Pages: [1]   Go Down

Author Topic: Different software producing different spectral data from the same target?  (Read 311 times)

Kyle D Jackson

  • Full Member
  • ***
  • Offline Offline
  • Posts: 103
    • Lone Leaf Photography

Hi all, normally I'd post in Printers forum but maybe this is more a CM / spectro question.

(Without detailing the month-deep rabbit hole I'm down...)

When using *different software* to collect print scan data from spectros from the *same printed target*, should the spectral data in their measurement files basically be the same?

So one printer, one target. Scan with different software packages to get the measurement data file (CGATS .txt) from each. Shouldn't the spectral data be very similar, within normal scan/device variation error? (dE 1-ish)

Or do different software packages apply some kind of "flavouring" to the spectral data? Which would imply that their CGATS files are only compatible with their own profile generation software and can't be used in anyone else's software.

Because on 1728 patches I'm seeing avg dE 4+, max dE 12+... (using Argyll colverify tool to compare them)


Thanks
Logged
Kyle D Jackson
Ottawa, Canada
Lone Leaf Photography

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20646
  • Andrew Rodney
    • http://www.digitaldog.net/

The scanned values (presumably ending up as Lab values in a CGATS) are the values, so no, the packages don't matter. The color patches defined and then measured need to match, other than that, the numbers are the numbers. What are you trying to accomplish?
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Kyle D Jackson

  • Full Member
  • ***
  • Offline Offline
  • Posts: 103
    • Lone Leaf Photography


Hi Andrew, thanks for reply. Not different LAB values, I mean upstream of those, actual different spectral readings in each nm-band (I've no idea what units they use in the file, whatever spectral density is measured in I guess).

Anyhow good to know that they should be the same. Sort of realized that after, I mean the whole point of being able to take spot measurements would be to get an accurate measurement regardless of which software was used with the device. Should be no reason for any particular software to mess with the spectral measurements.

Your question would involve diving into that month-deep rabbit hole that I doubt anyone wants to... In brief it started as testing the built-in spectros in 2 different HP Z-series printers to assess whether one of them was failing or not. Unfortunately because of software differences between the 2 printer models, I ended up at this point. In a thread over in Printing I described how I haven't been able to get them to use identical targets, one insists on arranging the patches slightly differently, so I'm stuck for the moment with scanning 2 different printed targets instead of one. It *shouldn't* be a problem -- same colour patches, same paper, same printer -- however it's beginning to smell a lot like what you've alluded to: the 2 printed targets don't have the same colours.  I mean I've checked it 8 ways from Sunday and Argyll colverify still gives the dE's above. 4-12 is massive so I should be able to spot the differences if that's the case. The investigation continues tomorrow...

Thanks again, cheers
Logged
Kyle D Jackson
Ottawa, Canada
Lone Leaf Photography

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20646
  • Andrew Rodney
    • http://www.digitaldog.net/

Two different makes or models of spectros measuring the same patch could slightly vary depending on a number of factors (light source as one example) but the dE should not be that high.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Kyle D Jackson

  • Full Member
  • ***
  • Offline Offline
  • Posts: 103
    • Lone Leaf Photography


Agreed. And in this case those 4-12 dE are from the *same* spectro, so it definitely shouldn't be happening.

With fresh eyes on the 2 target sheets this morning, I definitely see different ink densities, most noticeable in the darkest patches. So now I just need to investigate why the 2 software apparently used different paper presets when they made the targets (probably something I did without realizing it!). I suspect that will cure the anomaly.


Side question: does anyone know if Argyll colverify has an option to output the dE values to a file so I can look for which specific patches have the largest error? (I know profcheck does this but I'm not actually making profiles, this is still scan data only.) Colverify calculates the dE values but it only shows a summary on-screen (avg, max, worst 10%, a histogram, etc), but not the patch-by-patch dE values. Trying to avoid calculating all the dE values myself... Here's what I'm running:

   colverify -v -h -s -w -x FILE1(TARGET).ti3 FILE2(MEASURED).ti3
(https://www.argyllcms.com/doc/colverify.html)

That -s option says it sorts the patches by dE value, yet I'm not getting any such display. Maybe I just need to pipe the output to a file of my choosing, with " > OUTFILE.txt " similar to using profcheck. But I thought that only writes what's already echoed to screen.

Thanks
Logged
Kyle D Jackson
Ottawa, Canada
Lone Leaf Photography

GWGill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 608
  • Author of ArgyllCMS & ArgyllPRO ColorMeter
    • ArgyllCMS

That -s option says it sorts the patches by dE value, yet I'm not getting any such display.
Maybe try -v2 instead of -v ?
Logged

Kyle D Jackson

  • Full Member
  • ***
  • Offline Offline
  • Posts: 103
    • Lone Leaf Photography


Ah! Thanks so much Graeme!  :) I totally didn't understand the meaning of "-v [n]  Verbose mode, n >= 2 print each value", but now I think I get it. I'll give it a try!

Thanks
Logged
Kyle D Jackson
Ottawa, Canada
Lone Leaf Photography
Pages: [1]   Go Up