, for example.  You cannot simulate Velvia from
Provia, either; just increasing saturation doesn't work.  Many other
real-world examples could be cited.

This is why some people become very frustrated trying to exactly duplicate
the "look" of a B&W film (or any film) using the image from an electronic
camera.  It can't be done.  In fact, you can't duplicate the look of an
electronic camera on film, either, but since hardly anyone tries to do that,
nobody has noticed.

Austin writes:

> You do NOT have only the red element (unless
> it's only red, that is), you have three values.

"Unless it's only red" is sufficient to prove the point.  And in any case,
even with three values, you still have the same problem; see my trivial
example for yellow light.  Jon appears to be a better mathematician than I
am (which isn't difficult, I'll admit), and perhaps his explanation will be
clearer to you than mine.  I understand the concepts perfectly, but I hate
math.

> Why?  The aggregation of A(F) into any number of
> outputs fewer than all of F results in a masking
> of A(F).  After that, it is impossible to
> reconstruct A(F).

Which, incidentally, is the basis of message digest algorithms in
cryptography.

Hello Anthony,

The spectral response of BW film is a fairly simplistic curve
that 'could' be somewhat represented by three points .

Here is one for PlusX125  ( really favors Blue )
http://www.kodak.com/global/en/professional/support/techPubs/f4018/f009_0431ac.gif

Here is one for Porta400  ( much more linear )
http://www.kodak.com/global/en/professional/support/techPubs/f4012/f009_0390ac.gif

Here a typical RGB filter set ( this one needs an IR filter )
http://www.edmundoptics.com/images/IOD_Product_Images/4726.gif

Assuming the CCD system is properly calibrated, I think it would
be possible to reshape a white reference to enhance the blue end
of the curve and get a close PlusX BW response when extracting a
luminance layer. Porta400 might need a little push in the red.

Am I missing something ?  BW film is not particularly frequency
selective and it will respond to single frequency energy in
pretty much the same  manner as a digital CCD. In astrophotography,
a spectrometer can be used to spread sometimes quite specific
wavelengh energies across a strip of BW film. The BW film has
no problem seeing the bright and absorbtion regions from IR to UV.

I agree, a green laser and red laser would show as two points on
a spectrometer, the eye resolves the mix as if it was a single 
yellow value. The BW film is just going to create a grey shade
based upon its spectra sens of those freqs as will the 'colorized'
CCD, but a luminance response is just the integration of the 
RGB levels.

OR, in retrospect, perhaps the film curves subtile sensitivity
curves Illustrate Your Point of not being able to reconstruct
them perfectly with 'three points' .... ???

Best,
Alex



--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony
Atkielski" <anthony@a...> wrote:
> You've got it!  That's exactly the problem, and that's why you can never
> simulate the response of one film or sensor using data from another
film or
> sensor.  You can't get a Tri-X look from a digital RGB sensor, and
you can't
> get a Velvia look from a Provia scan, and so on.
>

Alex writes:

> The spectral response of BW film is a fairly
> simplistic curve that 'could' be somewhat
> represented by three points.

Just because the curve seems simple doesn't mean that the conversion is
possible.  Even a flat response cannot be simulated when filters are used.
For example, using a yellow filter on film with a flat response will produce
results that cannot be duplicated by any manipulation of an RGB image.

> OR, in retrospect, perhaps the film curves
> subtile sensitivity curves Illustrate Your
> Point of not being able to reconstruct
> them perfectly with 'three points' .... ???

The fundamental problem is that the B&W film response (or indeed the
response of any image-capture device, be it film or digital, color or B&W)
is a function of the spectral distribution of light in the original scene,
which includes an infinite number of different frequencies of light with an
intensity for each frequency, and this for each point in the image.  No form
of image capture records all this information, so once an image is capture,
there is no way to recreate the distribution in the original image, and
since converting an image from one form to another in order to simulate the
response of a different image-capture device requires this information from
the original image, there is no way to convert a RGB image to accurately
represent a B&W capture, or anything else.

Here's an example that might help:

I have a machine that collects money.  It provides one total for all coins,
and one total for all bills.  I have another machine that also collects
money, but it collects only quarters and five-dollar bills.

Using the first machine, I collect all the money I see around the house.
The totals displayed by the machine are $8.25 and $21.  Based on these
totals, I want to figure out what totals the second machine would have
given, had I used it to collect the money.  Can you explain how I would do
this?

The money around the house represents the spectral distribution of light in
an original scene.  The first machine represents color film; the second
machine represents B&W film.  Do you see the problem now?

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony
Atkielski" <anthony@a...> wrote:
> Here's an example that might help:
> 
I disagree on one point :
The BW money machine collects all money, adding a few pennies
or dimes, loosing a five, whatever, according to its spectral
response.

The Color money machine collects the money in small change,
large change, and dollar bills.  But, disregarding its
similar spectral response gains or loss, there is no money
not accumulated.

The color filters are such that midway between the Red and
Green, their respective responses are down by 50%, but the
sum total of their light energy is still One.

Your coin analogy is an excellent way to represent a frequency
distribution, as long as the pennies are worth as much as the 
five dollar bills <laugh>

I think the real variant is the S curve amplitude non linearity
between films rather than the spectral domain..

Alex


> I have a machine that collects money.  It provides one total for all
coins,
> and one total for all bills.  I have another machine that also collects
> money, but it collects only quarters and five-dollar bills.
> 
> Using the first machine, I collect all the money I see around the house.
> The totals displayed by the machine are $8.25 and $21.  Based on these
> totals, I want to figure out what totals the second machine would have
> given, had I used it to collect the money.  Can you explain how I
would do
> this?
> 
> The money around the house represents the spectral distribution of
light in
> an original scene.  The first machine represents color film; the second
> machine represents B&W film.  Do you see the problem now?

I'll try again:

You have a machine that collects money.  You run it in your house and it
manages to find $10.49.  You also have a piggy bank that can accept only
quarters.  Based on this total from your machine, how many quarters do you
have for your piggy bank?  See the problem?

The machine is a color sensor or film.  The total amount is the RGB value.
The piggy bank is a different sensor, one that only accepts parts of the
spectrum.  It should be obvious that, since the machine (sensor) gives you
only a total (RGB value), and not the distribution of coins and bills (color
frequencies), there is no way for you to determine how the piggy bank
(another sensor or film with different spectral sensitivities) would fill
(or record light).

Thus, you cannot convert an RGB value in a way that precisely duplicates the
results from a different image-capture device.  Often, you can't even come
close.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Jon Dubovsky 
<entropy@i...> wrote:> *I* understand Anthony, at least on this point:
> 
> It is a basic mathematical fact that the process of binning is 
*not* 
> reversible under almost all circumstances.


But technically this isn't "binning", because true binning is 
mutually exclusive.   It's either in this bin or that one.  In this 
case you have OVERLAPPING bins, and not only that but the sensitivity 
curve in each bin is different.   So the same pixel has a signal in 
bin A and a signal in bin B.  There is only one unique point on the 
spectrum that would produce any give signal in both bins.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Roy Harrington>

 
> The part I think you are missing is that given a particular
> triplet of RGB densities you are NOT trying to go back to a
> "unique point on the spectrum".   The light in a real scene that
> exposed that point in the film is NOT one unique frequency.
> Every single point is composed of the entire  visible spectrum.
> You'll have an amplitude (probably different) for every single
> frequency.   The three density numbers can't possibly tell
> what all the amplitudes of all the frequencies are.  Sure it
> tells you a fair amount about the general shape, and it's close 
> enough for the human color perception.  But B&W film with a 
> colored filter can be much more selective about what 
> frequencies to be sensitve to and what ones not to be.  How
> dark each piece of file gets is based on integrating over the
> entire visible spectrum of the energy * filter * film sensitivity.
> One RGB triplett just doesn't have enough information to
> do that integration.

But the practical question in this discussion is whether it's good 
enough.   What Anthony has not established is that there is any 
PRACTICAL significance to that.   

When I have a point to make here I do some tests and scan in the 
results and post them. You can see them on my website.   Anthony 
could save us a lot of bandwidth if he would do the same.   

I'm an engineer and we have a motto: "The difference between theory 
and practice is a lot greater in practice than it is in theory."

I just recieved my MIS 7600 Archival Inkset yesterday. These are color
inks designed to replace the Ultrachromes. I bought them for the
occasional color print, and so far I'm very happy their performance in
that respect.

I don't intend to do any extensive testing of their BW performance since
I use the Ultratones for that. However, I did make a few BW prints on
EAM and here's my observations:

- There is still metamerism

- The shift is more towards blue in sunlight than towards green

- The shift is smaller than with the Ultrachromes, but visible enough
that I wouldn't be comfortable selling or giving away BW prints made
with them.

- I like blue better than green, so if I had to use a color inkset with
the Epson driver for BW printing I would prefer this over the
Ultrachromes.

--
Daniel Staver
http://daniel.staver.no

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony 
Atkielski" <anthony@a...> wrote:
> I'll try again:
> 
> You have a machine that collects money.  You run it in your house 
and it
> manages to find $10.49.  You also have a piggy bank that can accept 
only
> quarters.  Based on this total from your machine, how many quarters 
do you
> have for your piggy bank?  See the problem?

Bad analogy.    Because the currency is discreet quanta (it's 25 or 5 
or 10 - there's no 7.25 cent piece) but wavelenthgs are continuous, 
and furthermore RGB sensors (and color film dyes) OVERLAP in their 
sensitivity. 

Science made a huge leap forward when it escaped from Aristotelian 
word-logic and went to an EXPERIMENTAL paradigm.   You could do the 
same.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony 
Atkielski" <anthony@a...> wrote:
> Truman writes:
> 
> > However, given a complex scene illuminated by
> > white light, I would expect you could come close.
> 
> Sometimes.  But there are important real-world cases that just 
don't work.

You're a fine one to talk about the real world, since in your dozens 
of posts on this you have yet to cite even one IOTA of real-world 
evidence or data.   Your entire commentary has been purely 
hypothetical.

It reminds me of a joke in my college math class.   Some student says 
to the professor, "OK, fine, but can you show us a real-world 
example?"  and the professor goes to his whiteboard and says, "Sure.  
Take two values, x and y, in r-space . . . "

Peter writes:

> But technically this isn't "binning", because
> true binning is mutually exclusive.

Yes, it is.  The RGB channels are mutually exclusive.  But even if you don't
look at it that way, the same principle applies.  You are discarding
information, and you can't recover it after it's gone.  Like a one-way hash
algorithm.

