Pages: [1]   Go Down

Author Topic: Printing from wide gamut color spaces: Obsessive confusion?  (Read 5827 times)

David Eckels

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3529
  • It's just a camera.
    • Website

I recently moved my Canon ipf8400 about a hundred miles away and 2500 feet higher and wanted to make sure it had survived. My initial test print using Andrew Rodney's (digitaldog) "Adobe RGB (1998)" test file looked OK and matched my calibrated Dell UH30 monitor well. However, a few weeks later I watched his excellent video advocating the use of wide gamut color spaces from which to print from PSCC2014 and wanted to test it out in LR6; you can find his video here. After converting a copy of his "Printer Gamut Test File" to sRGB in PSCC2014, I saved it as a 16-bit tiff along with his original, imported them into LR6, ensuring that I was in the ProPhoto working space. When I printed them with Relative Intent, I could immediately see the differences he discusses including the ProPhoto enhancements to saturation in the cyans and yellows, the smoothing of the magentas and other colors, as well as the improved shadow detail but for some reason I lost the blues in the ProPhoto version (outlined in red), but the sRGB blues looked OK (panels A and B). Note the "black" ring around the upper blue ball and the near black lower ball on the far right. The same thing happened printing the same file from PS. When I printed with Perceptual Intent, as Andrew recommended, I still saw the same noted differences as before, but now Bill's blue balls looked much better. The PI sRGB looked about the same (panel C), while the PI ProPhoto version looked more like those described by Andrew (panel D outlined in red), not quite, but better.
Please note this dreadful composite was created as follows. Not having a controlled viewing environment, I scanned each print at 300 dpi with 24-bit color on my Brother laser printer. Each file was saved as a png, then imported into PSCC2014 and combined, annotated, flattened, reduced, and saved as a max quality jpeg. Banding and slight colors shifts are artifacts of the scanning process, but my gestalt on viewing the composite is that it fairly approximates what I see in the prints.
In my second post, I show you screen captures of the LR6 Print module for each of the files and the two different rendering intents (see file names). For all of them, you can see the differences in saturation (less in sRGB, more in ProPhoto), same with shadow detail, but the black ring around the blues is missing! Again, some very subtle differences, but in essence as I have represented it. FWIW, these captures were all PNG files with no adjustments.
In my third post, I show the ProPhoto-Relative rendering screen capture and the preview that comes up on my Canon ipf8400: it clearly shows the black ring around the blues. My conclusion is that it's real: I see it, my printer software "sees it" and my printer prints it. I could print in PI, which I do all the time, but I want to understand this phenomenon, if possible. Maybe, I've screwed up somehow, but I don't think so. I am surmising that it may have something to do with emissivity, reflectivity, and how the different rendering intents are mapped at the time of output, but I am not sure and I would like to be able to obtain the same results as Andrew Rodney. In short:
Am I missing something?

PS Win 8.1-64 OS, Nvidia GeForce GTX 780
« Last Edit: May 25, 2015, 04:05:40 pm by David Eckels »
Logged

David Eckels

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3529
  • It's just a camera.
    • Website
Re: Printing from wide gamut color spaces: Obcessive confusion?
« Reply #1 on: May 25, 2015, 04:02:07 pm »

Screen captures. For some reason, I could not upload the PNG versions. These are jpeg files at max quality.

David Eckels

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3529
  • It's just a camera.
    • Website
Re: Printing from wide gamut color spaces: Obcessive confusion?
« Reply #2 on: May 25, 2015, 04:05:08 pm »

Comparison of ProPhoto RI view in LR6 Print Module with that which the Canon printer driver "sees."

hugowolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1001
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #3 on: May 25, 2015, 09:55:32 pm »

With what paper and profile?

Brian A
Logged

Royce Howland

  • Full Member
  • ***
  • Offline Offline
  • Posts: 155
    • Vivid Aspect Photography
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #4 on: May 25, 2015, 10:45:05 pm »

Yes, what paper and what profile? Just taking a guess, it looks to me like a matter related to black point compensation. But it also could be a flawed profile. Or both together. Generally speaking, choice of BPC or not impacts use of the Relative Colorimetric rendering intent, but not Perceptual. Choosing not to use BPC with the RC intent is not something I would typically do; I practically always use RC + BPC. Not using BPC with RC will lead to problems in dark tones.

