Yahoo Groups archive

Datacolor User to User Support Group.

Index last updated: 2026-04-28 23:18 UTC

Thread

Sudden loss of accuracy of Datacolor #1005 after 2 years

Sudden loss of accuracy of Datacolor #1005 after 2 years

2009-02-11 by nadar1900

I have used the spectrocolorimeter datacolor #1005 of my PrintFixPro
package for 2 years, profiling about 40 different papers for color and
grayscale with good results. But, just in the last days, measurements
of dark patches seem quite strange on the color coordinates in Lab
values. For example, here 5 Lab measurements of the step 96 in a 51
steps grayscale, waiting about 5-10 second between them:
1) 7,99 | -2,91 | 1,26
2) 8,25 | -1,29 | -0,9
3) 8,11 | -0,85 | -1,1
4) 8,29 | -0,22 | -2,13
5) 8,00 | -1,47 | 1,28

B value range from 1,28 to -2,13 and I can notice an evident loss of
contiguity, for a and b values for dark patches, comparing same paper
profiles created in the last days and one month ago.
I have repeated the same tests with different sw upgrades, from
PrintFix 1.0 to Spyder3Print, on different computers and different usb
ports, but nothing to do.
Have you any suggestions?

Re: [colorvision_group] Sudden loss of accuracy of Datacolor #1005 after 2 years

2009-02-11 by C D Tobie

On Feb 11, 2009, at 5:08 AM, nadar1900 wrote:

> I have used the spectrocolorimeter datacolor #1005 of my PrintFixPro
> package for 2 years, profiling about 40 different papers for color and
> grayscale with good results. But, just in the last days, measurements
> of dark patches seem quite strange on the color coordinates in Lab
> values. For example, here 5 Lab measurements of the step 96 in a 51
> steps grayscale, waiting about 5-10 second between them:
> 1) 7,99 | -2,91 | 1,26
> 2) 8,25 | -1,29 | -0,9
> 3) 8,11 | -0,85 | -1,1
> 4) 8,29 | -0,22 | -2,13
> 5) 8,00 | -1,47 | 1,28
>
> B value range from 1,28 to -2,13 and I can notice an evident loss of
> contiguity, for a and b values for dark patches, comparing same paper
> profiles created in the last days and one month ago.
> I have repeated the same tests with different sw upgrades, from
> PrintFix 1.0 to Spyder3Print, on different computers and different usb
> ports, but nothing to do.
> Have you any suggestions?


As I stated in your support ticket, and in other lists where you have  
posted this, if the variance is beyond what your usage requires, then  
you will have to purchase another spectro, since your current unit is  
beyond warranty. Sorry!

C. David Tobie
Global Product Technology Manager
Digital Imaging & Home Theater
CDTobie@...

Re: Sudden loss of accuracy of Datacolor #1005 after 2 years

2009-02-11 by nadar1900

It sounds so bad. So, this tool can be used for two years or for
producing 40 profiles only. However thanks for the response.
Bye
Daniel




--- In colorvision_group@yahoogroups.com, C D Tobie <CDTobie@...> wrote:
Show quoted textHide quoted text
> 
> As I stated in your support ticket, and in other lists where you have  
> posted this, if the variance is beyond what your usage requires, then  
> you will have to purchase another spectro, since your current unit is  
> beyond warranty. Sorry!
> 
> C. David Tobie
> Global Product Technology Manager
> Digital Imaging & Home Theater
> CDTobie@...
> 
> 
> 
> 
> Datacolor
> www.datacolor.com/Spyder3
>

Re: [colorvision_group] Re: Sudden loss of accuracy of Datacolor #1005 after 2 years

2009-02-11 by C D Tobie

On Feb 11, 2009, at 10:20 AM, nadar1900 wrote:

> It sounds so bad. So, this tool can be used for two years or for
> producing 40 profiles only. However thanks for the response.

Yours is actually the first, out of thousands, that I have heard of  
doing this. But its well beyond warranty, so if its a hardware issue,  
then thats the end of the story. After all, the way such devices are  
built affordably is by making them fixed units. There is no mechanism  
for repair. I've been carrying one around in my briefcase for years,  
and it still works as expected. So please do not assume your  
individual case is universal. Its also worth noting that your unit may  
still do fine for building color profiles, you are using for a  
different process.