> So the same pixel has a signal in
> bin A and a signal in bin B.  There is only one
> unique point on the spectrum that would produce
> any give signal in both bins.

Not so.  Yellow light will produce exactly the same signal as a certain
combination of red and green light, as I've explained over, and over, and
over.

Maybe I should try flash cards and puppets.

----- Original Message -----
From: "Peter Nelson" <peter@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 28, 2003 14:15
Subject: Re: [Digital BW] Digital, film, Spectral Sensitivity


> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony
> Atkielski" <anthony@a...> wrote:
> > I'll try again:
> >
> > You have a machine that collects money.  You run it in your house
> and it
> > manages to find $10.49.  You also have a piggy bank that can accept
> only
> > quarters.  Based on this total from your machine, how many quarters
> do you
> > have for your piggy bank?  See the problem?
>
> Bad analogy.    Because the currency is discreet quanta (it's 25 or 5
> or 10 - there's no 7.25 cent piece) but wavelenthgs are continuous,
> and furthermore RGB sensors (and color film dyes) OVERLAP in their
> sensitivity.
>
> Science made a huge leap forward when it escaped from Aristotelian
> word-logic and went to an EXPERIMENTAL paradigm.   You could do the
> same.
>
>
>
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

Peter, I'm with you - this whole debate by Anthony is a troll. Arguing my
analogy quickly descends into arguing whether the analogy is valid. Until
Anthony buys a scanner and an inkjet printer, all this is crazy!

Julian
----- Original Message ----- 
From: "Peter Nelson" <peter@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 28, 2003 2:15 PM
Subject: Re: [Digital BW] Digital, film, Spectral Sensitivity


> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony
> Atkielski" <anthony@a...> wrote:
> > I'll try again:
> >
> > You have a machine that collects money.  You run it in your house
> and it
> > manages to find $10.49.  You also have a piggy bank that can accept
> only
> > quarters.  Based on this total from your machine, how many quarters
> do you
> > have for your piggy bank?  See the problem?
>
> Bad analogy.    Because the currency is discreet quanta (it's 25 or 5
> or 10 - there's no 7.25 cent piece) but wavelenthgs are continuous,
> and furthermore RGB sensors (and color film dyes) OVERLAP in their
> sensitivity.
>
> Science made a huge leap forward when it escaped from Aristotelian
> word-logic and went to an EXPERIMENTAL paradigm.   You could do the
> same.
>
>
>
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

Jon,

> Mathematically:  Beyond trivial response functions (i.e. an impulse
> function), multiple A(F)'s can produce the same outputs.  I.e. an output
> does *not* map to a unique input.

It does NOT have to map to a unique input, it has to map to the same
response as the B&W film does, and as you should know, B&W film renders
different colors the same.

This is all quite interesting, as MANY people convert RGB images to
grayscale, and they look quite good.  If this was so inaccurate and
horrible, then how DOES that work so well?

Austin

Anthony,

> It's very easy to demonstrate.  Take a picture in black and white using a
> narrow-band color filter.  Those are not difficult to find.  Then take a
> picture of the same scene in color without a filter.  Then try to convert
> the color image to B&W in a way that duplicates the same shades
> of gray you
> see in the B&W image taken with the filter.  You'll find that it cannot be
> done.

Who cares about that, but you?  I don't use colored filters for my B&W
images.  Ever.

You know, if this just doesn't work, then how come it DOES in fact work?  I
see B&W images made from color ALL the time, and they look VERY close to
"same" B&W images shot with B&W film...  So, how is it that thousands of
people already do what you claim impossible?

> The real-world relevance is self-evident to me and
> to several
> other people.

As it is to me too.  I can take a color image.  I can take a B&W image of
the same scene.  I can scan them both.  I can manipulate the color image
tonality such that it VERY closely matches the tonality from the scanned B&W
image.  I can do this manually, or I can do this programmatically, as I
believe it's a deterministic mapping.  As I said, thousands of people make
B&W images from RGB data every day.  Of course, some better than others.
But, as I said, if so many people are doing this so successfully, how can
you believe it's impossible?  I know, you'll argue the meaning of
"duplicate" and "successful" etc.

Austin

Anthony,

> > I agree with Austin.   Tell us precisely what
> > information is missing.
>
> I have, again and again.

No, actually, you haven't...  I know you believe you have, but you're
missing the point.

> I've just tried again with Austin, although I'm
> not optimistic.

But, as I've said, I see what you claim can't be done, done...

> > Or put another way:  because we know the shape
> > of the film's response curves there is only one
> > unique point on the spectrum that will produce
> > a given density in all three dyes.
>
> Wrong.  There are multiple distributions of light energy in the original
> scene that will produce identical results in the dyes.

Name them, please, and then after you do, show that this is significant.
You are aware that B&W film gives the same tonality for many different
colors?

> If you are convinced that every unique distribution of light
> frequencies in
> the original scene produces a unique RGB result, then explain why
> mixing red
> and green light produces exactly the same result as yellow light, even
> though they have completely different spectral distributions.

Again, show that that is significant.  I do not believe it is, you are not
trying to replicate colors, just how B&W film would have, which, as I've
pointed out, does not require uniqueness in color rendition.

Austin

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Julian Thomas" 
<julianthomas@t...> wrote:
> Peter, I'm with you - this whole debate by Anthony is a troll. 
Arguing my
> analogy quickly descends into arguing whether the analogy is valid. 
Until
> Anthony buys a scanner and an inkjet printer, all this is crazy!
> 
> Julian

Anyone get anything usefull out of this tread or am I missing 
something ?
Cheers,
Andre

Anthony,

> > I have all the information I need, the frequency
> > and the intensity.
>
> "The" frequency?  But for every pixel in the original scene, there are
> multiple frequencies of light, each with its own intensity.

What, precisely, does B&W film "sense"?  What "information" is recorded on
B&W film, and what information do the scanned values 0-255, let's say, give
you?  NOTHING but intensity, period.  That's it.  Light to dark.

How is this information derived from the original "scene"?  Both the film,
and the CCD, sense the number of photons that "hit" it.  They are spectral
independent.  Look at the film response curve for B&W film.  It's reasonably
flat, up to the point of fall-off.  That means that all photons, no matter
of what frequency, are treated, for all practical purposes, equally, and
generate the same results on the film, recording the average intensity of
light at "that" point in space.

Now, does the RGB data not have this "intensity" information?  How accurate
can you extract intensity from the three RGB values?  Does it matter that
different combinations of RGB produce the same intensity?  Only if that's
not how the B&W film would have seen it as well...which, is why different
colors produce the same results on B&W film...so the answer seems to be no,
it doesn't matter.

Austin

>>It is a basic mathematical fact that the process of binning is
 >>*not* reversible under almost all circumstances.
 >
 > But technically this isn't "binning", because true binning is
 > mutually exclusive.   It's either in this bin or that one.  In this
 > case you have OVERLAPPING bins, and not only that but the sensitivity
 > curve in each bin is different.   So the same pixel has a signal in
 > bin A and a signal in bin B.  There is only one unique point on the
 > spectrum that would produce any give signal in both bins.

Mathematically, it is the same.  Whether the response functions overlap or 
not, except for trivial cases (again, such as an impulse function for the 
response function), the impossibility of conversion still holds.

-- 
Jon Dubovsky ( entropy@... )

> -----Original Message-----
> From: Anthony Atkielski [mailto:anthony@...]
> Sent: Wednesday, May 28, 2003 8:24 AM
> To: DigitalBlackandWhiteThePrint@yahoogroups.com
> Subject: Re: [Digital BW] Digital, film, Spectral Sensitivity
> 
> 
> Maybe I should try flash cards and puppets.

I'm game ;-)

Austin

Anthony,

> Sometimes.  But there are important real-world cases that just don't work.
> You cannot simulate the results of shooting B&W through a filter by using
> any transformation of RGB, for example.

That's not relevant to the discussion.  Let's just stick to simple grayscale
conversion from RGB values.

> You cannot simulate Velvia from
> Provia, either; just increasing saturation doesn't work.  Many other
> real-world examples could be cited.

But again, that's not relevant for this discussion...

> This is why some people become very frustrated trying to exactly duplicate
> the "look" of a B&W film (or any film) using the image from an electronic
> camera.  It can't be done.

Hum.  I see it done all the time.

> In fact, you can't duplicate the look of an
> electronic camera on film, either, but since hardly anyone tries
> to do that,
> nobody has noticed.

The ability to "process" the data is significantly reduced trying to do that
with film.

Austin

Anthony,

> You can't get a Tri-X look from a digital RGB sensor,
> and you can't
> get a Velvia look from a Provia scan, and so on.

Those are two entirely different problems.  One is spectrally sensitive, and
the other is simply intensity sensitive, with no spectral content.

From what I can tell from the Tri-X film response curve, it responds
equally, until falloff, to the number of photons that hit it, and this
response is spectrally independent.  Is that not the case?

Austin

CCD's and film are energy detectors - their response is directly 
proportional to the incident energy in a given frequency range. 
Different types of sensors/film, X-ray, UV, IR etc. are sensitive to 
different frequency ranges. Imaging radar, MRI, etc. perform the same 
function except they are active and the processing of the returns 
measures the energy profile to produce an image.

Film, CCD's, etc. are not coherent in that they don't record any phase 
information. RGB channels are nothing more than the energy in three 
bands of the spectrum. The only way you could not reconstruct the 
intensity of white light would be if there was a frequency that fell 
outside the passband of all three of the RGB filters. This is the key 
question. Otherwise the RGB output can be calibrated to produce a 
equivalent gray scale of any film if you know the characteristics of all 
the sensors involved. There are some questions in my mind about fidelity 
of this conversion. How does the noise floor of the combination of the 
three filters covert to the noise floor of the gray scale generated from 
the RGB conversion compare to the noise floor of the same image taken 
directly on a intensity only (B&W) gray scale sensor?  I consider that 
the real question.

