Pages: 1 2 [3] 4 5   Go Down

Author Topic: arbitrary RGB profiles for opening RAW  (Read 39898 times)

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Re: arbitrary RGB profiles for opening RAW
« Reply #40 on: April 07, 2014, 12:55:41 am »

So is someone going to tell me that Soft Proof in Photoshop doesn't render anything but relative?

It will if you use an output profile with a perceptual table in it (most do) but not when using a matrix color space profile–unless it's a V4 profile with a perceptual rendering in it–such as the such as the special sRGB v4 Appearance (beta).
Logged

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Re: arbitrary RGB profiles for opening RAW
« Reply #41 on: April 07, 2014, 09:08:44 am »

It will if you use an output profile with a perceptual table in it (most do) but not when using a matrix color space profile–unless it's a V4 profile with a perceptual rendering in it–such as the such as the special sRGB v4 Appearance (beta).

WE are informed that perceptual rendering is performed via lookup tables in the receiving profile (typically a printer profile with perceptual lookup tables) and this may be the case with V2 profiles. However, with V4 profiles things may be different as I interpret the example in White Paper 26. As explained in the text immediately below Figure 2 in that paper, it states that when the Ver 4 sRGB profile is used as a source, the sRGB colors are re-rendered into the PRM (Perceptual Reference Medium), apparently using the PRM information located in the Ver 4 source profile. When a Ver 4 profile is used as a destination, the colors are re-rendered from the PRM into the destination space. This is a two step process, as compared to one step with Ver 2 profiles.

The link that Jeff provides is an example of using the Ver 4 sRGB profile to print an image rendered into sRGB by a digital camera. This is not how most photographers using raw files would operate. Rather, the source master file would be in ProPhotoRGB and one might wish to use the Ver 4 sRGB profile to re-render into sRGB for use on the web. This is not covered in White Paper 26, but they do cover the case of the original file being in Adobe RGB, where one could convert to sRGB with clipping, assign the Ver 4 sRGB profile and then perform the perceptual rendering.

With a Prophoto source, converting to sRGB could involve quite a lot of clipping and the results might not be pleasing. It would seem better to assign a Ver 4 ProphotoRGB profile (perhaps ISO22028-3_RIMM-RGB-exCR.icc, as mentioned here), which would re-render the colors into the PRM and then render into sRGB Ver 4 with the perceptual intent. Does anyone have any experience or comments on this proposed work flow?

Bill









« Last Edit: April 07, 2014, 09:21:19 am by bjanes »
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20614
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: arbitrary RGB profiles for opening RAW
« Reply #42 on: April 07, 2014, 10:33:30 am »

As explained in the text immediately below Figure 2 in that paper, it states that when the Ver 4 sRGB profile is used as a source, the sRGB colors are re-rendered into the PRM (Perceptual Reference Medium), apparently using the PRM information located in the Ver 4 source profile.
Yes, if that V4 profile uses the PRMG of which I know of none that do, at least those built by X-rite, GMB, Copra and DataColor. There may be some profile builder out there that does, I'm not aware of any nor using any such products.

Quote
With a Prophoto source, converting to sRGB could involve quite a lot of clipping and the results might not be pleasing.
That's possible but I've not see it. There is still an issue of a lot of possible OOG colors getting clipped with this process. The gallon of water has to fit into the pint container using either RI, one is supposed to do a 'better/more pleasing' job perceptually but everyone's mileage may vary. It is quite possible a Perceptual RI in this example could make the image less pleasing than a RelCol intent.

Would it be nice if we had the options? Yes. Is it a huge problem? I think not, especially since so few are implementing V4 profiles with the PRMG.
« Last Edit: April 07, 2014, 10:36:16 am by digitaldog »
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Lundberg02

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 379
Re: arbitrary RGB profiles for opening RAW
« Reply #43 on: April 07, 2014, 03:46:42 pm »

This has got me so ticked off that I'm actually going to get off my a** and do some test prints to make sure that my Epson renders. Ticked off because none of this  valuable information has seen the light of day before now to my knowledge, at least not in places accessible to the average person struggling with color management.
One of the authors  posting here could do community service by writing an article ( with copious thanks to me in the introduction) about this hole in the fabric of space-time.
Logged

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Re: arbitrary RGB profiles for opening RAW
« Reply #44 on: April 07, 2014, 03:50:20 pm »

This has got me so ticked off that I'm actually going to get off my a** and do some test prints to make sure that my Epson renders. Ticked off because none of this  valuable information has seen the light of day before now to my knowledge, at least not in places accessible to the average person struggling with color management.