C. David Tobie
Global Product Technology Manager
Digital Imaging & Home Theater
CDTobie@...

Re: Sudden loss of accuracy of Datacolor #1005 after 2 years

2009-02-11 by nadar1900

This is evident, I have been unlucky with Datacolor, then I will try
another brand.


--- In colorvision_group@yahoogroups.com, C D Tobie <CDTobie@...> wrote:
Show quoted textHide quoted text
>
> 
> On Feb 11, 2009, at 10:20 AM, nadar1900 wrote:
> 
> > It sounds so bad. So, this tool can be used for two years or for
> > producing 40 profiles only. However thanks for the response.
> 
> Yours is actually the first, out of thousands, that I have heard of  
> doing this. But its well beyond warranty, so if its a hardware issue,  
> then thats the end of the story. After all, the way such devices are  
> built affordably is by making them fixed units. There is no mechanism  
> for repair. I've been carrying one around in my briefcase for years,  
> and it still works as expected. So please do not assume your  
> individual case is universal. Its also worth noting that your unit may  
> still do fine for building color profiles, you are using for a  
> different process.
> 
> C. David Tobie
> Global Product Technology Manager
> Digital Imaging & Home Theater
> CDTobie@...
> 
> 
> 
> 
> Datacolor
> www.datacolor.com/Spyder3
>

RE:Sudden loss of accuracy of Datacolor #1005 after 2 years

2009-02-12 by Daniel Vézina

Posted by: "nadar1900" nadar1900@... nadar1900
Wed Feb 11, 2009 11:24 am (PST)

>This is evident, I have been unlucky with Datacolor, then I will try
>another brand.

Don\u2019t jump in the air too rapidly... You\u2019re trying to read a very dark patch right?

Your readings are:

> 1) 7,99 | -2,91 | 1,26
> 2) 8,25 | -1,29 | -0,9
> 3) 8,11 | -0,85 | -1,1
> 4) 8,29 | -0,22 | -2,13
> 5) 8,00 | -1,47 | 1,28

You said that you\u2019re waiting 5-10 seconds between each measurement.

First, when L reach 8 and below, depending on the texture/finish of the media you\u2019re trying to profile, it can be difficult for ANY spectro to give a correct reading. Beware of it. Others more expensive spectro could also show sign of weakness in this area.

Spectro are often a bit less accurate for the first reading. When the sensor is \u201ccold\u201d. So usually, you should start to read your patches sequence this way:

Your read the first 2 patches, then promptly reposition the spectro and the software on the first patch and start to read continuously all patches with less than 5 sec delay between patch.

In your case the delay between the readings is too long. Just for the fun, use the Tools/Measure function, put the spectro on the patch and then read it two time per sec and see if the values swing a lot or not.You should get very constant values.

Now, you will say that it is not written nowhere to proceed this way. Datacolor never said that the first reading is less accurate. Why?

  1. Because they\u2019re probably not alone in this case.
  2. Because this variation is so small that it will change nothing to the resulting profile. The printed picture will look identical to the eyes.

Now, if you have some sensitive work to do, you can always proceed like I said. If you want to read a specific patch/object with the Spectro, then simply read the patch rapidly, discard the first 2 values, average the 3 next ones. (which should be near).

If the Spectro canR17;t provide consistent values when used like cited above, then go to buy a new one, or a different one as you wish. But don\u2019t expect that a so cheap device (all brands) will last forever and will be as consistent than a far more expensive device costing thousands of $. But my own experience is that the datacolor unit is as good as a far more expensive spectro.

Also others Spectro come with a different software, good luck to find another software in the price range of S3Print that will give you profiles as nice as what you can build with it. Again, more expensive software don217;t necessary gives you a better profile.

I run a Fine Art Reproduction business and S3Print is my workhorse to make my profiles. I still have to find a competitor in my area that can match my repro quality and color accuracy. Of course the result of a repro job is not just related to a printer profile. But the printer\u2019s profiles must be good otherwise it will not works

Good luck!