Truman

Austin Franklin wrote:

>Anthony,
>
>  
>
>>    
>>
>
>What, precisely, does B&W film "sense"?  What "information" is recorded on
>B&W film, and what information do the scanned values 0-255, let's say, give
>you?  NOTHING but intensity, period.  That's it.  Light to dark.
>
>How is this information derived from the original "scene"?  Both the film,
>and the CCD, sense the number of photons that "hit" it.  They are spectral
>independent.  Look at the film response curve for B&W film.  It's reasonably
>flat, up to the point of fall-off.  That means that all photons, no matter
>of what frequency, are treated, for all practical purposes, equally, and
>generate the same results on the film, recording the average intensity of
>light at "that" point in space.
>
>Now, does the RGB data not have this "intensity" information?  How accurate
>can you extract intensity from the three RGB values?  Does it matter that
>different combinations of RGB produce the same intensity?  Only if that's
>not how the B&W film would have seen it as well...which, is why different
>colors produce the same results on B&W film...so the answer seems to be no,
>it doesn't matter.
>
>Austin
>
>
>
>Please visit the Group Homepage to check the Files, Bookmarks, Polls and other resources as they are often being updated. The page is at:
>
>http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
>If you wish to receive no emails or just a daily digest, or you wish to unsubscribe, please edit your Membership preferences by visiting this same page.
>
>Please follow these basic guidelines:
>- Include your full name with your message.
>- Include the address of your website, if you have one.
>- As threads develop, trim off excess portions of earlier messages to keep them short.
>- As the topic of a thread changes remember to change the subject header.
>- Good manners are required at all time. No personal attacks or flames
>- Complete your Yahoo profile.
>- Before posting a question, search the message archives and the various resources on the homepage. 
>
>
> 
>
>Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ 
>
>
>
>  
>



[Non-text portions of this message have been removed]

>>Mathematically:  Beyond trivial response functions (i.e. an impulse
 >>function), multiple A(F)'s can produce the same outputs.  I.e. an output
 >>does *not* map to a unique input.
 >
 > It does NOT have to map to a unique input, it has to map to the same
 > response as the B&W film does, and as you should know, B&W film renders
 > different colors the same.
 >
 > This is all quite interesting, as MANY people convert RGB images to
 > grayscale, and they look quite good.  If this was so inaccurate and
 > horrible, then how DOES that work so well?

Ah, kind sir, I never asserted that one cannot get great results from the 
process.  :)  In fact, I've printed a large number of quite pleasing b&w 
prints from color originals.  However, one of your original statements,

 >I have all the information I need, the frequency and the intensity.
 >Both the color and B&W films response is deterministic to the
 >frequency and intensity...so I believe I have the information
 >necessary to map one to the other.

..is a mathematical and practical fallacy.

Simple example:

Let's say that you're using Kodak's sensor listed here:
http://members.aol.com/modernimaging/mi/Kodak_DCS-620x_Technology.htm
and Tri-X:
http://www.kodak.com/global/en/professional/support/techPubs/f9/f9.pdf 
(see page 9 for the spectral response curve)

Let's zoom in on one little cel of our test image... you have a nice blue 
object.. let's say it comes out to one very dominant at 470 nm.  Let's say 
that it's about as bright as the whole of the scene.  (We do the latter 
just so we can skip over talking about development times, over- or 
under-exposure, etc.)  Let's say that right next to it you have another 
blue object at 430 nm, same brightness.

On the CCD, these two blues excite the sensor the same amount.  They will 
both be "weighted" the same amount.  You'll get some output value B, with 
no output values R or G (their response curve is 0 at that point.. the 
argument holds even if they aren't).  Compared to the overall brightness 
of the scene, you'll get some number for B, say.. oh.. 100, for each of them.

On the Tri-X, notice that it's spectral response curve is falling through 
the blue region.  It will weight the 470 nm color less than the 430 nm 
color.  The first cel will get a brightness value (compared to the overall 
brightness of the scene) of, oh, say, 100, and the second will get a 
brightness value of, oh, say, 90.

So does RGB(0,0,100) map to B&W(90) or B&W(100)?  Ah, there's the rub... 
it maps to both, but you no longer have enough information about the input 
to be able to determine which.  Again: the mapping is not unique.

So, when you convert your RGB image to black and white to try to emulate 
Tri-X, are you going to map that triplet to 90 or 100?  Or any of the 
other numerous values it might be?

Let me state again:  it is a simple mathematical and real-world-proven 
fact that one cannot convert the output of a binning process into the 
output of another non-identical binning process in all but the most 
trivial of cases (that is, trivial response functions, such as impulse 
functions).  I see this most every week at work.

Again: this is *not* to say one can't get great b&w images by desaturating 
color images.  See my disclaimer above.  I love my trusty G2, my NPH, you 
name it.  However, one cannot claim to be able to reproduce sensor X's 
output by shaping the output curves of non-identical sensor Y, whether 
that is comparing film to film or CCD to CCD or CCD to film.  That's just 
bad science.

Peace be with you all, and signing off of this discussion to go take some 
pictures in beautiful San Diego,
-- 
Jon Dubovsky ( entropy@... )

Hi Truman,

> CCD's and film are energy detectors - their response is directly
> proportional to the incident energy in a given frequency range.

Correct.

> Different types of sensors/film, X-ray, UV, IR etc. are sensitive to
> different frequency ranges. Imaging radar, MRI, etc. perform the same
> function except they are active and the processing of the returns
> measures the energy profile to produce an image.

Agreed.

> Film, CCD's, etc. are not coherent in that they don't record any phase
> information. RGB channels are nothing more than the energy in three
> bands of the spectrum.

Agreed.

> The only way you could not reconstruct the
> intensity of white light would be if there was a frequency that fell
> outside the passband of all three of the RGB filters.

Well, that's what I'd like to believe...  I'm sure that somewhere along the
spectral curve that the response needs to be "adjusted", as there is
sensitivity falloff at the ends of each colors response curve.

> This is the key
> question.

Exactly.

> Otherwise the RGB output can be calibrated to produce a
> equivalent gray scale of any film if you know the characteristics of all
> the sensors involved.

BINGO.

> There are some questions in my mind about fidelity
> of this conversion.

It's a concern, certainly, but I'm not sure it's significant, given,
remember, we are only sensitive to around 100 gray levels anyway...

> How does the noise floor of the combination of the
> three filters covert to the noise floor of the gray scale generated from
> the RGB conversion compare to the noise floor of the same image taken
> directly on a intensity only (B&W) gray scale sensor?  I consider that
> the real question.

Let's say it's insignificant, for arguments sake.  So, I take it you now
agree that:

1) B&W film merely records intensity only, and has no spectral "influence"
(that can't be corrected)
2) That RGB data contains this intensity information (to within a degree of
accuracy, of course)
3) That the same (to within a degree of accuracy) intensity information that
would have been "seen" by the B&W film can be extracted (with appropriate
LUTs to correct for "distortion") from the RGB data?

So, if this is true, then two questions remain...one I've incessantly asked,
with no response...

1) Is there any frequency in the B&W film's response range that is not
covered by the RGB response range?

2) Characterizing noise/distortion, and can it be corrected such that the
grayscale image is "replicated"?

Regards,

Austin

Jon,

>  > This is all quite interesting, as MANY people convert RGB images to
>  > grayscale, and they look quite good.  If this was so inaccurate and
>  > horrible, then how DOES that work so well?
>
> Ah, kind sir, I never asserted that one cannot get great results from the
> process.  :)  In fact, I've printed a large number of quite pleasing b&w
> prints from color originals.  However, one of your original statements,
>
>  >I have all the information I need, the frequency and the intensity.
>  >Both the color and B&W films response is deterministic to the
>  >frequency and intensity...so I believe I have the information
>  >necessary to map one to the other.
>
> ..is a mathematical and practical fallacy.

Apparently not.  For two reasons.  One, the spectral information is
irrelevant, and two, the intensity information IS contained in the RGB data.

> Simple example:
>
> Let's say that you're using Kodak's sensor listed here:
> http://members.aol.com/modernimaging/mi/Kodak_DCS-620x_Technology.htm

Well, that's a Bayer pattern sensor, and isn't what is used in film
scanners...but fine, that works for your example.

> you have a nice blue
> object.. let's say it comes out to one very dominant at 470 nm.
> Let's say
> that it's about as bright as the whole of the scene.  (We do the latter
> just so we can skip over talking about development times, over- or
> under-exposure, etc.)  Let's say that right next to it you have another
> blue object at 430 nm, same brightness.
>
> On the CCD, these two blues excite the sensor the same amount.  They will
> both be "weighted" the same amount.  You'll get some output value B, with
> no output values R or G (their response curve is 0 at that point.. the
> argument holds even if they aren't).  Compared to the overall brightness
> of the scene, you'll get some number for B, say.. oh.. 100, for
> each of them.
>
> On the Tri-X, notice that it's spectral response curve is falling through
> the blue region.

Er, insignificantly...

> It will weight the 470 nm color less than the 430 nm
> color.  The first cel will get a brightness value (compared to
> the overall
> brightness of the scene) of, oh, say, 100, and the second will get a
> brightness value of, oh, say, 90.

Your numbers are fictitious.  How do you know it's not 1026 vs 1032, which,
as far as your eyes can tell, show NO difference?

> So does RGB(0,0,100) map to B&W(90) or B&W(100)?  Ah, there's the rub...

Only if you show that it's significant in reality, which I do not believe
you can.

> it maps to both, but you no longer have enough information about
> the input
> to be able to determine which.  Again: the mapping is not unique.

And it does not have to be!

> So, when you convert your RGB image to black and white to try to emulate
> Tri-X, are you going to map that triplet to 90 or 100?  Or any of the
> other numerous values it might be?
>
> Let me state again:  it is a simple mathematical and real-world-proven
> fact that one cannot convert the output of a binning process into the
> output of another non-identical binning process in all but the most
> trivial of cases (that is, trivial response functions, such as impulse
> functions).  I see this most every week at work.

And, as I've said, I believe that's not relevant.

> Again: this is *not* to say one can't get great b&w images by
> desaturating
> color images.  See my disclaimer above.  I love my trusty G2, my NPH, you
> name it.  However, one cannot claim to be able to reproduce sensor X's
> output by shaping the output curves of non-identical sensor Y, whether
> that is comparing film to film or CCD to CCD or CCD to film.  That's just
> bad science.

I believe you are mis-applying "science" here.  Let's just leave it to good
engineering ;-)  I've designed similar systems that work just fine, where I
reconstruct, via characterization, the original response.  I'm not saying
that works for every instance, but for THIS instance, I believe the correct
information is there in the RGB data to do so.

The thing that you're missing is significance.  Yes, there are differences,
but is this difference discernable in the final output?  I say no, and you,
or anyone else, has yet to prove that not true.

I'll go back to pointing out that development and exposure are CRITICAL to
what you get for a "look" out of Tri-X.  Given the WIDE variance that Tri-X
can provide, I believe that shows that I in fact have a wide variance to
which I can "simulate" to.

Austin

Julian writes:

> Until Anthony buys a scanner and an inkjet printer,
> all this is crazy!

