Pages: [1] 2   Go Down

Author Topic: Can't match two monitors  (Read 1358 times)

David Mantripp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 822
    • :: snowhenge dot net ::
Can't match two monitors
« on: November 26, 2018, 01:54:14 pm »

I'm going to ask a question here which is probably stupid and has probably been answered a million times before, but anyway...

I'm trying, and totally failing, to colour match two monitors attached to a Mac Pro trashcan.  One is an Eizo CS270, about 2 years old, the other is a Quato 240LE, about 8 years old.  I'm using an i1 Display Pro, ColorNavigator on the Eizo, i1Profiler on the Quato. i1Profiler cannot access the Quato's hardware control so I have to set RGB & Brightness manually.  I'm calibrating to 6500K, 100cd, using L* on the Eizo and default Gamma on the Quato.

Both calibrate without error and give good results. But the calibrated displays, showing Apple's Solid Gray Dark, are completely different.  If I look at either on their own, they seem ok.  But together, the Eizo is much more contrasty. On the Quato there seems to be a green-ish tinge, but the grey is anyway more like what I expect a Mid-Gray to look like.   The problem is I have no idea which is correct.  I suppose in both cases the monitors think they are displaying neutral grey, but clearly either one or both is not.

My expectation is that calibration should being me into the ballpark.  I don'r expect an exact match, but neither do I expect a glaring difference.  Up until a few years ago I also used an ageing 1st generation Apple Cinema Display, and even that I could get close enough to the Quato to not bother me.

Are my expectations excessive ?  Any suggestions on how I can troubleshoot ?

Thanks!
Logged
--
David Mantripp

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15490
    • http://www.digitaldog.net/
Re: Can't match two monitors
« Reply #1 on: November 26, 2018, 02:10:54 pm »

You expectations that two dissimilar displays should match with the same target calibration is the problem.
As for contrast, you need control over contrast ratio which you may or may not have on both:
http://blog.xritephoto.com/2011/07/x-rite-i1display-pro-advanced-features-contrast-ratio-with-coloratti-andrew-rodney/
You're also crippling the full capabilities of the Eizo which I think is a bad idea.
Logged
Andrew Rodney
Author “Color Management for Photographers"

David Mantripp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 822
    • :: snowhenge dot net ::
Re: Can't match two monitors
« Reply #2 on: November 26, 2018, 04:07:56 pm »

but....  ok, so I use one, or the other.  They're very noticeably different, but both my profiles are "good".  So, if I then make a print using either, say using ImagePrint, to remove one variable, which display will ImagePrint match?   The Quato doesn't even let me adjust contrast using the HUD.  Not sure why...  Unfortunately Quato are defunct, and their hardware calibration software, which is built on Java, is no longer supported.

And sorry, but if just for now, I plead "tldr", do you mean I should not use ColorNavigator for the Eizo ?

I do appreciate your reply, Andrew, the time you take to help people through these mazes is really a gift.
Logged
--
David Mantripp

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15490
    • http://www.digitaldog.net/
Re: Can't match two monitors
« Reply #3 on: November 26, 2018, 04:10:24 pm »

Both profiles are good but reflect differing calibration, calibration that doesn't produce a match. You'll have to futz around and set each product to result in as close a match as possible. But don't expect the same settings for cd/m^2, white point, contrast ratio etc to produce a visual match just because the numbers used match.
Logged
Andrew Rodney
Author “Color Management for Photographers"

David Mantripp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 822
    • :: snowhenge dot net ::
Re: Can't match two monitors
« Reply #4 on: November 27, 2018, 04:51:11 am »

Ok, understood, marketing and reality are different realities.

Sorry if I’m being dense, but what did you mean by “you're also crippling the full capabilities of the Eizo” ?
Logged
--
David Mantripp

Czornyj

  • Sr. Member
  • ****
  • Online Online
  • Posts: 1857
    • zarzadzaniebarwa.pl
Re: Can't match two monitors
« Reply #5 on: November 27, 2018, 09:44:47 am »

try to enable:
Preferences>Measurement device>Multiple Monitor Matching
« Last Edit: November 27, 2018, 12:50:12 pm by Czornyj »
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 15490
    • http://www.digitaldog.net/
Re: Can't match two monitors
« Reply #6 on: November 27, 2018, 12:48:19 pm »

Ok, understood, marketing and reality are different realities.

Sorry if I’m being dense, but what did you mean by “you're also crippling the full capabilities of the Eizo” ?
Sorry, I meant if you were not using the Eizo software but it appears for that display you are, but not the other which you can't.
Logged
Andrew Rodney
Author “Color Management for Photographers"

Ethan Hansen

  • Full Member
  • ***
  • Offline Offline
  • Posts: 132
    • Dry Creek Photo
Re: Can't match two monitors
« Reply #7 on: November 28, 2018, 01:24:14 pm »

Your main problem is lack of hardware control over the Quato. Ideally you can set the white point identically between the two monitors, but you will rely on the profile LUTs to massage all darker tones to the correct color temperature. If you measure the contrast of both monitors, set the white point luminance and color temperature identically, and use a target contrast of the lesser max value of the two monitors, there may be hope.