--
Daniel Vézina
Directeur Technique
Les productions Numart Inc.
PQ, Canada
email: numart@...

www.numart.ca
La technologie numérique au service de l\u2019Art!


RE: [colorvision_group] RE:Sudden loss of accuracy of Datacolor #1005 after 2 years

2009-02-12 by PJS

GREAT information! Thank you!

pjs
"Light makes photography. Embrace light. Admire it.
Love it. But above all, know light.
Know it for all you are worth,
and you will know the key to photography."
George Eastman

-----Original Message-----
From: Daniel V é zina
Cc: nadar1900@...
Subject: [colorvision_group] RE:Sudden loss of accuracy of Datacolor #1005 after 2 years

Don’t jump in the air too rapidly... You’re trying to read a very dark patch right?

Your readings are:

> 1) 7,99 | -2,91 | 1,26
> 2) 8,25 | -1,29 | -0,9
> 3) 8,11 | -0,85 | -1,1
> 4) 8,29 | -0,22 | -2,13
> 5) 8,00 | -1,47 | 1,28

You said that you’re waiting 5-10 seconds between each measurement.

First, when L reach 8 and below, depending on the texture/finish of the media you’re trying to profile, it can be difficult for ANY spectro to give a correct reading. Beware of it. Others more expensive spectro could also show sign of weakness in this area.

Spectro are often a bit less accurate for the first reading. When the sensor is “cold”. So usually, you should start to read your patches sequence this way:

Your read the first 2 patches, then promptly reposition the spectro and the software on the first patch and start to read continuously all patches with less than 5 sec delay between patch.

In your case the delay between the readings is too long. Just for the fun, use the Tools/Measure function, put the spectro on the patch and then read it two time per sec and see if the values swing a lot or not.You should get very constant values.

Now, you will say that it is not written nowhere to proceed this way. Datacolor never said that the first reading is less accurate. Why?
  1. Because they’re probably not alone in this case.
  2. Because this variation is so small that it will change nothing to the resulting profile. The printed picture will look identical to the eyes.

Now, if you have some sensitive work to do, you can always proceed like I said. If you want to read a specific patch/object with the Spectro, then simply read the patch rapidly, discard the first 2 values, average the 3 next ones. (which should be near).

If the Spectro can’t provide consistent values when used like cited above, then go to buy a new one, or a different one as you wish. But don’t expect that a so cheap device (all brands) will last forever and will be as consistent than a far more expensive device costing thousands of $. But my own experience is that the datacolor unit is as good as a far more expensive spectro.

Also others Spectro come with a different software, good luck to find another software in the price range of S3Print that will give you profiles as nice as what you can build with it. Again, more expensive software don’t necessary gives you a better profile.

I run a Fine Art Reproduction business and S3Print is my workhorse to make my profiles. I still have to find a competitor in my area that can match my repro quality and color accuracy. Of course the result of a repro job is not just related to a printer profile. But the printer’s profiles must be good otherwise it will not works

Good luck!

--
Daniel Vézina
Directeur Technique
Les productions Numart Inc.
PQ, Canada
email: numart@videotron.ca

www.numart.ca
La technologie numérique au service de l’Art!

Re: RE:Sudden loss of accuracy of Datacolor #1005 after 2 years

2009-02-12 by nadar1900

Yes Daniel, of course. You are right about the procedure, but I canshow you, with a practical example, the differences concerning agrayscale profile on the same paper made a month ago and another made afew days ago. You can notice how dark patches have substantially randomvalues for color coordinates, in particular "b" values. The contiguityof "a" and "b" values in the dark side of the curve is completely lost,the standard deviation is changed... and my procedure to creategrayscale or ICC profiles is always the same. Usually, how many timesyou read a patch making a profile? I read the patches in sequence, thenif I want, I repeat reading and calculate the average. Well, now I haveto do the average of 4 or 5 reading in a grayscale to get an acceptablecurve. Something is happened, do not you think so?
Thanks for reply.
Daniel

(sorry for proportional characters, I have adjusted the alignment for a better graphical view)