Anthony already has three film scanners, two ink-jet printers, a PostScript
laser printer, a flatbed scanner, and a dye-sub printer.

Austin writes:

> ... as you should know, B&W film renders
> different colors the same.

That depends on the exact colors, and on the exact film.  And filters can
further modify the results.

> This is all quite interesting, as MANY people
> convert RGB images to grayscale, and they look
> quite good.  If this was so inaccurate and
> horrible, then how DOES that work so well?

Nobody ever said it wouldn't look good.  It just will never look identical
to a B&W capture of the same original scene.  In some cases, it won't even
be close.

Austin writes:

> Who cares about that, but you?

A lot of people.  Many photographers use colored filters for black and
white, in order to change the response to light.  A red filter, for example,
darkens the sky considerably.

> You know, if this just doesn't work, then how
> come it DOES in fact work?

It _doesn't_ work.

> I see B&W images made from color ALL the time,
> and they look VERY close to "same" B&W images
> shot with B&W film...

Very close is not the same.  And sometimes it's not even very close.

> So, how is it that thousands of people already
> do what you claim impossible?

They don't.

> I can manipulate the color image tonality such that
> it VERY closely matches the tonality from the scanned
> B&W image.

Show me.  Use Tri-X as the black and white film.  Also, do the same thing
with a narrowband yellow filter over the B&W exposure.

Anthony,

> > ... as you should know, B&W film renders
> > different colors the same.
>
> That depends on the exact colors, and on the exact film.

Well, the statement was that it occurs, and it does occur, and it does occur
all the time.  It don't think it's relevant what colors and what film.

> And filters can
> further modify the results.

Please drop the filter thing.  It's not part of the discussion.  If it
interests you, fine, but it has not a WIT to do with my claim.

> > This is all quite interesting, as MANY people
> > convert RGB images to grayscale, and they look
> > quite good.  If this was so inaccurate and
> > horrible, then how DOES that work so well?
>
> Nobody ever said it wouldn't look good.  It just will never look identical
> to a B&W capture of the same original scene.  In some cases, it won't even
> be close.

Well, it depends on what you mean by identical and close.  I STILL believe
exactly what I've said, that the information is in RGB to produce a visually
identical B&W tonality, and no one has shown that it won't work...and for
good reason, because it DOES work.

The only reason it won't be close is if someone did something incorrectly.

Austin

Austin writes:

> No, actually, you haven't...  I know you believe
> you have, but you're missing the point.

Explain to me why this statement would be true when you utter it, but not
true when I utter it.

> Name them, please ...

I already have.  Spectral yellow is rendered identically to a certain
combination of spectral red and green.

> ... and then after you do, show that this is significant.

If it were not significant, no color display or printing system that exists
today would work.

> You are aware that B&W film gives the same tonality
> for many different colors?

Yes.  And it also gives different tones for many different color as well.
And often the response of the film is essentially random across the
spectrum.

> Again, show that that is significant.

I have, probably at least a dozen times by now.  If you didn't understand it
the first eleven times, why would you understand it now?

Anthony,

> > Who cares about that, but you?
>
> A lot of people.

Again, so what?  That in NO WAY negates the premise that you CAN in fact
convert the RGB information into grayscale tonality in such a way that it is
visually indistinguishable than if the scene was photographed using Tri-X.

> Many photographers use colored filters for black and
> white, in order to change the response to light.  A red filter,
> for example,
> darkens the sky considerably.

Yes, but so what?  IT ISN'T PART OF THE DISCUSSION.  This is a separate
issue, and in no way mitigates what I've said.

> > You know, if this just doesn't work, then how
> > come it DOES in fact work?
>
> It _doesn't_ work.

OK, if you say so ;-)

> > I see B&W images made from color ALL the time,
> > and they look VERY close to "same" B&W images
> > shot with B&W film...
>
> Very close is not the same.  And sometimes it's not even very close.

Well, if you use amorphous terms that go unqualified, then of course.  What
I claim is very close, you will claim is not.  How about you quantify "very
close" for us?

> > So, how is it that thousands of people already
> > do what you claim impossible?
>
> They don't.

OK, if you say so.  Then I'll have to let them know.  I'll tell them you
said what they are doing can't be done, and that they should just stop doing
it.

> > I can manipulate the color image tonality such that
> > it VERY closely matches the tonality from the scanned
> > B&W image.
>
> Show me.  Use Tri-X as the black and white film.  Also, do the same thing
> with a narrowband yellow filter over the B&W exposure.

Why?  That isn't part of the discussion.  The discussion is NOT about
filters.  You are only making it about filters because my premise IS in fact
correct, and you know it, and this is your way of trying to mitigate what
I've said.

Austin

Austin writes:

> What, precisely, does B&W film "sense"?

Light falling upon it.  Light in a wide range of frequencies will cause the
film to react, but the magnitude of the reaction per unit exposure varies by
frequency in a largely random way.  B&W film essentially integrates the area
under an irregular curve, and produces a single number as a result.

> Both the film, and the CCD, sense the number of
> photons that "hit" it.  They are spectral independent.

No, they are not.  Photons of certain energies produce a reaction more
readily than others.

> Look at the film response curve for B&W film.  It's reasonably
> flat, up to the point of fall-off.

It looks irregularly curved to me.  It can be dramatically changed by
filters, too.

> Now, does the RGB data not have this "intensity"
> information?

RGB data, like B&W data, is an integration of the area under one or more
curves.  The curves themselves are lost in the process.

> Does it matter that different combinations of RGB
> produce the same intensity?

What matters is the spectral distribution in the original scene.

I'll try again, although I'm not optimistic that anyone who hasn't already
understood will understand from this:

Set up two omnidirectional microphones in the middle of an orchestra, and
record some music.  At the same time, set up two other microphones, in two
completely different positions, and record the music with those, too.

Now take the recording from one set of microphones, and transform it into a
recording of the music as heard from the other two microphones.  You may be
surprised to discover that it cannot be done.  For example, an instrument
exactly equidistant from the first two microphones may be in one of two
positions, but that instrument will not be equidistant from the other two
microphones.  Therefore, the instrument will sound just as loud in both
channels of the first pair of mics, whereas it will sound unequally loud
across the two channels of the second pair.  Explain how you can transform a
recording from the first pair so as to accurately reproduce the imbalance in
the channels that was heard by the second pair.  It can't be done.

In fact, even if you use two mics for the first recording, and only one mic
for the second (like RGB to B&W), it still can't be done.

Austin writes:

> That's not relevant to the discussion.

Of _course_ it is relevant.  The whole purpose of this list is to discuss
black and white printing.  Black and white prints normally come from black
and white images, and the latter are often produced with black and white
film and filters.  It's about as relevant as anything can get.

This entire discussion, in fact, is far more important than you seem to
realize, because it has some pretty serious implications for those trying to
reproduce one type of B&W print from something other than one type of B&W
image.  In particular, it is critically important that anyone shooting
color--such as digital color--and trying to get B&W from it understand these
principles, so that he or she can understand why certain results can be
produced, whereas others cannot.

> Hum.  I see it done all the time.

No, you see approximations.

Anthony,

> > No, actually, you haven't...  I know you believe
> > you have, but you're missing the point.
>
> Explain to me why this statement would be true when you utter it, but not
> true when I utter it.

Since you've conveniently clipped what this statement is related to, and I'm
not going to waste my time looking it up.

> > Name them, please ...
>
> I already have.  Spectral yellow is rendered identically to a certain
> combination of spectral red and green.
>
> > ... and then after you do, show that this is significant.
>
> If it were not significant, no color display or printing system
> that exists
> today would work.

Again, you have failed to show the significance of this as it relates to
B&W.  I don't care about color printing systems, we are not talking about
color printing systems.

> > You are aware that B&W film gives the same tonality
> > for many different colors?
>
> Yes.  And it also gives different tones for many different color as well.

Er, right...but so what?  It's intensity based, not tonal based.

> And often the response of the film is essentially random across the
> spectrum.

So if I take the same picture 10 times, with the same film, exposure, camera
etc. I get random tonality?

> > Again, show that that is significant.
>
> I have, probably at least a dozen times by now.  If you didn't
> understand it
> the first eleven times, why would you understand it now?

Well, it's because claiming something is significant doesn't SHOW it is, it
merely SAYS it is, and fact is, you have not shown the significance.  You
can't, that's why you don't.  You continue to bring in filters, which have
absolutely no bearing what so ever on the topic.  And the only reason I can
guess is because you know your challenge to my original premise is now
incorrect, so you are trying to find a way to misdirect that.

Austin

Truman writes:

> The only way you could not reconstruct the
> intensity of white light would be if there
> was a frequency that fell outside the passband
> of all three of the RGB filters.

True, as long as the light is truly white, that is, as long as the intensity
of the light at each frequency in the spectrum is the same (the spectral
distribution is a flat line).  In this case, you can easily reproduce the
distribution in order to transform RGB into anything else.

If the light is not absolutely white, the rules change, and most
transformations are not possible.  If you can somehow reconstruct the
curve--e.g., if you know it has a certain shape--you can do the
transformation, but since there is typically no way to know this, there
isn't much you can do.

In real world scenes, of course, the light is never white.  The only time
all the light is white is if you take a picture of a white sheet of paper
illuminated by a perfect white light source, or the equivalent.

> > That's not relevant to the discussion.
>
> Of _course_ it is relevant.  The whole purpose of this list is to discuss
> black and white printing.  Black and white prints normally come from black
> and white images, and the latter are often produced with black and white
> film and filters.  It's about as relevant as anything can get.

No, Anthony.  The premise what that you can turn scanned color film data
into a visually indistinguishable image (tonal wise) from that of the same
image taken with Tri-X and scanned, period.  It was not about the use of
filters.  That is another discussion.  FEW people actually use filters
compared to the number that don't.  And as I said, it is NOT relevant to my
premise, and the reason you want to make it relevant is because you somehow
believe it mitigates the correctness of my statement, and it does not.

> This entire discussion, in fact, is far more important than you seem to
> realize,

Being that I've been in the digital imaging business for over 25 years, any
discussion on digital imaging is very important to me.

> because it has some pretty serious implications for
> those trying to
> reproduce one type of B&W print from something other than one type of B&W
> image.

Yes, and that's why it's done all the time, whether you concur or not.

> In particular, it is critically important that anyone shooting
> color--such as digital color--and trying to get B&W from it
> understand these
> principles, so that he or she can understand why certain results can be
> produced, whereas others cannot.

That's very true...but my premise is still correct.

> > Hum.  I see it done all the time.
>
> No, you see approximations.