Another issue, however, is the different calibration packages. Even with the same codebase, ColornNavigator 7 and i1Profiler give different takes on both calibration and profiling. Quato's legacy software was iColor Display. If my memory isn't too fuzzy, this was a rebranded version of BasICColor Display. It might be worth downloading an evaluation version of Display 5 to see if it can talk via DDC to the Quato as well as your Eizo.

Dave Rosser

  • Full Member
  • ***
  • Offline Offline
  • Posts: 228
    • My Website
Re: Can't match two monitors
« Reply #8 on: November 28, 2018, 06:04:35 pm »

You're probably on a hiding to nothing, I have a NEC PA241W and a NEC P242W display and despite calibrating both with Spectraview2 software and achieving deltaE of less than 1 for both neutral grey is visibly different. The reason is the PA241 has a ccfl backlight while the P242W has a LED backlight.
The Spectraview software does a!low visual matching of the two displays but it is really not worth the effort.
Logged

faberryman

  • Sr. Member
  • ****
  • Online Online
  • Posts: 856
Re: Can't match two monitors
« Reply #9 on: November 28, 2018, 06:17:43 pm »

You're probably on a hiding to nothing, I have a NEC PA241W and a NEC P242W display and despite calibrating both with Spectraview2 software and achieving deltaE of less than 1 for both neutral grey is visibly different. The reason is the PA241 has a ccfl backlight while the P242W has a LED backlight. The Spectraview software does a!low visual matching of the two displays but it is really not worth the effort.
If they are visibly different, which one, if either, is correct.

Dave Rosser

  • Full Member
  • ***
  • Offline Offline
  • Posts: 228
    • My Website
Re: Can't match two monitors
« Reply #10 on: November 29, 2018, 01:40:28 pm »

If they are visibly different, which one, if either, is correct.
A very good question, as most of the world uses LED back light these days go with that.  Of course I have a difficulty there as my ccfl device is Adobe RGB while my LED device is sRGB  :-\

Dave
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1789
Re: Can't match two monitors
« Reply #11 on: November 29, 2018, 02:03:20 pm »

You're probably on a hiding to nothing, I have a NEC PA241W and a NEC P242W display and despite calibrating both with Spectraview2 software and achieving deltaE of less than 1 for both neutral grey is visibly different. The reason is the PA241 has a ccfl backlight while the P242W has a LED backlight.
The Spectraview software does a!low visual matching of the two displays but it is really not worth the effort.

Not often talked about but well known to color scientists is that people with otherwise normal color vision can and do often see different colors that have identical CIEXY 1931 average observer values. This is because light with different spectrums may match for one person but be quite a bit off for another person. And the grays are particularly sensitive to individual differences.

Monitors are intrinsically spectrally spikey and the wider the gamut the more spike separation is required to increase the gamut. But more than that, the spikes associated with LEDs are differently located than CFLs. As a result, even if you could calibrate and profile a monitor perfectly, with a $10k lab grade reference instrument, grays that match on a LED monitor and CFL monitor in terms of the CIE Standard Observer model will very often appear different. For some individuals, quite different. The current state of the art has no way to address this because it's individual differences. The future? It's quite possible decades from now that you can run your dna genotype through a database that estimates these slight color vision differences. Brave New World.

For now the best advice is use monitors that have the same backlight tech.
Logged

Dave Rosser

  • Full Member
  • ***
  • Offline Offline
  • Posts: 228
    • My Website
Re: Can't match two monitors
« Reply #12 on: November 29, 2018, 05:24:43 pm »

Not often talked about but well known to color scientists is that people with otherwise normal color vision can and do often see different colors that have identical CIEXY 1931 average observer values. This is because light with different spectrums may match for one person but be quite a bit off for another person. And the grays are particularly sensitive to individual differences.

Monitors are intrinsically spectrally spikey and the wider the gamut the more spike separation is required to increase the gamut. But more than that, the spikes associated with LEDs are differently located than CFLs. As a result, even if you could calibrate and profile a monitor perfectly, with a $10k lab grade reference instrument, grays that match on a LED monitor and CFL monitor in terms of the CIE Standard Observer model will very often appear different. For some individuals, quite different. The current state of the art has no way to address this because it's individual differences. The future? It's quite possible decades from now that you can run your dna genotype through a database that estimates these slight color vision differences. Brave New World.

For now the best advice is use monitors that have the same backlight tech.
Thanks very much indeed for that very clear explanation.