OLD PROFILE
Step Dens Lab A B
0.00 0.024 97.84 1.26 -2.09 - b ; | a L +
2.00 0.071 93.86 1.74 -2.70 - b | a ; L +
4.00 0.115 90.19 ; 1.19 -2.49 - ; b | a ; L +
6.00 0.153 87.16 1.33 -2.25 - b | a L +
8.00 0.191 84.17 0.91 -2.00 - b | a ; L +
10.00 0.235 80.82 1.79 -2.36 - b | a L +
12.00 0.273 78.09 2.22 -2.77 - b | a L +
14.00 0.315 75.07 2.36 -3.22 - ; b | a L +
16.00 0.359 72.03 2.92 -3.56 - b ; | a L ; +
18.00 0.401 69.28 2.30 -3.17 - b | a L +
20.00 ; 0.456 65.77 1.94 -2.79 - b | a L ; +
22.00 0.502 62.90 1.85 -2.74 - b | a +
24.00 0.547 60.24 1.31 -2.31 - b | aL +
26.00 0.596 57.41 1.02 -2.04 - ; b | a +
28.00 0.645 54.68 0.78 -1.70 - b | La +
30.00 0.684 52.61 0.89 -1.57 - b |L a +
32.00 0.733 ; 50.10 1.08 -1.80 - b L a +
34.00 0.780 47.73 1.30 -2.50 - b L | a +
36.00 0.819 45.85 1.01 -2.17 - ; b L | a +
38.00 0.880 43.02 0.66 -2.82 - b L | a +
40.00 0.940 40.38 0.31 -2.96 - b L |a +
42.00 0.992 38.17 0.23 -3.26 - b L a +
44.00 1.041 36.18 0.35 -3.52 - b L |a +
46.00 1.102 33.80 0.15 -3.09 - b L a ; +
48.00 1.149 32.02 0.53 -2.15 - ; L b | a +
50.00 1.195 30.36 0.49 -1.88 - L b |a +
52.00 1.239 ; 28.82 0.75 -1.52 - L b | ; a +
54.00 1.269 27.80 0.78 -1.57 - L b | a ; +
56.00 1.311 26.40 0.71 -1.37 - ; L b | a +
58.00 1.354 25.04 0.96 -1.51 - L b | a +
60.00 1.390 23.90 0.85 -1.30 - L ; b | a ; +
62.00 1.420 23.00 0.92 -1.33 - L b | a +
64.00 ; 1.457 21.91 0.92 -1.28 - L b | a +
66.00 1.490 20.96 1.11 -0.62 ; - L b | a ; +
68.00 1.522 20.07 0.78 -0.46 - ; L b | a +
70.00 1.554 19.20 0.83 -0.20 - L b| a ; +
72.00 1.591 18.22 0.79 -0.12 - ; L ; b| a +
74.00 1.634 17.10 0.48 -0.04 - L b|a +
76.00 1.682 15.89 0.60 -0.30 - L b | a +
78.00 1.720 14.98 0.53 -0.57 - L ; b | a ; +
80.00 1.754 14.19 0.11 -0.84 - L b a +
82.00 1.791 ; 13.33 0.11 -0.78 - L b a +
84.00 1.830 12.47 0.22 -1.00 - L b a ; ; +
86.00 1.861 11.81 -0.18 -1.10 - L ; b a| +
88.00 1.885 11.29 -0.22 -1.23 - L b a| +
90.00 1.929 10.40 -1.07 -1.03 - L ; b | +
92.00 ; 1.976 9.45 -1.34 -1.24 - L ab | +
94.00 2.005 8.90 -1.17 -0.81 - L ab | +
96.00 2.054 7.98 -1.24 -0.29 - L a b | +
98.00 2.080 7.51 -0.89 -0.24 - L a b| ; +
100.00 2.083 ; 7.46 -0.35 0.19 - L a b ; +