Everything is an approximation.

Austin

Austin writes:

> Well, it's because claiming something is significant
> doesn't SHOW it is, it merely SAYS it is, and fact is,
> you have not shown the significance.

But you are doing exactly the same thing, and yet you expect me to believe
you, even as you refuse to believe me.  Why is that?

> And the only reason I can guess is because you
> know your challenge to my original premise is now
> incorrect, so you are trying to find a way to
> misdirect that.

Show me an example of a transformation that works, and make me a liar.

Austin writes:

> Being that I've been in the digital imaging
> business for over 25 years, any discussion on
> digital imaging is very important to me.

That's why it mystifies me that you don't seem to understant this.  It
_must_ have come up many times in the past, I should think.

> Yes, and that's why it's done all the time,
> whether you concur or not.

Show me an example.  Show me an algorithm or table or anything that will
transform an RGB image from Provia, or from a digital camera, into a B&W
image that exactly duplicates Tri-X, for example.  I want to see the
miracle.

> Everything is an approximation.

No.  Some transformations are exact.  For example, an inversion of an image
to its negative form is lossless and exact; every positive image has one and
only one negative image, and vice versa.

If I was in a bar and a few guys were arguing about this stuff while most of
the other people were trying to watch something on the tv, I would
respectfully ask them to take it outside.

How about it?

Anthony,

> No, they are not.  Photons of certain energies produce a reaction more
> readily than others.

Look at the B&W photoresponse curve.  Along the ENTIRE spectrum of the film,
they produce nearly similar reaction.  This is what's important, as the B&W
film is NOT color sensitive because of this response.

> > Look at the film response curve for B&W film.  It's reasonably
> > flat, up to the point of fall-off.
>
> It looks irregularly curved to me.  It can be dramatically changed by
> filters, too.

Well, it's not irregular enough to be significant.  Yes, it can be
dramatically changed by a lense cap as well.  They both have the same
relevance to this discussion.

> > Now, does the RGB data not have this "intensity"
> > information?
>
> RGB data, like B&W data, is an integration of the area under one or more
> curves.  The curves themselves are lost in the process.

And, that's not relevant.  If the intensity information is maintained, then
you've got all you need.

> > Does it matter that different combinations of RGB
> > produce the same intensity?
>
> What matters is the spectral distribution in the original scene.

Not at all.  What matters is the INTENSITY of the original scene.

> I'll try again, although I'm not optimistic that anyone who hasn't already
> understood will understand from this:
>
> Set up two omnidirectional microphones in the middle of an orchestra, and
> record some music.  At the same time, set up two other microphones, in two
> completely different positions, and record the music with those, too.

Yeah, they will be very different in amplitude, and it also depends on how
they are aimed, but no matter what, since the distances between the sound
and the microphones are different, which changes the amplitude, the results
will be different.

> Now take the recording from one set of microphones, and transform
> it into a
> recording of the music as heard from the other two microphones.

I don't believe it can be done, since there are complex frequencies
involved...which is an entirely different case from converting RGB color
data into B&W data.  Audio is NOT point sourced, images are.  Audio data is
also not static, it requires surrounding data to evaluate a single point.
Image data does not.

> You may be
> surprised to discover that it cannot be done.

Not surprised at all, being that I designed a lot of digital imaging
equipment, and consulted to a lot of top sound engineers (both live and
studio) who used our equipment.

> In fact, even if you use two mics for the first recording, and
> only one mic
> for the second (like RGB to B&W), it still can't be done.

And once again, your analogy is entirely irrelevant.  Image data is static,
audio data is not.  Image data is point sourced, audio data is not.

Austin

Anthony,

> > Being that I've been in the digital imaging
> > business for over 25 years, any discussion on
> > digital imaging is very important to me.
>
> That's why it mystifies me that you don't seem to understand this.

The other option is that I'm correct, and you are wrong.  Strongly consider
that.

> It
> _must_ have come up many times in the past, I should think.

It has come up many times, and as I've said, my premise is correct.

> > Yes, and that's why it's done all the time,
> > whether you concur or not.
>
> Show me an example.  Show me an algorithm or table or anything that will
> transform an RGB image from Provia, or from a digital camera, into a B&W
> image that exactly duplicates Tri-X, for example.  I want to see the
> miracle.

I suggest you do a web search.

> > Everything is an approximation.
>
> No.  Some transformations are exact.  For example, an inversion
> of an image
> to its negative form is lossless and exact; every positive image
> has one and
> only one negative image, and vice versa.

The original data is not exact, and that is what the premise is based on, so
the transformation is also an approximation of the original scene.

Austin

My two cents- it's really not critically important that I understand the
principles in creating b&w images from digital color in order to create art.

Capp




[capuozzo]

 From: Austin Franklin [mailto:darkroom@...]
Sent: Wednesday, May 28, 2003 2:32 PM
To: DigitalBlackandWhiteThePrint@yahoogroups.com
Subject: RE: [Digital BW] Digital, film, scanning comparisons




  > > That's not relevant to the discussion.
  >
  > Of _course_ it is relevant.  The whole purpose of this list is to
discuss
  > black and white printing.  Black and white prints normally come from
black
  > and white images, and the latter are often produced with black and white
  > film and filters.  It's about as relevant as anything can get.

  No, Anthony.  The premise what that you can turn scanned color film data
  into a visually indistinguishable image (tonal wise) from that of the same
  image taken with Tri-X and scanned, period.  It was not about the use of
  filters.  That is another discussion.  FEW people actually use filters
  compared to the number that don't.  And as I said, it is NOT relevant to
my
  premise, and the reason you want to make it relevant is because you
somehow
  believe it mitigates the correctness of my statement, and it does not.

  > This entire discussion, in fact, is far more important than you seem to
  > realize,

  Being that I've been in the digital imaging business for over 25 years,
any
  discussion on digital imaging is very important to me.

  > because it has some pretty serious implications for
  > those trying to
  > reproduce one type of B&W print from something other than one type of
B&W
  > image.

  Yes, and that's why it's done all the time, whether you concur or not.

  > In particular, it is critically important that anyone shooting
  > color--such as digital color--and trying to get B&W from it
  > understand these
  > principles, so that he or she can understand why certain results can be
  > produced, whereas others cannot.

  That's very true...but my premise is still correct.

  > > Hum.  I see it done all the time.
  >
  > No, you see approximations.

  Everything is an approximation.

  Austin


        Yahoo! Groups Sponsor





  Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:

  http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint

  If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.

  Please follow these basic guidelines:
  - Include your full name with your message.
  - Include the address of your website, if you have one.
  - As threads develop, trim off excess portions of earlier messages to keep
them short.
  - As the topic of a thread changes remember to change the subject header.
  - Good manners are required at all time. No personal attacks or flames
  - Complete your Yahoo profile.
  - Before posting a question, search the message archives and the various
resources on the homepage.




  Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.


[Non-text portions of this message have been removed]

Anthony,

> > Well, it's because claiming something is significant
> > doesn't SHOW it is, it merely SAYS it is, and fact is,
> > you have not shown the significance.
>
> But you are doing exactly the same thing, and yet you expect me to believe
> you, even as you refuse to believe me.  Why is that?

Because YOU are the one who claims it can't be done, and it's up to YOU to
show an example of why, not up to me to show an example that it can.

> > And the only reason I can guess is because you
> > know your challenge to my original premise is now
> > incorrect, so you are trying to find a way to
> > misdirect that.
>
> Show me an example of a transformation that works, and make me a liar.

I don't believe you're a liar at all.  I've never said that.  I just think
you think you know more than you actually do.

I have real work to do now...so you're on your own.

Austin

Austin writes:

> Yeah, they will be very different in amplitude,
> and it also depends on how they are aimed ...

They are omnidirectional; they cannot be aimed.

> I don't believe it can be done, since there are
> complex frequencies involved...which is an entirely
> different case from converting RGB color data
> into B&W data.

It's the same thing.  But as I said, I wasn't optimistic that yet another
analogy would work where its predecessors had not.

Stephen writes:

> How about it?

You're not in a bar, and you have a Delete key on your computer.

Austin writes:

> The other option is that I'm correct, and you are
> wrong.  Strongly consider that.

Why?  You've provided even less support for your position than I have.

> I suggest you do a web search.

No, I want to see you provide an example.  You say you see it all the time,
it should be trivial.  How about a program or something that does it?  Plug
in three RGB values, and out comes the grayscale value.  It must work for
any type of light in the original scene.  It must duplicate the grayscale
image obtained by photographing the scene directly with Tri-X.

Austin writes:

> Because YOU are the one who claims it can't be done ...

You are the one who claims it can.  What's the difference?

> ... and it's up to YOU to show an example of why,
> not up to me to show an example that it can.

Why?  They are both positive assertions that carry a burden of proof.  Why
ask others to assume this burden if you will not?

> I don't believe you're a liar at all.

Well, then, show me an example of such a transformation, and don't make me a
liar.  Just show me the example.

Capp writes:

> My two cents- it's really not critically important
> that I understand the principles in creating b&w
> images from digital color in order to create art.

Perhaps for you, but some artists are extremely finicky about getting things
just so; and for them, understanding why certain things cannot be done is
important to sparing them a lot of future grief and frustration.

On 5/28/03 11:55 AM, "Anthony Atkielski" <anthony@...> wrote:

> Austin writes:
> 
>> The other option is that I'm correct, and you are
>> wrong.  Strongly consider that.
> 
> Why?  You've provided even less support for your position than I have.
> 
>> I suggest you do a web search.
> 
> No, I want to see you provide an example.  You say you see it all the time,
> it should be trivial.  How about a program or something that does it?  Plug
> in three RGB values, and out comes the grayscale value.  It must work for
> any type of light in the original scene.  It must duplicate the grayscale
> image obtained by photographing the scene directly with Tri-X.

Anthony,

All you need to do is search the archives of this list.  We've discussed
this many times before.  There are several companies that make photoshop
plugins to convert RBG images into various type of BW film.  Try
silveroxide.com for one. They work very well.  Of course what they do not
simulate is the difference in grain structure...and this would be much more
difficult...but most people that are interested in doing the RGB to BW
conversion are interested in doing it either for Digital capture or for
transparency film images that are relatively free of grain...so they
probably don't care about this.

Robert

on 5/28/03 2:53 PM, Anthony Atkielski at anthony@... wrote:

> 
> You're not in a bar, and you have a Delete key on your computer.

And I would love to use it a lot less than I've had to lately.  Anyone else
agree?