If the initial set of 4 images are scans made from actual prints, then besides the dark blues getting blocked up to black with RC prints, there is another problem. There's a huge chunk of black at the bottom of the green gradient in the centre left of the test pattern. It also shows up in the dark greens in the swatches of fabrics in the upper right part of the test, where there appear to be lots of patches of black that are dark green in the original image. This appears in both the RC and Perceptual versions of the test prints, both sRGB and original ProPhoto RGB.

I don't use Lightroom myself. But if I recall correctly, it doesn't provide a choice to use BPC or not in the print module; it just enables BPC all the time. If that's the case, it seems to me something is wrong with this printer profile.

What happens if you soft proof with your printer profile in Photoshop? Do you see the same blocked up regions in the dark blues and dark greens?
« Last Edit: May 25, 2015, 10:46:49 pm by Royce Howland »
Logged

David Eckels

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3529
  • It's just a camera.
    • Website
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #5 on: May 26, 2015, 07:44:00 am »

Canon Premium Gloss 280, Canon supplied profile
BPC set in PS produces the same result
Soft proofing does not show the blocked up regions
The scans are admittedly crappy (best I could do) and the "black chunk" does not show as badly on the prints
Thanks for taking a stab

Royce Howland

  • Full Member
  • ***
  • Offline Offline
  • Posts: 155
    • Vivid Aspect Photography
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #6 on: May 26, 2015, 11:06:44 am »

Okay, the issue with the greens may be related to scanning the test prints. Perhaps the scanner simply doesn't register them. Looking at the Canon print preview display you showed, it's not blocking all the dark greens to black, so the Canon software thinks the printer should be able to print those colours. Let's set that aside and go back to the issue with the blues.

I looked at some Canon iPF profiles in the GamutVision tool. I checked CN_iPF9400_PremiumSemiGlossyPaper2-280.icc and CN_iPF9400_PremiumGlossyPaper2-280.icc, among others. When simulating conversion from the ProPhoto RGB colour space to the printer space with these profiles, they have a severe problem in the core blues, from about the mid-tone L all the way down to black. This is true with or without BPC. I believe these Canon profiles are broken, at least with source images in ProPhoto RGB. To get around it you'll have to print from a lesser gamut source space like Adobe RGB, which doesn't exhibit the same problem. Or else get custom profiles created for your Canon papers.

Slobodan Blagojevic

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 18115
  • When everyone thinks the same, nobody thinks
    • My website
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #7 on: May 26, 2015, 11:50:03 am »

... After converting a copy of his "Printer Gamut Test File" to sRGB in PSCC2014, I saved it as a 16-bit tiff along with his original, imported them into LR6, ensuring that I was in the ProPhoto working space...


I am confused...

1. How do you "ensure you are in the ProPhoto working space" in LR?

2. Why would you take an sRGB file and then "ensure you are in the ProPhoto working space"

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20838
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #8 on: May 26, 2015, 12:18:20 pm »


I am confused...
1. How do you "ensure you are in the ProPhoto working space" in LR?
2. Why would you take an sRGB file and then "ensure you are in the ProPhoto working space"
ANY LR/ACR processing is done in ProPhoto RGB primaries (gamut). From there, what you ask for in terms of color space upon export is what the final image is converted into.
If a rendered image is cataloged and just printed (no editing), the color space is honored for printing in the print module.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Slobodan Blagojevic

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 18115
  • When everyone thinks the same, nobody thinks
    • My website
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #9 on: May 26, 2015, 12:48:19 pm »

ANY LR/ACR processing is done in ProPhoto RGB primaries (gamut)...

Indeed, and that was my point... there is nothing to "ensure," it is in the ProPhoto by default.

