Digital BW, The Print

Allesandro-

Interesting.


> That's what I used to think and do, too, but very recently I changed my
> mind.
> These days I'm scanning Portra400BW 6x7 with an Epson3200 using 
> Vuescan, and
> my very simple workflow was to get a raw scan, extract the green 
> channel and
> Photoshop away, setting BP and WP etc.
> I was a bit concerned, though, on the huge amount of stretching I had 
> to do
> setting BP/WP in order to bring the usually *very* flat scan to cover 
> the
> whole 0-255 range, but given that I was starting from the raw scan,

The raw data is all crammed up against one end of the histogram.  
Somewhere along the workflow is needs to be 'stretched' by setting the 
white and black points.  You can do it in photoshop or the scanner 
software can do it.

I have trouble visually setting the WP and BP accurately on a raw scan 
because it looks so compressed.  I use vuescan to do that initially for 
me.  By setting the clip points to 0 vuescan sets the darkest part of 
the negative in preview frame as the white point and the clearest part 
of the negative in the preview frame as the BP.

> I
> thought I had no control over it. I was wrong: I found that the 
> automatic
> exposure set by Vuescan is, at least with this film-scanner 
> combination,
> usually way too short: talking Vuescan numbers, the automatic setting 
> is
> just below 4, the optimal setting ranges from 6 to 11 (you can 
> verify/change
> this number by checking the Lock exposure checkbox at the bottom of the
> Device tab: the value is set after the preview).
> I used to think that increasing exposure would only shift the histogram
> curve from the light to the dark end, but it's not true. Assuming that 
> the
> response is linear (I'll stand corrected on this: Austin, your 
> knowledged
> opinion is?), if exposure x gives a 20-100 range, exposure 2x gives 
> 40-200,
> thus resulting in a less flat scan. The result is a smoother image 
> (nothing
> outstanding, but noticeable), especially after steep curves and levels 
> (e.g.
> to bring dull skies to life)

I wonder if  the exposure actually increased by or if it is a software 
compensation.  Do you know if the Epson 3200 can change the intensity 
of the light source or the duration of the time the CCD is capturing 
info?  From memory the RGB exposure is a multiplier not like the analog 
gain in some of the Nikon scanners that actually change the exposure 
time.

I can't figure out why changing the actual exposure would do anything 
other than move the range of values left or right in the histogram.  If 
the RGB exposure in vuescan is just a multiplier then what you describe 
makes sense.

> Bottom line, my workflow now includes a little extra step at the 
> begininng,
> which is one or more 100dpi scan, checking the resulting histogram in
> Vuescan, with a locked exposure, until I get a reasonable range.


--
Kevin Gulstene
http://www.dockwalker.com