I've written about this issue in The Digital Print (and did so well before you asked your questions).
Logged

Lundberg02

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 379
Re: arbitrary RGB profiles for opening RAW
« Reply #45 on: April 07, 2014, 04:38:51 pm »

Stout lad! What page?
Logged

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Re: arbitrary RGB profiles for opening RAW
« Reply #46 on: April 08, 2014, 12:23:09 am »

Color Management (Chapter 2) starting on page 40 and a Tip: Matrix Profiles with Perceptual Rendering on page 44.
Logged

Lundberg02

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 379
Re: arbitrary RGB profiles for opening RAW
« Reply #47 on: April 08, 2014, 01:38:46 am »

Henry Ford once you can have any color car you want as long as it's black.
So I can have any intent I want as long as it's relative. I just printed the ProPhoto Printer test image using Photoshop manages colors in all four intents and there is NO difference between them. This seems to mean that the printer profile does not have anything but relative. Very disappointing.  Color Settings in PS was set to Pro Photo working space, Perceptual intent but this should have no bearing on the print flow.
The only profile I can find on the web that alleges to be perceptual is the ICC Appearance beta for sRGB.
No idea whether Best RGB, PhotoGamut or any others have anything but relative, as it is not mentioned. I feel like I've been swindled. I know of no way to obtain a perceptual rendering intent in my setup. I'll have to try PhotoGamut and  see if has the four intents. You would think it would.  I was unable to to find a PRGM profile in the ICC site oe anywhere else and I had the impression that it has to be built in to an icc profile in order to do its thing.
BTW, Burkhard Huber, developer of PhotoLine, says that you can add arbitrary profiles  to PL using the Camera Profile setting or the document setting.  He claims that all profiles, v2 or v4 will do relative and absolute, but did not offer any solution for perceptual. I doubt his statement about intents.
I read the section of your book, but I don't get a sense of outrage.  Adobe and Epson have failed me, my faith is crumbling like an Obama voter with a new health insurance policy. I'm beginning to doubt gamma  now. Color agnosticism.
Logged

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Re: arbitrary RGB profiles for opening RAW
« Reply #48 on: April 08, 2014, 01:42:26 am »

I read the section of your book, but I don't get a sense of outrage.

I don't exhibit "outrage" in a public form, I prefer to express my outrage in private.

And, if you are outputting from PP RGB to an Epson pro printer through all 4 rendering intents and you see no difference, you are doing something wrong (unless your target image falls inside the gamut of the printer). Where did you get the target you printed?
Logged

Lundberg02

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 379
Re: arbitrary RGB profiles for opening RAW
« Reply #49 on: April 08, 2014, 05:12:06 pm »

My current printer is an Epson Artisan 725 six color. I know you have at least three Epson pro level printers and I suppose you can confirm that they all do all four renderings. Why would Epson not provide that capability at any price point?
The test image is      PrinterEvaluationImage_V002_ProPhoto.tif       available from several sites on the web such as     
books.outbackphoto.com/DOP2010_03/printer_tests/‎
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20614
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: arbitrary RGB profiles for opening RAW
« Reply #50 on: April 08, 2014, 05:14:51 pm »

My current printer is an Epson Artisan 725 six color. I know you have at least three Epson pro level printers and I suppose you can confirm that they all do all four renderings. Why would Epson not provide that capability at any price point?
The test image is      PrinterEvaluationImage_V002_ProPhoto.tif       available from several sites on the web such as     
books.outbackphoto.com/DOP2010_03/printer_tests/‎
It's the profile we'd need to see. I doubt it doesn't have a perceptual table, I've never seen an output profile that didn't have one.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Re: arbitrary RGB profiles for opening RAW
« Reply #51 on: April 08, 2014, 05:54:08 pm »

The test image is      PrinterEvaluationImage_V002_ProPhoto.tif       available from several sites on the web such as     
books.outbackphoto.com/DOP2010_03/printer_tests/‎