Anthony,

> > Because YOU are the one who claims it can't be done ...
>
> You are the one who claims it can.  What's the difference?

You were the first to make a claim.  You can't do that without expecting to
be challenged.

> > I don't believe you're a liar at all.
>
> Well, then, show me an example of such a transformation, and
> don't make me a
> liar.  Just show me the example.

It won't make you a liar, it will simply make you wrong.

Austin

Anthony,

> > Yeah, they will be very different in amplitude,
> > and it also depends on how they are aimed ...
>
> They are omnidirectional; they cannot be aimed.

How high above the floor?  Omni directional does not mean they don't have
"aim".  It means their aim is very large.

> > I don't believe it can be done, since there are
> > complex frequencies involved...which is an entirely
> > different case from converting RGB color data
> > into B&W data.
>
> It's the same thing.

YOU believe it's the same thing, but in fact, it is not.

> But as I said, I wasn't optimistic that yet another
> analogy would work where its predecessors had not.

No analogies will work, because you are arguing against a premise that is
simply correct.

Austin

> If I was in a bar and a few guys were arguing about this stuff
> while most of
> the other people were trying to watch something on the tv, I would
> respectfully ask them to take it outside.
>
> How about it?

I agree.  Anthony's claim has been challenged, and I believe until he can
demonstrate that his claim has merit in REALITY, it simply doesn't, and
discussing it any more is futile.

Austin

I know what you mean- on the other hand, one great thing about email lists
is the *plonk* filter- one of these guys was already on mine, others are
newly added!

Bill

on 5/28/03 2:48 PM, Stephen Petegorsky wrote:

> If I was in a bar and a few guys were arguing about this stuff while most of
> the other people were trying to watch something on the tv, I would
> respectfully ask them to take it outside.
> 
> How about it?
> 
> 
> 
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and other
> resources as they are often being updated. The page is at:
> 
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
> 
> If you wish to receive no emails or just a daily digest, or you wish to
> unsubscribe, please edit your Membership preferences by visiting this same
> page.
> 
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
> them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
> resources on the homepage.
> 
> 
> 
> 
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
> 
>

me too. Anthony was kicked from the sp list for this kind of trolling... the
guy hasn't tried printing or scannign BW and has opinions about everything.
ignore him - he'll go away

Julian
----- Original Message ----- 
From: "Stephen Petegorsky" <petegorsky@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 28, 2003 9:09 PM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> on 5/28/03 2:53 PM, Anthony Atkielski at anthony@... wrote:
>
> >
> > You're not in a bar, and you have a Delete key on your computer.
>
> And I would love to use it a lot less than I've had to lately.  Anyone
else
> agree?
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

Not a chance. Seems like it belongs on the Techie and Geek" list.

Seth

=-----Original Message-----
=From: sceptre12345 [mailto:am1000@...] 
=Sent: Wednesday, 28 May, 2003 08:32
=To: DigitalBlackandWhiteThePrint@yahoogroups.com
=Subject: Re: [Digital BW] Digital, film, Spectral Sensitivity
=
=
=--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Julian Thomas" 
=<julianthomas@t...> wrote:
=> Peter, I'm with you - this whole debate by Anthony is a troll.
=Arguing my
=> analogy quickly descends into arguing whether the analogy is valid.
=Until
=> Anthony buys a scanner and an inkjet printer, all this is crazy!
=> 
=> Julian
=
=Anyone get anything usefull out of this tread or am I missing 
=something ?
=Cheers,
=Andre
=
=
=------------------------ Yahoo! Groups Sponsor 
=---------------------~--> Get A Free Psychic Reading! Your 
=Online Answer To Life's Important Questions. 
=http://us.click.yahoo.com/Lj3uPC/Me7FAA/CNxFAA/=ucIolB/TM
=
=---------------------------------------------------------------
=------~->
=
=Please visit the Group Homepage to check the Files, Bookmarks, 
=Polls and other resources as they are often being updated. The 
=page is at:
=
=http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
=
=If you wish to receive no emails or just a daily digest, or 
=you wish to unsubscribe, please edit your Membership 
=preferences by visiting this same page.
=
=Please follow these basic guidelines:
=- Include your full name with your message.
=- Include the address of your website, if you have one.
=- As threads develop, trim off excess portions of earlier 
=messages to keep them short.
=- As the topic of a thread changes remember to change the 
=subject header.
=- Good manners are required at all time. No personal attacks or flames
=- Complete your Yahoo profile.
=- Before posting a question, search the message archives and 
=the various resources on the homepage. 
=
=
= 
=
=Your use of Yahoo! Groups is subject to 
=http://docs.yahoo.com/info/terms/ 
=
=
=

> Anyone get anything useful out of this tread or am I missing something ?

Actually, I did.  I found out that the spectral response of Portra 400 is
wider than Tri-X.

Austin

It's a good topic, but the negatives (due to handling) now outweigh the upside.
I agree with you, Stephen.
John

  ----- Original Message ----- 
  From: Stephen Petegorsky 
  To: DigitalBlackandWhiteThePrint@yahoogroups.com 
  Sent: Wednesday, May 28, 2003 3:09 PM
  Subject: Re: [Digital BW] Digital, film, scanning comparisons


  on 5/28/03 2:53 PM, Anthony Atkielski at anthony@... wrote:

  > 
  > You're not in a bar, and you have a Delete key on your computer.

  And I would love to use it a lot less than I've had to lately.  Anyone else
  agree?


        Yahoo! Groups Sponsor 
              ADVERTISEMENT
             
       
       

  Please visit the Group Homepage to check the Files, Bookmarks, Polls and other resources as they are often being updated. The page is at:

  http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint

  If you wish to receive no emails or just a daily digest, or you wish to unsubscribe, please edit your Membership preferences by visiting this same page.

  Please follow these basic guidelines:
  - Include your full name with your message.
  - Include the address of your website, if you have one.
  - As threads develop, trim off excess portions of earlier messages to keep them short.
  - As the topic of a thread changes remember to change the subject header.
  - Good manners are required at all time. No personal attacks or flames
  - Complete your Yahoo profile.
  - Before posting a question, search the message archives and the various resources on the homepage. 




  Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service. 



[Non-text portions of this message have been removed]

Count me in as an aye vote. Is it over yet? I was away for a couple of days
and came back to this. I guess the Delete is better than sticking tooth
picks in your eyes. Pretty close.

bgs
----- Original Message -----
From: "Stephen Petegorsky" <petegorsky@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 28, 2003 3:09 PM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> on 5/28/03 2:53 PM, Anthony Atkielski at anthony@... wrote:
>
> >
> > You're not in a bar, and you have a Delete key on your computer.
>
> And I would love to use it a lot less than I've had to lately.  Anyone
else
> agree?
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

> I guess the Delete is better than sticking tooth
> picks in your eyes. Pretty close.
>
> bgs

Gosh, pressing that button some 50 times...oh the horror!  You sound like
George Jetson...

;-)

Did you guys like, go to college or something?

----- Original Message -----
From: "Jon Dubovsky" <entropy@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 28, 2003 8:47 AM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> >>It is a basic mathematical fact that the process of binning is
>  >>*not* reversible under almost all circumstances.
>  >
>  > But technically this isn't "binning", because true binning is
>  > mutually exclusive.   It's either in this bin or that one.  In this
>  > case you have OVERLAPPING bins, and not only that but the sensitivity
>  > curve in each bin is different.   So the same pixel has a signal in
>  > bin A and a signal in bin B.  There is only one unique point on the
>  > spectrum that would produce any give signal in both bins.
>
> Mathematically, it is the same.  Whether the response functions overlap or
> not, except for trivial cases (again, such as an impulse function for the
> response function), the impossibility of conversion still holds.
>
> --
> Jon Dubovsky ( entropy@... )
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>

Pressing a button 50 times at my age is called vigorous exercise!
----- Original Message -----
From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 28, 2003 8:58 PM
Subject: RE: [Digital BW] Digital, film, scanning comparisons


>
> > I guess the Delete is better than sticking tooth
> > picks in your eyes. Pretty close.
> >
> > bgs
>
> Gosh, pressing that button some 50 times...oh the horror!  You sound like
> George Jetson...
>
> ;-)
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

Try a mass selection - in outlook click on from to reorder by sender, roll
to first post hold shift key, scroll down to last still holding shift key ,
delete and its done.

-----Original Message-----
From: bgs [mailto:bgs@...]
Sent: Thursday, May 29, 2003 10:13 AM
To: DigitalBlackandWhiteThePrint@yahoogroups.com
Subject: Re: [Digital BW] Digital, film, scanning comparisons


Pressing a button 50 times at my age is called vigorous exercise!
----- Original Message -----
From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, May 28, 2003 8:58 PM
Subject: RE: [Digital BW] Digital, film, scanning comparisons


>
> > I guess the Delete is better than sticking tooth
> > picks in your eyes. Pretty close.
> >
> > bgs
>
> Gosh, pressing that button some 50 times...oh the horror!  You sound like
> George Jetson...
>
> ;-)
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or flames
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>



Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:

http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint

If you wish to receive no emails or just a daily digest, or you wish to
unsubscribe, please edit your Membership preferences by visiting this same
page.

Please follow these basic guidelines:
- Include your full name with your message.
- Include the address of your website, if you have one.
- As threads develop, trim off excess portions of earlier messages to keep
them short.
- As the topic of a thread changes remember to change the subject header.
- Good manners are required at all time. No personal attacks or flames
- Complete your Yahoo profile.
- Before posting a question, search the message archives and the various
resources on the homepage.




Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

----- Original Message ----- 

From: "A. Huntley" <leicam6@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Monday, May 26, 2003 7:34 PM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> Martin writes:
>
> > Alan,
> >
> > I remember Morley Baer letting me look through his old 8x10 and it was a
> > real eye opener in comparison to the puny 4x5 I was working with. I
never
> > went to 8x10 due to the enlarger barrier but now with digital the
> economics
> > shift and 8x10 lends itself to less expensive scanning solutions than
4x5.
> >
> > Martin
>
> I'm a little jealous and envious now...I absolutely love Morley's work and
> have at least one video where he is interviewed. He seemed like an
> unpretentious man...one of those rare photographers that is both GREAT and
> approachable. I would have loved to do a workshop with him or, at least,
> have met him.

Alan,

I was fortunate enough to attend one of his last workshops. Four days in San
Juan Bautista to photograph the mission and town with Frank Talbot as
co-instructor and Gordon Hutchings along for the fun of it. Lots of fond
memories. He also did a "post vivo" of the workshop at his house and we got
the chance to show are prints. A very nice man and absolutely dedicated to
his personal vision. His book "Light Years" is a must have those who love
B&W landscape photography.