David Eckels

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3529
  • It's just a camera.
    • Website
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #10 on: May 26, 2015, 01:13:47 pm »

1. How do you "ensure you are in the ProPhoto working space" in LR?
I just made sure that nothing had been inadvertently changed; tiff files were set to work on in ProPhoto. Minor point, but wanted to make sure nobody had to ask the question: Had I unintentionally changed the LR working space? No.
I looked at some Canon iPF profiles in the GamutVision tool. I checked CN_iPF9400_PremiumSemiGlossyPaper2-280.icc and CN_iPF9400_PremiumGlossyPaper2-280.icc, among others. When simulating conversion from the ProPhoto RGB colour space to the printer space with these profiles, they have a severe problem in the core blues, from about the mid-tone L all the way down to black. This is true with or without BPC. I believe these Canon profiles are broken, at least with source images in ProPhoto RGB. To get around it you'll have to print from a lesser gamut source space like Adobe RGB, which doesn't exhibit the same problem. Or else get custom profiles created for your Canon papers.
I think this may well be right. I'll have to get with Andrew for some custom profiles   ;D

Thanks for the help, folks and Andrew, thanks for the exercise!

Dale_Cotton2

  • Full Member
  • ***
  • Offline Offline
  • Posts: 109
    • http://daystarvisions.com
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #11 on: May 27, 2015, 01:56:23 pm »

I did the same test a few weeks ago on my Epson 3880 after watching Andrew's pro-ProPhoto tutorial. I only did one print in ProPhoto rel/col and one in AdobeRGB rel/col. Same dark ring on the blue ball in the ProPhoto but not in the AdobeRGB. Used the Epson profile and Premium Luster. Are we not simply seeing out-of-gamut for the paper/ink? When I do a out-of-gamut display in Photoshop it shows more than half the image as out-of-gamut in ProPhoto.

Is there a way to know whether this is the actual ink+paper or just the limitations of the profile?
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20838
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #12 on: May 27, 2015, 02:45:38 pm »

I did the same test a few weeks ago on my Epson 3880 after watching Andrew's pro-ProPhoto tutorial. I only did one print in ProPhoto rel/col and one in AdobeRGB rel/col. Same dark ring on the blue ball in the ProPhoto but not in the AdobeRGB.
I think that's a limitation in the canned Epson profile. I don't see this at all in my custom profile and did see it when testing Epson's profile (Luster).
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

David Eckels

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3529
  • It's just a camera.
    • Website
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #13 on: May 27, 2015, 02:46:05 pm »

I did the same test a few weeks ago on my Epson 3880 after watching Andrew's pro-ProPhoto tutorial. I only did one print in ProPhoto rel/col and one in AdobeRGB rel/col. Same dark ring on the blue ball in the ProPhoto but not in the AdobeRGB. Used the Epson profile and Premium Luster. Are we not simply seeing out-of-gamut for the paper/ink? When I do a out-of-gamut display in Photoshop it shows more than half the image as out-of-gamut in ProPhoto.

Is there a way to know whether this is the actual ink+paper or just the limitations of the profile?
See! I'm NOT crazy ;)

David Eckels

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3529
  • It's just a camera.
    • Website
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #14 on: May 27, 2015, 02:47:54 pm »

I think that's a limitation in the canned Epson profile. I don't see this at all in my custom profile and did see it when testing Epson's profile (Luster).
This makes it virtually certain, we're dealing with profile issues. Thanks, Andrew!

Royce Howland

  • Full Member
  • ***
  • Offline Offline
  • Posts: 155
    • Vivid Aspect Photography
Re: Printing from wide gamut color spaces: Obsessive confusion?
« Reply #15 on: May 27, 2015, 10:55:58 pm »

It is not the ink+paper, or any kind of "normal" gamut limitation. A black ring in this specific case is the fault of a poorly constructed profile. Good generic profiles downloaded from paper vendors who know what they're doing, will not have this side effect. Custom profiles created by someone competent like Andrew won't have the problem either. Certainly my own don't have it. :)

The whole point of the Relative Colorimetric and Perceptual rendering intents in particular is to move out-of-gamut colours back into gamut in some kind of reasonable & pragmatic way. Clipping them to pure black is not reasonable or useful. Such profiles are broken IMO. Good profiles won't do this.
Pages: [1]   Go Up