Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: Adobe98 vs ProPhoto gamut question?  (Read 2493 times)
Redcrown
Full Member
***
Offline Offline

Posts: 153


« on: November 29, 2012, 05:18:52 PM »
ReplyReply

While editing an image in Photoshop (CS6) I stumbled across a strange out-of-gamut condition than confuses me.

My image was in Adobe98 colorspace, and when I went to Soft Proof the image (View/Proof colors), my default happened to be ProPhoto colorspace. The gamut warning showed some small out-of-gamut areas. How can that be? Everything I've seen says that the Adobe98 colorspace is fully contained inside the ProPhoto colorspace. So how can an Adobe98 value be outside of the gamut of ProPhoto?

I found these examples:

Create an Adobe98 document, fill it with the RGB values of 29/0/0 ( a very dark red). Turn on soft proofing and gamut warning using the ProPhoto colorspace. Use any rendering intent, does not make any difference. My system shows that color to be out-of-gamut, does yours?

I found the following Adobe98 RGB values to show as out-of-gamut in ProPhoto: 28/0/0, 29/0/0. No red value below 28 or above 29 shows as out-of-gamut. Also, the green values of 25 thru 28 with red and blue at zero show out-of-gamut. I can't find any blue values that are out-of-gamut in ProPhoto.
Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 9241



WWW
« Reply #1 on: November 29, 2012, 05:23:12 PM »
ReplyReply

The gamut warning is both buggy and not real useful. Best to ignore it.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Redcrown
Full Member
***
Offline Offline

Posts: 153


« Reply #2 on: November 29, 2012, 11:03:48 PM »
ReplyReply

Andrew, thanks for taking the time to reply. A lot of what I've learned about color spaces, soft proofing, and gamut warnings came from your various publications over the years. Your recent videos on Lightroom 4 soft proofing, for example, are very good tutorials. They are the only source I think I've seen that advise us to simply let the profile conversion take care of gamut issues instead of trying to do a bunch of manual edits to fix out-of-gamut conditions.

But this is the first time I've heard you say that Adobe's soft proof gamut warning is buggy and useless. I take from that reply that you think the mystery I described in my original post is the result of a bug, or otherwise inaccurate display of out-of-gamut values in Adobe software?
Logged
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3884


« Reply #3 on: November 30, 2012, 03:18:33 AM »
ReplyReply

While editing an image in Photoshop (CS6) I stumbled across a strange out-of-gamut condition than confuses me.

My image was in Adobe98 colorspace, and when I went to Soft Proof the image (View/Proof colors), my default happened to be ProPhoto colorspace. The gamut warning showed some small out-of-gamut areas. How can that be? Everything I've seen says that the Adobe98 colorspace is fully contained inside the ProPhoto colorspace. So how can an Adobe98 value be outside of the gamut of ProPhoto?

I found these examples:

Create an Adobe98 document, fill it with the RGB values of 29/0/0 ( a very dark red). Turn on soft proofing and gamut warning using the ProPhoto colorspace. Use any rendering intent, does not make any difference. My system shows that color to be out-of-gamut, does yours?

I found the following Adobe98 RGB values to show as out-of-gamut in ProPhoto: 28/0/0, 29/0/0.

Hi,

According to Bruce Lindbloom's CIE calculator, AdobeRGB 29/0/0 translates via 'Lab' to ProPhoto RGB 15/6/2 (after rounding to integer values). Should be in gamut. Seems to support Andrew's statement about the PS implementation, in fact the implementation is not buggy but wrong ... Even a clock that is not functioning, is still correct twice a day, but I wouldn't call that buggy either.