He later let me come down and took a look at my portfolio. I planned to go
back the following year but kept putting it off thinking I needed to get one
more print done and he passed away before I made it back.
>
> Did you know that Brett Weston referred to the 4x5 as his miniature format
> camera? I know that he used MF during his later years, but quite a few of
us
> seem to migrate downward in format as we age. I'm trying to talk my son
into
> lugging the 8x10, but he's not biting yet...smart boy!

I have a much younger brother that I have conned into playing camera bearer
for me a couple of times.<G>
>
> Once I started using the 8x10 I pretty much gave up on 4x5 for about 10
> years. To my eye, B&W contact prints are simply stunning and the 8x10 is
> just big enough to see the image well. Though, my late friend Fred Picker
> did some really lovely 4x5 contacts of snow/ice/water details of brooks in
> Vermont. However, as you said with digital the economics of using 8x10 is
> looking really good. The cost of film is way up there now--I used to buy
50
> sheet boxes of Tri-X for about $50US--but, processing can be done with a
few
> cheap plastic trays, and probably just about any flatbed that can handle
> this size film would be fine. My Epson 1680 does an acceptable job. Most
of
> the B&W quads I have produced from 8x10 have been roughly A4 size. I guess
> the only real downside to the economics involved is that you'll want a LF
> printer! <g> Epson 7000 anyone?
>
I know what you mean about the 8x10 contacts. Morley's were really superb.
My own favorite print size was 11x14 for many years and moving up to that
format was too much to undertake. The sheer weight of camera and film
holders pretty much ties you to a car. I couldn't shoot 8x10 either since
the cost of an 8x10 enlarger was impossible. Even 5x7 was prohibitive
because of the enlarger expense. From scanning my 4x5 negs on the Howtek I
am really amazed at the amount of detail on the film. I am pretty convinced
you would need to print it out at 30x40 to see all of it in the print. I bet
a 4000 dpi scan of an 8x10 neg would be really something!

With digital printing in mind I think the real advantage of the 8x10 may
simple be that it is much, much easier to compose on a bright 8x10 ground
glass than the wimpy little 4x5 glass. My current thought, that I have not
acted on, is to move up from 4x5 to 5x7. Two sheets should easily wet mount
on the Howtek and the increase in camera size and weight would not be as
great as going up to 8x10. B&H is charging about $70 for 50 sheets of TriX,
about half of the 8x10 price. I like the proportions better too.

In all honesty though I have done most of my work in the last few years with
the Pentax 67.

Martin Wesley

Martin writes:

> I was fortunate enough to attend one of his last workshops. Four days in
San
> Juan Bautista to photograph the mission and town with Frank Talbot as
> co-instructor and Gordon Hutchings along for the fun of it. Lots of fond
> memories. He also did a "post vivo" of the workshop at his house and we
got
> the chance to show are prints. A very nice man and absolutely dedicated to
> his personal vision. His book "Light Years" is a must have those who love
> B&W landscape photography.

I had planned to do a workshop with Morley but, as you said, time has a way
of getting away from us and before ya know it. I have "Light Years" and it
remains one of my favorite photography books. I never tire of looking
through it!

> I know what you mean about the 8x10 contacts. Morley's were really superb.
> My own favorite print size was 11x14 for many years and moving up to that
> format was too much to undertake. The sheer weight of camera and film
> holders pretty much ties you to a car. I couldn't shoot 8x10 either since
> the cost of an 8x10 enlarger was impossible. Even 5x7 was prohibitive
> because of the enlarger expense. From scanning my 4x5 negs on the Howtek I
> am really amazed at the amount of detail on the film. I am pretty
convinced
> you would need to print it out at 30x40 to see all of it in the print. I
bet
> a 4000 dpi scan of an 8x10 neg would be really something!

I almost moved to 11x14 after seeing first-hand some of Brett Weston's ULF
(Ultra Large Format) contact prints and when RH Philips was making their
"flyweight" camera in this size. But, even with the lighter weight of the
Philips camera you're still looking at quite a mass of stuff to move about!
I toyed briefly with enlarging my 8x10 negs; I bought an Aristo 12x12
coldlight, fashioned a homemade neg holder that fit on the back of my
camera, built a base for holding everything together, and used my 300mm
G-Claron lens as an enlarging lens (ala AA.) I bought a wall mount metal
"easel" with magnets from Wisner. It all worked pretty well though perfect
alignment was nearly impossible! However, these large film formats provide
quite a bit of "fudge factor" for these things. I almost laugh now because
when Zone VI introduced their 8x10 enlarger I considered it...$3K to $4K
depending on the enlarging lens chosen. Considering what I've spent on
digital, now, that would have been an absolute bargin! <G> I bet the Howtek
scans are really nice. I've only ever used my Epson Expression 1680 for LF
because film scanners for this size are way outta reach for me; though deals
can probably be found on eBay. I can't imagine the sheer processing power
and memory requirements to handle a 4000 dpi 8x10 scan. My 1600 dpi 16-bit
gray scans are large enough!

> With digital printing in mind I think the real advantage of the 8x10 may
> simple be that it is much, much easier to compose on a bright 8x10 ground
> glass than the wimpy little 4x5 glass. My current thought, that I have not
> acted on, is to move up from 4x5 to 5x7. Two sheets should easily wet
mount
> on the Howtek and the increase in camera size and weight would not be as
> great as going up to 8x10. B&H is charging about $70 for 50 sheets of
TriX,
> about half of the 8x10 price. I like the proportions better too.

The 8x10 groundglass is like a TV screen! After using it for so many years I
can hardly see my 4x5! <G> I think you might like the 5x7. I, personally,
seem to be moving away from the squattier frames presented by 4x5, 8x10,
etc. Seems more and more I prefer the elongated formats of 35mm, 6x9, etc.
I've often thought of the 5x7 myself because the weight increase over 4x5 is
negligible and the surface area of the film is so much greater...somewhat
like comparing 6x4.5 to 6x7. 5x7 contacts are jewel-like and any higher end
flatbed scanner would handle these nicely. One of these days I'm going to
FTP a couple 8x10 image files to West Coast Imaging and see what a large
Piezo print looks like.

> In all honesty though I have done most of my work in the last few years
with
> the Pentax 67.

I've got a nice Pentax 67 system, too. Great camera. However, I always seem
to come back to weight vs. quality. If I'm going to lug the Pentax outfit, I
may as well take the 4x5. Besides...my Toho 4x5 is about half the weight of
the 67 body alone!

Regards,
Alan Huntley

Alan writes:

> I can't imagine the sheer processing power
> and memory requirements to handle a 4000 dpi
> 8x10 scan. My 1600 dpi 16-bit gray scans are
> large enough!

You need 3 GB just to load a 4000-dpi scan of 8x10, in B&W.  In color, you
need 8 GB.  To do anything besides load the image, you'd need 20 GB of RAM
or so.

Unfortunately, many computers are limited to far less memory than that.
Typically they accept only a few memory modules, and even with the largest
modules available, it may be impossible to get beyond 1.5-3 GB.  Worse yet,
currently available Windows systems do not support more than a few gigabytes
of RAM (the 32-bit processor architecture makes support of more than 4 GB
problematic, and most versions of Windows limit applications to half this
amount).

So it will be a while before affordable computers exist to manipulate
high-resolution scans of LF film.  It sure will be nice when it happens,
though.  Finding a way to scan the film is also a problem, and scanners,
unlike computers, do not drop in price rapidly, so even if a desktop LF
scanner becomes available, it may be unaffordable for a long time.

Note that 4000-dpi scans from 8x10 would allow prints literally the size of
a barn.  Make those Duratrans and light them from behind, and the whole
neighborhood can admire your work!

----- Original Message ----- 

From: "A. Huntley" <leicam6@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, June 05, 2003 5:47 PM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


>
> Martin writes:
>
(snip earlier)
>
> I almost moved to 11x14 after seeing first-hand some of Brett Weston's ULF
> (Ultra Large Format) contact prints and when RH Philips was making their
> "flyweight" camera in this size. But, even with the lighter weight of the
> Philips camera you're still looking at quite a mass of stuff to move
about!
> I toyed briefly with enlarging my 8x10 negs; I bought an Aristo 12x12
> coldlight, fashioned a homemade neg holder that fit on the back of my
> camera, built a base for holding everything together, and used my 300mm
> G-Claron lens as an enlarging lens (ala AA.) I bought a wall mount metal
> "easel" with magnets from Wisner. It all worked pretty well though perfect
> alignment was nearly impossible! However, these large film formats provide
> quite a bit of "fudge factor" for these things.

That is an incredible effort!

> I almost laugh now because
> when Zone VI introduced their 8x10 enlarger I considered it...$3K to $4K
> depending on the enlarging lens chosen. Considering what I've spent on
> digital, now, that would have been an absolute bargin! <G>

I know! I have spent way more on digital than I ever considered reasonable
for my wet darkroom.

> I bet the Howtek
> scans are really nice. I've only ever used my Epson Expression 1680 for LF
> because film scanners for this size are way outta reach for me; though
deals
> can probably be found on eBay. I can't imagine the sheer processing power
> and memory requirements to handle a 4000 dpi 8x10 scan. My 1600 dpi 16-bit
> gray scans are large enough!

My 16-bit grayscale, 4000 dpi scans of 4x5 negs run 600+ MB. Some 16-bit RGB
scans top 1.8GB. So 8x10 would hit 2.4GB and 7.2GB! Actually even 600MB is
too much to work with comfortably. I did some test printing and printed the
same file at 13x19 from the original 4000 dpi scan and then again from the
same file down sampled to 2400 dpi. I could not see any difference in a
print this size at all so I now just down sample right away. Even so by the
time I have an 8-bit grayscale PS file with layers the size creeps back up
to 300 to 400 MB.

If I was scanning from 8x10 I suspect I would have to scan at 2000 dpi to
keep file sizes down to something I could even open. Further downsampling
would probably be necessary to have a managable working file. At 1600 dpi
you probably are at the limit current computers and PS can comfortably
swallow. I think that you would still be amazed at the difference in quality
of a 1600 dpi scan from a drum scanner vs. the flatbed. The cost is probably
along the lines of the Zone VI enlarger. <G>
>
> > With digital printing in mind I think the real advantage of the 8x10 may
> > simple be that it is much, much easier to compose on a bright 8x10
ground
> > glass than the wimpy little 4x5 glass. My current thought, that I have
not
> > acted on, is to move up from 4x5 to 5x7. Two sheets should easily wet
> mount
> > on the Howtek and the increase in camera size and weight would not be as
> > great as going up to 8x10. B&H is charging about $70 for 50 sheets of
> TriX,
> > about half of the 8x10 price. I like the proportions better too.
>
> The 8x10 groundglass is like a TV screen! After using it for so many years
I
> can hardly see my 4x5! <G> I think you might like the 5x7. I, personally,
> seem to be moving away from the squattier frames presented by 4x5, 8x10,
> etc. Seems more and more I prefer the elongated formats of 35mm, 6x9, etc.
> I've often thought of the 5x7 myself because the weight increase over 4x5
is
> negligible and the surface area of the film is so much greater...somewhat
> like comparing 6x4.5 to 6x7. 5x7 contacts are jewel-like and any higher
end
> flatbed scanner would handle these nicely. One of these days I'm going to
> FTP a couple 8x10 image files to West Coast Imaging and see what a large
> Piezo print looks like.

When you do, give us some feedback on the experience. I would be very
interested.
>
> > In all honesty though I have done most of my work in the last few years
> with
> > the Pentax 67.
>
> I've got a nice Pentax 67 system, too. Great camera. However, I always
seem
> to come back to weight vs. quality. If I'm going to lug the Pentax outfit,
I
> may as well take the 4x5. Besides...my Toho 4x5 is about half the weight
of
> the 67 body alone!

I agree! Thing weighs a ton but when I bought my system in the early 90's it
was an incredible bargain in that format. If it wasn't for the cost I would
trade for the Mamiya at this point. The 67 just offers speed and the
possibilty of hand held shots. It works out better when I am traveling with
people who don't have the patience to wait while you do a view camera shot.

Martin

----- Original Message ----- 

From: "Anthony Atkielski" <anthony@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Friday, June 06, 2003 2:13 AM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


> Alan writes:
>
> > I can't imagine the sheer processing power
> > and memory requirements to handle a 4000 dpi
> > 8x10 scan. My 1600 dpi 16-bit gray scans are
> > large enough!
>
> You need 3 GB just to load a 4000-dpi scan of 8x10, in B&W.  In color, you
> need 8 GB.  To do anything besides load the image, you'd need 20 GB of RAM
> or so.
>
> Unfortunately, many computers are limited to far less memory than that.
> Typically they accept only a few memory modules, and even with the largest
> modules available, it may be impossible to get beyond 1.5-3 GB.  Worse
yet,
> currently available Windows systems do not support more than a few
gigabytes
> of RAM (the 32-bit processor architecture makes support of more than 4 GB
> problematic, and most versions of Windows limit applications to half this
> amount).

Anthony,

I too am disappointed at the memory limitations on current systems and lack
of OS support. In this case though, Photoshop is still the limiting factor.
I am running 2GB of RAM under Win 2000 and Photoshop will initially open say
a 1.2GB file entirely in RAM but as soon as you try to do any adjustments or
manipulations it starts spooling to the swap file and only uses a fraction
of the available memory. Adobe's insistence on using a separate swap file in
addition to the Windows page file results in sub par memory performance. As
things stand now there does seem to be much benefit to exceeding 1.5 to
2.0GB of RAM. Hopefully this will change.

Martin Wesley

(snip earlier)

Martin writes:

> In this case though, Photoshop is still the limiting
> factor.

Photoshop still has the brain-dead memory management that it had under the
old Mac OS (out of necessity, back then).  Adobe is unwilling to fix it, and
it can become a serious problem with large images.

> I am running 2GB of RAM under Win 2000 and Photoshop
> will initially open say a 1.2GB file entirely in RAM
> but as soon as you try to do any adjustments or
> manipulations it starts spooling to the swap file
> and only uses a fraction of the available memory.

I have encountered a persistent bug in PS that causes it to give up all
memory when you try to save a large TIFF as a JPEG.  It goes from 1 GB down
to 46 MB or so, then gives up.  You have to close and reopen PS to do the
save.

> Adobe's insistence on using a separate swap file
> in addition to the Windows page file results in sub
> par memory performance.

Agreed, although Adobe refuses to recognize this.

On Friday, Jun 6, 2003, at 20:09 America/New_York, Martin Wesley wrote:

> Adobe's insistence on using a separate swap file in
> addition to the Windows page file results in sub par memory 
> performance.

I would guess that the main reason it was done this way was because 
prior to Mac OSX, the mac didn't have paging memory so Adobe needed 
their own memory management system. In order to keep the code base 
easier to maintain they just ported the whole system to Windows. Now 
that PS 8 will be strictly OSX from what I hear, maybe their swap file 
won't be needed (or at least they can change the way it works).

Jim

Martin writes:

> My 16-bit grayscale, 4000 dpi scans of 4x5 negs run 600+ MB. Some 16-bit
RGB
> scans top 1.8GB. So 8x10 would hit 2.4GB and 7.2GB! Actually even 600MB is
> too much to work with comfortably. I did some test printing and printed
the
> same file at 13x19 from the original 4000 dpi scan and then again from the
> same file down sampled to 2400 dpi. I could not see any difference in a
> print this size at all so I now just down sample right away. Even so by
the
> time I have an 8-bit grayscale PS file with layers the size creeps back up
> to 300 to 400 MB.
>
> If I was scanning from 8x10 I suspect I would have to scan at 2000 dpi to
> keep file sizes down to something I could even open. Further downsampling
> would probably be necessary to have a managable working file. At 1600 dpi
> you probably are at the limit current computers and PS can comfortably
> swallow. I think that you would still be amazed at the difference in
quality
> of a 1600 dpi scan from a drum scanner vs. the flatbed. The cost is
probably
> along the lines of the Zone VI enlarger. <G>

My 16-bit 4x5 grayscale scans on the Epson 1680 flatbed usually top out at
about 100MB. My 450Mhz G4 Mac seems to handle this size without a hiccup. I
usually scan my 8x10 B&W negs at 800 ppi figuring that with this size neg I
don't need anything more than that for my typical print size; generally
about 11x14, or so, on 13x19 paper. However, for best quality I should
probably be scanning at max optical density (1600) of my scanner. No doubt
that any drum scan would produce better quality than my flatbed, but the
Epson 1680 seems to do pretty well for LF work. I'm actually toying with the
idea of getting some Kami fluid, or equivalent, and trying a quasi wet-mount
directly on the glass. I've read some things about the 'net of people using
mineral oil, etc., and reporting higher resolution scans with better shadow
detail. All good things! But, then there's the cleanup. <G>

> When you do, give us some feedback on the experience. I would be very
> interested.

Will do.

> I agree! Thing weighs a ton but when I bought my system in the early 90's
it
> was an incredible bargain in that format. If it wasn't for the cost I
would
> trade for the Mamiya at this point. The 67 just offers speed and the
> possibilty of hand held shots. It works out better when I am traveling
with
> people who don't have the patience to wait while you do a view camera
shot.

Got mine in the early 90's, too. At the time, I ordered only the body,
metering prism (actually I wanted the standard prism because I always use a
spot meter - habit from all those years with LF, but they only had the prism
with the meter), and a 105/2.4 lens. Upon receipt, I tested the Pentax
optics fairly extensively against Zeiss on a Hasselblad. Though the Zeiss
glass was marginally better, the Pentax optic was no slouch! Certainly no
difference that would be discernable by even the most critical eye at my
standard print sizes. Maybe, if I was printing to billboard size? Sounds
like we got our 67s for the same reason...it's a little nerve wracking to
have the family standing around tapping their feet while you(I) set up a LF
camera and take a shot! <G> That's why, for LF work, I work alone. To this
day, I still enjoy the slow, methodical nature of working with those large
cameras.

Regards,
Alan Huntley

Martin Wesley

http://www.borderless-photos.de/guests.html

----- Original Message ----- 
From: "A. Huntley" <leicam6@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Saturday, June 07, 2003 9:16 AM
Subject: Re: [Digital BW] Digital, film, scanning comparisons


(snip)
>
> My 16-bit 4x5 grayscale scans on the Epson 1680 flatbed usually top out at
> about 100MB. My 450Mhz G4 Mac seems to handle this size without a hiccup.
I
> usually scan my 8x10 B&W negs at 800 ppi figuring that with this size neg
I
> don't need anything more than that for my typical print size; generally
> about 11x14, or so, on 13x19 paper. However, for best quality I should
> probably be scanning at max optical density (1600) of my scanner. No doubt
> that any drum scan would produce better quality than my flatbed, but the
> Epson 1680 seems to do pretty well for LF work. I'm actually toying with
the
> idea of getting some Kami fluid, or equivalent, and trying a quasi
wet-mount
> directly on the glass. I've read some things about the 'net of people
using
> mineral oil, etc., and reporting higher resolution scans with better
shadow
> detail. All good things! But, then there's the cleanup. <G>

Alan,

George DeWolfe wrote a nice article with examples on using wet mounting with
a 1640. It was in View Camera or Camera Arts about 18 months ago. It really
does improve things. I would recommend you try something like Kami fluid or
Prazio Anti-Newton oil which are very volatile and easy to clean up. Need to
seal the edges of the glass plate on your scanner really well with mounting
tape to keep the oil out of the works. I think you would see a big
improvement.

Martin

(Snip)

Martin writes,

> Alan,
>
> George DeWolfe wrote a nice article with examples on using wet mounting
with
> a 1640. It was in View Camera or Camera Arts about 18 months ago. It
really
> does improve things. I would recommend you try something like Kami fluid
or
> Prazio Anti-Newton oil which are very volatile and easy to clean up. Need
to
> seal the edges of the glass plate on your scanner really well with
mounting
> tape to keep the oil out of the works. I think you would see a big
> improvement.

Thank you! For the life of me I couldn't remember George's name when I typed
my previous message. Yes, I am aware of his article and have it. I believe
he uses mineral oil which I just couldn't bring myself to spread all over my
LF negs so I never proceeded any further with this idea. Seems like mineral
oil would be quite a mess to cleanup.

It finally occurred to me one day--I'm a little slower than I should be with
these things! ;>)--that Kami fluid is used throughout the industry for drum
scanning, etc., so maybe that would work better. Based on a recent message
from one of our list members I visited www.aztek.com yesterday and it
appears that they have a "kit" consisting of Kami fluid, neg cleaning
solution, etc., specifically for use on a flatbed scanner no less! I plan to
contact Aztek this week. Didn't know about the Prazio oil...thanks. I will
seal the edges of the glass before putting any fluid on it, but thanks for
reminding me.

Alan Huntley