Hum, I downloaded both the PP RGB and ARGB test file and when I put the ARGB on top of the PP RGB file and set the blend to Difference, I see very, very little difference which leads me to believe that the test file was created in Adobe RGB and then converted to ProPhoto RGB. If each color space was created in a native color space, I would expect the differences to be greater. So, the PP RGB test file is, I think a file that started life as Adobe RGB and was merely converted to ProPhoto RGB. So, that's not a real PP RGB test file.
Logged

samueljohnchia

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 498
Re: arbitrary RGB profiles for opening RAW
« Reply #52 on: April 08, 2014, 09:07:00 pm »

Hum, I downloaded both the PP RGB and ARGB test file and when I put the ARGB on top of the PP RGB file and set the blend to Difference, I see very, very little difference which leads me to believe that the test file was created in Adobe RGB and then converted to ProPhoto RGB. If each color space was created in a native color space, I would expect the differences to be greater. So, the PP RGB test file is, I think a file that started life as Adobe RGB and was merely converted to ProPhoto RGB. So, that's not a real PP RGB test file.

Not on my end Jeff. I plotted the two test images in ColorThink, against Adobe RGB. The ProPhoto test image does contain colors clearly outside of Adobe RGB.

Edit: Added the difference blend result I got in Photoshop.
« Last Edit: April 08, 2014, 09:12:04 pm by samueljohnchia »
Logged

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Re: arbitrary RGB profiles for opening RAW
« Reply #53 on: April 09, 2014, 01:30:58 am »

Not on my end Jeff. I plotted the two test images in ColorThink, against Adobe RGB. The ProPhoto test image does contain colors clearly outside of Adobe RGB.

Good to know...I didn't plot the PP RGB against the ARGB image in ColorThink...that seems to be fairly conclusive except I added the ARGB image into the PP RGB image and converted the ARGB to PP RGB as a layer. So, either the original test target was done in PP RGB and transformed to ARGB or the original target was in ARGB and transformed into PP RGB. Only the author of the target would know for sure how the target was made.
Logged

samueljohnchia

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 498
Re: arbitrary RGB profiles for opening RAW
« Reply #54 on: April 09, 2014, 02:15:01 am »

Only the author of the target would know for sure how the target was made.

I think it's fairly suggestive given that the Adobe RGB version is a jpeg file.  ;)

Based on Outback Photo's webpage description, I'll bet that it was indeed a ProPhoto original, and someone dumbed it down to Adobe RGB. The latter cannot be found from the original page.
Logged

Lundberg02

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 379
Re: arbitrary RGB profiles for opening RAW
« Reply #55 on: April 09, 2014, 02:48:12 am »

Thanks guys, very interesting. I guess I can use the Colorsync utility to inspect the Epson printer profile for the rendering tags, right?
Logged

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Re: arbitrary RGB profiles for opening RAW
« Reply #56 on: April 09, 2014, 02:52:29 am »

Yes...if you look at the various tags...
Logged

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3387
Re: arbitrary RGB profiles for opening RAW
« Reply #57 on: April 09, 2014, 11:02:18 am »

I think it's fairly suggestive given that the Adobe RGB version is a jpeg file.  ;)

Based on Outback Photo's webpage description, I'll bet that it was indeed a ProPhoto original, and someone dumbed it down to Adobe RGB. The latter cannot be found from the original page.

That would be my interpretation also. The clipping in the AdobeRGB image is consistent with conversion from a ProPhotoRGB image with a relative colorimetric rendering intent. A test image in AdobeRGB is not suitable for testing with wide gamut inkjet printers as shown below in a Colorthink plot of the AdobeRGB gamut (solid) vs Epson 3880 with glossy Epson paper (wire frame). No knowledgeable photographer would use a test image in AdobeRGB for testing of wide gamut printers.

Bill
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 20614
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: arbitrary RGB profiles for opening RAW
« Reply #58 on: April 09, 2014, 11:22:50 am »

Many of those images come from Bill Atkinson who scanned them on his Tango into Lab. I have zero idea what happened after that in respect to these analysis from this one site. I have the original Lab files from Bill so of course it's possible to convert from Lab to whatever you wish.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Schewe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6229
    • http:www.schewephoto.com
Re: arbitrary RGB profiles for opening RAW
« Reply #59 on: April 09, 2014, 04:01:47 pm »

I have the original Lab files from Bill so of course it's possible to convert from Lab to whatever you wish.

Well, if they came from Bill, I have a much higher confidence that Bill did the right thing :~)
Logged
Pages: 1 2 [3] 4 5   Go Up