Cheers,
Bart
Logged
kers
Sr. Member
****
Offline Offline

Posts: 789


WWW
« Reply #4 on: November 30, 2012, 03:37:06 AM »
ReplyReply

So we have to wait for photoshop CS7 to get this right after 20 years of improvements like -content aware fills- that never work?
Some companies improve only on gadgets and forget the basics.. ( Apple is in such a phase)


Logged

Pieter Kers
www.beeld.nu
Rhossydd
Sr. Member
****
Offline Offline

Posts: 1993


WWW
« Reply #5 on: November 30, 2012, 03:49:38 AM »
ReplyReply

So we have to wait for photoshop CS7 to get this right.....
Given how long this has been an issue, I doubt it will get fixed.
Logged
john beardsworth
Sr. Member
****
Online Online

Posts: 2893



WWW
« Reply #6 on: November 30, 2012, 05:31:30 AM »
ReplyReply

So we have to wait for photoshop CS7 to get this right after 20 years of improvements like -content aware fills- that never work?
Some companies improve only on gadgets and forget the basics.. ( Apple is in such a phase)
Sarcasm: I guess they should cater to pixel peepers and not add features that often work very well....
« Last Edit: November 30, 2012, 07:39:00 AM by johnbeardy » Logged

digitaldog
Sr. Member
****
Offline Offline

Posts: 9241



WWW
« Reply #7 on: November 30, 2012, 08:44:05 AM »
ReplyReply

The bug is easy to observe, just open an sRGB document, load sRGB as the soft proof and put on the OOG overlay. Depending on the image content, you'll see the overlay which you should not.

I seriously doubt this will be fixed. It's low hanging fruit and besides, after the release of Photoshop 5 (NOT CS5), OOG wasn't necessary even without the bugs as we had soft proofing.

What I would like Photoshop to do is use the OOG preview to show us the differences between our display gamut and the working space, something LR does if you load a working space for soft proofing. At least you can observe what colors fall outside display gamut. Would make showing people what a wide gamut display brings to the party. Otherwise, OOG overlay is IMHO just short of useless. Now if some day we could have the OOG treat the overlay in a user defined scale: make colors way out of gamut or if you prefer, dE of 0-1 be yellow, 2-5 be green, anything over 6 red and so forth, as ColorThink does that could be somewhat useful.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Rob Reiter
Full Member
***
Offline Offline

Posts: 235



WWW
« Reply #8 on: November 30, 2012, 11:08:29 AM »
ReplyReply

This is not really an answer to your problem, but as someone who deals with this stuff everyday in a fine art printing studio, even if you have the best tools and techniques out there, soft proofing might get you 90-95% of the way there. In the end, you have to do a hard copy printed proof to really squeeze the best out of any image.

But what we have today sure beats the bad old days of darkroom color printing, IMHO.
Logged

http://www.lightroom.com Fine art printing for photographers and other artists
kers
Sr. Member
****
Offline Offline

Posts: 789


WWW
« Reply #9 on: November 30, 2012, 04:10:33 PM »
ReplyReply

Sarcasm: I guess they should cater to pixel peepers and not add features that often work very well....
sarcasm: why buy a 36mp camera if your work only has to serve on internet ?
( bottomline- photoshop is used by a very wide audience with different objectives- and i think 1+1 must be 2 - it cannot be very hard to repair this flaw)
« Last Edit: November 30, 2012, 04:17:44 PM by kers » Logged

Pieter Kers
www.beeld.nu
john beardsworth
Sr. Member
****
Online Online

Posts: 2893



WWW
« Reply #10 on: December 01, 2012, 02:21:31 AM »
ReplyReply

sarcasm: why buy a 36mp camera if your work only has to serve on internet ?
( bottomline- photoshop is used by a very wide audience with different objectives- and i think 1+1 must be 2 - it cannot be very hard to repair this flaw)
Sacrasm: So let's not buy 36mp cameras and cease printing until this somewhat obscure flaw is resolved? OK, pixel peepers deserve a niche as one of Photoshop's many groups of users, but it's nonsense to suggest that "20 years of improvements like - content aware fills - ...never work".
Logged

jeremypayne
Guest
« Reply #11 on: December 01, 2012, 06:38:31 AM »
ReplyReply

Does it REALLY matter?

As with any software development effort of this scale, there literally thousands upon thousands of "things" they COULD work on ... They have to pick their battles.

What would the ROI be for this bug fix?  What would be the true practical benefit of fixing this issue to whom and under what circustances?

And ... Unless you know the code intimately, you really can't tell how hard this is to fix.  It may be "wrong" because the code is shared with some other aspect of the application that needs it to to be "wrong" to get something else ... Something more important ... "Right".  Not saying that IS the case ... Just saying I've authorized plenty of similar kinds of "hacks" in complex applications in my time as a manager of software development projects ...

Logged
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad