Lightroom and Photoshop color mismatch

[E_Edwards]

Can anyone explain why the image I see in the Lightroom Develop window doesn't quite match when I open it in Photoshop after  export from Lightroom?

 It's pretty close, and easy to correct, but I would prefer the Lightroom file to look identical when opened in Photoshop.


My Lightroom Export setting is set to AbobeRGB and the Photoshop Color Setting is set to a Working Space:  AdobeRGB.


As I understand it, Lightroom doesn't assign a working space to the raw file until export, where you can then choose to export it as sRGB, ProPhotoRGB or AdobeRGB.

So does that mean that what I see in the Lightroom Develop window is the profiled Monitor color space? If so, is Lightroom converting from Monitor color Space to AdobeRGB color space on Exporting?. And if that is the case, why is there a small disparity in the conversion?


This happens on the new Beta2 as well.


Random sample file attached.(Mac and calibrated Nec2180  with Spectraview and iOne pro).

Thanks

Edward


[attachment=5961:attachment]

[Hermie]

Can anyone explain why the image I see in the Lightroom Develop window doesn't quite match when I open it in Photoshop after  export from Lightroom?

 It's pretty close, and easy to correct, but I would prefer the Lightroom file to look identical when opened in Photoshop.
My Lightroom Export setting is set to AbobeRGB and the Photoshop Color Setting is set to a Working Space:  AdobeRGB.
As I understand it, Lightroom doesn't assign a working space to the raw file until export, where you can then choose to export it as sRGB, ProPhotoRGB or AdobeRGB.

So does that mean that what I see in the Lightroom Develop window is the profiled Monitor color space? If so, is Lightroom converting from Monitor color Space to AdobeRGB color space on Exporting?. And if that is the case, why is there a small disparity in the conversion?
This happens on the new Beta2 as well.
Random sample file attached.(Mac and calibrated Nec2180  with Spectraview and iOne pro).

Thanks

Edward
[attachment=5961:attachment]
[{POST_SNAPBACK}][/a]

What you're seeing in Lightroom's develop mode is the image rendered to Lightroom's internal color space Melissa RGB. See for example [a href=\"http://www.xs4all.nl/~teeuwen/hermiergb.html]http://www.xs4all.nl/~teeuwen/hermiergb.html[/url]

[E_Edwards]

Thank you, Hermie.

I tried assigning your sample profile, but it doesn't improve it.

I was under the impression that both the raw as viewed in Lightroom Develop window and the file exported as ProPhotoRGB file and then viewed in Photoshop also with ProPhotoRGB, preserving profiles, should look exactly the same. In this case, they don't.

It could be my monitor profiling, but I reprofiled the Spectraview2180 and the same small difference still persists, so I will give up at this point.

Let me say that this only happens with captured raw files (which don't contain embedded profiles). With other files, for example tif files from other sources which already have an embedded profile, they look identical in Lightroom.

Many thanks.

Edward

[Hermie]

> I was under the impression that both the raw as viewed in Lightroom Develop window and the file exported as ProPhotoRGB file and then viewed in Photoshop also with ProPhotoRGB, preserving profiles, should look exactly the same.

They should indeed look exactly the same.
It must be something else then.

> I tried assigning your sample profile, but it doesn't improve it.

Assigning? I don't get it. Assigning this sample profile to aRGB file in PS?; that won't help.

Also note that different background colors in LR and PS may influence your perception.


Imageingester's latest version of the standalone LRViewer (Lightroom viewer) is now color managed and uses this hermiergb profile to render images a la Lightroom (for screen diplay and profile embedding in exported JPEGs): http://basepath.com/ImageIngester/LRViewer-info.php.

[E_Edwards]

I posted the little sample file in the initial post to show that it is a real, though minor difference. It certainly is not imaginary! The bottom file is very slightly going towards the blue/magenta side. The slight yellowness at the top of the top file has gone in the bottom file, for instance.

Someone mentioned that it could be to do with my monitor profile and its rendering intent. But it's the same calibrator and the same profiling software that I use for other monitor brands that don't seem to show the same discrepancy. I have four Spectraview2180 and they all have this difference.

However, on a Apple 30" Cinema Display, I can't see the difference, the files seem pretty identical. Does that give you a possible clue?

It's not a big deal, I can live with it, but it's nice to know the reason why.

[digitaldog]

Doesn't matter the color spaces don't match, the color appearance should (and for me does). You do need to view both previews at the same size and at least zoomed to 100%.

[E_Edwards]

Doesn't matter the color spaces don't match, the color appearance should (and for me does). You do need to view both previews at the same size and at least zoomed to 100%.
[a href=\"index.php?act=findpost&pid=188301\"][{POST_SNAPBACK}][/a]

Hi Andrew,

I've gone by the book, 100% in Develop window, correct profile matching, etc. and to no avail.


However, good news. I think I have discovered the culprit.

This time, instead of using my own monitor profile, I went and used the original factory supplied monitor profile for the Spectraview. I then suddenly had perfect match between the two programs. I actually took a screenshot of both windows  (the picture was a digital scale - black-midgreys-white) so I could measure the difference. And using the factory profile, the measument was identical in both Photoshop and Lightroom, like it should be.

I then swapped to my own created monitor profile, and the difference was measurable in the black patch only, the mid-grey and white patches were pretty much identical, or very close.

So my conclusion, some setting in the Spectraview profiler or the profiles it creates, interfere with the correct matching in Lightroom/Photoshop in the shadow area mainly.
(The difference is about 2-4 units, so not much, but visible).

Does this make sense?


Here is the difference

[attachment=6032:attachment]

[Hermie]

> This time, instead of using my own monitor profile, I went and used the original factory supplied monitor profile for the Spectraview. I then suddenly had perfect match between the two programs.