Dave
Logged

GWGill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 528
  • Author of ArgyllCMS & ArgyllPRO ColorMeter
    • ArgyllCMS
Re: Can't match two monitors
« Reply #13 on: November 29, 2018, 06:27:31 pm »

The future? It's quite possible decades from now that you can run your dna genotype through a database that estimates these slight color vision differences.
Hmm. Last time I heard Mark Fairchild talk about this idea, he seemed more pessimistic than he had been in the past (Boston CIC tute.). As I understand it, DNA analysis can get you in the ballpark, but there are other aspects that can't be predicted from the genome that have at least equal influence (Macula density, lens & cornea yellowing etc. ?). So in practical terms, measuring CMF's might be the best approach. On the plus side, the models for CMF variability are now well developed.
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1789
Re: Can't match two monitors
« Reply #14 on: November 29, 2018, 07:16:05 pm »

Hmm. Last time I heard Mark Fairchild talk about this idea, he seemed more pessimistic than he had been in the past (Boston CIC tute.). As I understand it, DNA analysis can get you in the ballpark, but there are other aspects that can't be predicted from the genome that have at least equal influence (Macula density, lens & cornea yellowing etc. ?). So in practical terms, measuring CMF's might be the best approach. On the plus side, the models for CMF variability are now well developed.
I agree. And I think the additional factors beyond the genetic are quite significant as well. Particularly as most testing is done on the available population: students.  Not exactly representative of the broader range of folks out there.

I suspect, unfortunately, an instrument that could do this will be a long time coming. If ever.
Logged

daicehawk

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 88
Re: Can't match two monitors
« Reply #15 on: November 30, 2018, 02:23:51 am »

I am guessing an instrument with a higher resolution than i1pro like jeti would minimize the visual difference, though that would leave the individual CMFs unaddressed.
Logged

JRSmit

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 618
    • Jan R. Smit Fine Art Printing Specialist
Re: Can't match two monitors
« Reply #16 on: November 30, 2018, 02:43:17 am »

What we see and what f.i. I1 display pro sees is different. I once calibration and profiled two monitors, both ccfl wide gamut. Looked slightly different, so i took my camera and made a picture with both together, showing 50% grey. On the image the monitors looked identical, and measured identical. The one with the difference (see vs measure) was the NEC pa271w. For me it was the coating, which added some crystallic like “thing” .
Nowadays i use an EIZO 248, which has a different coating. The NEC i still use as an office monitor, and still see this “crystallic thing”. The EIZO does not.
Logged
Fine art photography: janrsmit.com
Fine Art Printing Specialist: www.fineartprintingspecialist.nl


Jan R. Smit

32BT

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3012
    • Pictures
Re: Can't match two monitors
« Reply #17 on: November 30, 2018, 03:34:36 am »

I have always wondered whether you could make this work if you just put more distance between the sensing device and the monitor. In addition, the sensing device and computational model are designed for the 2degr observer problem, whereas this seemed to be a problem for either the 10degr observer or even wider?

I think that discussing dna differences is kind of odd since it is a solution with an extreme precission whereas the problem suggests a much coarser solution should be possible. In addition it raises the question of who determines the reference norm...
Logged
Regards,
~ O ~
If you can stomach it: pictures

Czornyj

  • Sr. Member
  • ****
  • Online Online
  • Posts: 1857
    • zarzadzaniebarwa.pl
Re: Can't match two monitors
« Reply #18 on: November 30, 2018, 04:14:32 am »

There's no practical solution to such problem other than matching both displays visually. You can use super expensive high resolution spectroradiometer like JETI Specbos, Photo Research, Konica-Minolta etc., calibrate from a distance, and the difference will be virtually identical, or similar - in most cases i1Display Pro is surprisingly accurate.

It's not about observer variability issues - when you see magenta/green tint on calibrated display and ask a couple of other people if it's neutral, they will also tell you it has magenta/green tint.

The issue is most probably caused by postreceptoral chromatic adaptation.
Quote
In the same cross-media colour matching conditions, we find significant systematic discrepancies between the predictions of the CIE Standard Colorimetric Observer and the mean matches made by the group of observers. We attribute these discrepancies to additivity failures caused by postreceptoral adaptation, leading to nonlinear change in sensitivity of S/(L+M) chromatic channel.
http://oicherman.com/Boris/Science/Boris_Oicherman_Phd_thesis.pdf

32BT

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3012
    • Pictures
Re: Can't match two monitors
« Reply #19 on: November 30, 2018, 04:39:05 am »

The issue is most probably caused by postreceptoral chromatic adaptation.

But the problem is immediately obvious and very significant. So, would it be possible to simplify the sensing to solve the problem? I.e. the right tricolor sensor, or even some kind of bicolor sensor, might be able to solve this on the cheap.
Logged
Regards,
~ O ~
If you can stomach it: pictures
Pages: [1] 2   Go Up