NEW PROFILE
Step Dens Lab A B
0.00 0.016 98.58 1.37 -1.94 - b | a L+
2.00 0.066 94.30 1.72 -2.63 ; - b | a L +
4.00 0.110 90.57 1.18 -2.51 - ; ; b | a ; L +
6.00 0.148 87.58 ; 1.18 -2.07 - b | a L +
8.00 0.186 84.60 0.80 -1.86 - b | a L +
10.00 ; 0.231 81.15 1.79 -2.34 - b | a L +
12.00 0.267 78.47 2.23 -2.66 ; - ; b ; | a L ; +
14.00 0.310 75.43 2.24 -3.02 - ; b | ; a L +
16.00 0.355 72.33 2.93 -3.50 - b | a L +
18.00 0.396 69.59 2.14 -3.09 - b | a L +
20.00 0.449 66.17 1.92 -2.86 - b | a L +
22.00 0.495 63.36 1.70 -2.58 - ; b | a L +
24.00 0.541 60.59 1.26 ; -2.26 - b ; | aL +
26.00 0.590 57.78 1.05 -1.95 - b | a +
28.00 0.639 55.01 0.68 -1.77 - ; b | aL +
30.00 0.677 ; 52.99 0.80 -1.63 - b |L a +
32.00 0.726 50.46 0.89 -1.78 - ; b L a ; +
34.00 0.772 48.16 1.20 -2.45 - ; b L | a +
36.00 0.808 46.38 0.71 ; -1.99 - b L | a +
38.00 0.872 43.40 0.59 -2.90 - b ; L | a ; +
40.00 0.931 40.77 0.20 -2.88 - b L a ; +
42.00 0.979 38.72 -0.24 -2.79 - b L a| +
44.00 1.029 36.66 0.20 -3.44 - b L a +
46.00 1.089 34.30 -0.24 -2.56 - bL a| ; +
48.00 1.139 32.39 -0.25 -1.53 - L b a| +
50.00 1.184 30.75 0.31 -2.03 ; - L b ; |a ; +
52.00 1.230 29.14 -0.10 -0.63 - L b a| +
54.00 1.258 28.16 0.98 -1.80 - L b | a +
56.00 ; 1.301 26.73 0.61 -1.44 - L ; b | a ; +
58.00 1.342 25.40 0.79 -1.29 - L b | a +
60.00 1.380 24.21 0.79 -0.78 - L b | a +
62.00 1.410 23.32 ; 0.41 -0.59 - L b |a +
64.00 1.448 22.17 0.05 -0.20 - L ba +
66.00 1.483 21.17 0.49 0.08 - L ba +
68.00 1.512 20.36 0.76 0.09 - L b a +
70.00 1.549 19.34 0.46 ; 0.49 - L ; |b ; +
72.00 1.570 18.76 0.71 -0.46 - ; L ; b | a ; +
74.00 1.625 17.32 0.65 0.15 - L b a +
76.00 1.679 15.98 1.18 -1.34 - L b | a ; +
78.00 1.706 15.31 0.20 0.10 - L ; b +
80.00 1.751 14.25 -0.56 0.26 - L a |b ; +
82.00 1.773 13.74 0.86 -1.49 - L ; b | a ; +
84.00 1.813 12.84 1.05 -1.05 - L b | a ; +
86.00 1.851 12.01 -0.14 -1.50 - L ; b a| +
88.00 1.893 11.13 -1.88 0.52 - L a | b +
90.00 1.910 10.78 -0.98 -1.14 - L ba | +
92.00 1.959 ; 9.80 -0.11 -1.64 - L b a| +
94.00 1.982 9.33 0.45 -2.39 - L b |a ; +
96.00 2.053 8.00 -2.54 2.35 - L a | b +
98.00 2.052 8.02 0.61 -1.75 - L b | a +
100.00 2.053 7.99 0.05 -0.40 ; - L b a +

The Lab values are not in order.
Cannot make a profile.




--- In colorvision_group@yahoogroups.com, Daniel Vézina wrote:
>
> Posted by: "nadar1900" nadar1900@... nadar1900
> Wed Feb 11, 2009 11:24 am (PST)
>
> >This is evident, I have been unlucky with Datacolor, then I will try
> >another brand.
>
> Don¹t jump in the air too rapidly... You¹re trying to read a very dark patch
> right?
>
> Your readings are:
>
> > 1) 7,99 | -2,91 | 1,26
> > 2) 8,25 | -1,29 | -0,9
> > 3) 8,11 | -0,85 | -1,1
> > 4) 8,29 | -0,22 | -2,13
> > 5) 8,00 | -1,47 | 1,28
>
> You said that you¹re waiting 5-10 seconds betwee (Message over 64 KB, truncated)

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.