I'm not familiar with SpectraView software, but what type of profile is your custom profile, matrix, LUT?

Photoshop used to have some issues with LUT based profiles, I'm not sure what the current status is however.
See e.g. http://lists.apple.com/archives/Colorsync-...n/msg00129.html

[E_Edwards]

Thank you.

From what I've read, it has a 10-bit internal LUT, though now we are getting very technical!

I'll try other settings but mainly, I think I'll try another software to calibrate, as a process of elimination.

[Hermie]

> From what I've read, it has a 10-bit internal LUT, though now we are getting very technical!

That's about the display, I meant the ICC profile type.

I just noticed on a NEC brochure that SpectraView support both LUT-based and matrix profiles.

I would check my settings in SpectraView and if it turns out that your current profile is a LUT-based profile, I would try a matrix version.

[E_Edwards]

I would check my settings in SpectraView and if it turns out that your current profile is a LUT-based profile, I would try a matrix version.
[a href=\"index.php?act=findpost&pid=188390\"][{POST_SNAPBACK}][/a]

I tried the Matrix setting on the Spectraview program but after waiting for the profile stages for ages, when I opened Photoshop it tells me that this particular monitor profile appears to be faulty. Indeed, the colors were all over the place. I made another one on matrix just in case, same faulty message, so I gave up.

I then made a monitor profile using good old ProfileMaker 5 and it's all working beautifully first time round, matching colors and values beautifully. It also created a profile considerably faster than Spectraview. So I'll probably stick with ProfileMaker.

Many thanks for your help and interest.

Edward

[digitaldog]

So I'll probably stick with ProfileMaker.

I'd advise against this idea. The SpectraView system is going to do a much better job, PMP's display module is the weak link in their suite and, you're not using the capabilities of the display to anywhere near its potential.

I'll add, I have a 2690 driven by SpectraView II and have no issues between PS and LR.

[E_Edwards]

I'll add, I have a 2690 driven by SpectraView II and have no issues between PS and LR.
[a href=\"index.php?act=findpost&pid=188449\"][{POST_SNAPBACK}][/a]

Since you have no issues, I decided to download the latest version of Spectraview (I was using the older v3.0.3 and apparently they have redesigned the color engine)).

I'm happy to report that it works like a dream, now calibrated with Spectraview 4.1.8

I used Matrix, but maybe LUT is just as good.

So all is well. Thanks

[digitaldog]

Spectraview 4.1.8

Could that be 1.04?

[E_Edwards]

Could that be 1.04?
[a href=\"index.php?act=findpost&pid=188496\"][{POST_SNAPBACK}][/a]

Andrew,

The latest version of SpectraView Profiler is v4.1.8 and I've just been to their website to download it.

If you own one of their monitors and the sofware serial number (or TAN as they call it) they will send you a license to activate it.

Maybe for your monitor there is a different software, or maybe they call it differently in the States.

I have not seen SpectraviewII listed on the Spectraview website, they only list SpectraView Profiler.

[digitaldog]

Andrew,

The latest version of SpectraView Profiler is v4.1.8 and I've just been to their website to download it.

OK we're using different products. I'm using the NEC software (SpectraView II) that is supplied to the US. You're using the European stuff. Mines better <g>.

[Fabrice]

OK we're using different products. I'm using the NEC software (SpectraView II) that is supplied to the US. You're using the European stuff. Mines better <g>.
[{POST_SNAPBACK}][/a]

The latest version of Nec Spectraview Profiler is v4.1.9, dated May 5 '08 and available from [a href=\"http://www.nec-display-solutions.com/specials/spectraview/index-en.html]http://www.nec-display-solutions.com/speci...w/index-en.html[/url].  I am using Spectraview II v1.0.41 Build 62101.

I don't think SpectraView II has been updated as recently as SpectraView Profiler (the copyright date for SpectraView II v1.0.41 is 2003-2007).  I'm using SpectraView II to calibrate a NEC 1980sxi.  Would I benefit from upgrading to Spectraview Profiler v4.1.9?

The SpectraView Profiler v4.1.9 release apparently has the following:

... new calibration algorithm (CIECAM02),
....automatic hardware adjustment of JUST viewing booth via software,
....Universal Application (Mac with Intel processors support)
....Colorvision Spyder 3

Is this new calibration algorithm something that would improve the results over what we're getting with the older Spectraview II v1.0.41?

[Fabrice]

Update:  searching the forums I found the link Andrew posted to the US version of the software and was able to download the patch for SpectraView II v1.0.42: http://www.necdisplay.com/SupportCenter/Mo...s/spectraview2/  

The copyright date is now 2003-2008, and I'm assuming SpectraView II has all the functionality of the Euro version (Spectraview Profiler).

The latest version of Nec Spectraview Profiler is v4.1.9, dated May 5 '08 and available from http://www.nec-display-solutions.com/speci...w/index-en.html.  I am using Spectraview II v1.0.41 Build 62101.

I don't think SpectraView II has been updated as recently as SpectraView Profiler (the copyright date for SpectraView II v1.0.41 is 2003-2007).  I'm using SpectraView II to calibrate a NEC 1980sxi.  Would I benefit from upgrading to Spectraview Profiler v4.1.9?

The SpectraView Profiler v4.1.9 release apparently has the following:

... new calibration algorithm (CIECAM02),
....automatic hardware adjustment of JUST viewing booth via software,
....Universal Application (Mac with Intel processors support)
....Colorvision Spyder 3

Is this new calibration algorithm something that would improve the results over what we're getting with the older Spectraview II v1.0.41?
[a href=\"index.php?act=findpost&pid=201074\"][{POST_SNAPBACK}][/a]