Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Color Picker and ProPhotoRGB - false calculation

0 views
Skip to first unread message

Mike_Sc...@adobeforums.com

unread,
Mar 21, 2009, 4:36:58 AM3/21/09
to
This is sort of academic question. When using ProPhotoRGB as working space, Photoshop愀 color picker calculates R=0, G=0, B=255 (pure, unprintable blue) and the Lab colorspace is corresponding with L=0,a=90, b=-128. Most confusing is the L-value "0". Zero L(uminance) means black, absolutly black.
I checked other working spaces and none of them behaves like ProPhotoRGB.

Could someone clarify this?

Ann_She...@adobeforums.com

unread,
Mar 21, 2009, 2:29:18 PM3/21/09
to
Those L*ab readings will change depending on your working CMYK space.

Try converting to different CMYK spaces and see what happens

pfigen

unread,
Mar 21, 2009, 3:58:30 PM3/21/09
to
That does sound like a bug to me, although Photoshop has long be inconsistent in how it has dealt with colors spec'd outside the range of L*a*b or that of human vision.

Changing your CMYK space only changes the mix of CMYK in the Color Picker. The L*a*b values stay the same. The OP is concerned with how a ProPhoto 255 blue translates to L*a*b, a conversion that is independent of CMYK altogether. I get those same values when I fill a document with 255 blue and convert to L*a*b. It's clearly not a zero Luminance.

Ann_She...@adobeforums.com

unread,
Mar 21, 2009, 4:28:48 PM3/21/09
to

Changing your CMYK space only changes the mix of CMYK in the Color Picker.
The L*a*b values stay the same.


Not on my machine:

ProPhoto RGB 00; 00; 255 results in:

L*ab: 37; 11; -48 with Web Coated SWOP 2

L*ab: 39; 7; -35 with Web UnCoated 2

L*ab: 34; 9; -42 with Sheetfed Coated

Even in the Color Picker, RGB 00; 00; 255: equates to: L=30, a=68, b=-112

I don't know what settings Mike can be using to get that L=0, a=90, b=-128 reading?

pfigen

unread,
Mar 21, 2009, 5:26:17 PM3/21/09
to
He has typed in 0,0,255 in the RGB color picker using ProPhoto. It's the same here on an old laptop with CS1.

You have to physically retype the CMYK numbers in the color picker dialog to have the color engine refigure the L*a*b number, which it does when you do. But type in the prescribed RGB numbers in ProPhoto and you'll see the same wrong L*a*b numbers. Typing in the CMYK numbers also forces the picker to re-figure the RGB numbers as well. This has been either a feature or a bug for as long as I can remember.

Mike_Sc...@adobeforums.com

unread,
Mar 21, 2009, 5:29:37 PM3/21/09
to
Try the Color Picker with:

- RGB working space is ProPhotoRGB
- no document opened or
- active document is RGB mode AND in working space ProPhotoRGB.

The Color Picker is working in the working space / profile of the active document.

My primar workflow is in RGB with RGB output devices. For the above conditions: The mode change to CMYK is ok (blue) and the defined color in RGB is blue, too. It is just the wrong value of L=0 in the Color Picker, that makes me sceptical.

And if you like try this, too:

- Define R=0, G=0, B=255 while in ProPhoto working space
- Create a new document with this colr as a solid fill
- Do a mode change RGB --> Lab
- Review the L-channel (Cmnd-3)

My PS CS4 shows me a complete black L-channel, while the Lab-composite is in blue.

Ann_She...@adobeforums.com

unread,
Mar 22, 2009, 12:29:14 AM3/22/09
to
I think I see when it happens:

If I open a new file in ANY other color space (apart from Prophoto RGB) and then Convert to Prophoto before filling with Blue,
the L* channel does NOT register zero.

I opened a new AdobeRGB file and immediately converted to Prophoto RGB / CMYK web coated SWOP 2
Typed 00; 00; 255 in RGB in Picker which then showed L*=30; a=69; b=-114

Did an Edit menu/Convert to L*ab; and Lab Colors are still reading L=30; a=69; b=-114 in the info Palette while the L* Channel alone shows a mid-gray.

Tried it again with Image/Mode change to L*ab with the same results.

And it is the same in both 8- and 16- bits per Channel.
--------
BUT then I tried opening a new file in Prophoto RGB / CMYK web coated SWOP 2:

Did the same as above and THEN I got the strange L*= 0 reading;
I also got it if I converted that file directly to L*ab.

But if I convert directly from Prophoto to CMYK:
I get 100C; 80M; 1Y; 1K and a L*ab reading of L*=33; a=7; b=-51

So it seems to be a bug that you only see if you actually open the original file in ProPhoto. You probably should file a Bug Report if you haven't already done so?
---------
Using CS4 on a G5 with OSX 10.4.11.

Mike_Sc...@adobeforums.com

unread,
Mar 22, 2009, 6:05:43 AM3/22/09
to
I think I扉e got an explanation now:

ProPhotoRGB is exceeding the Lab color space, especially in the blue area. Since Photoshop is internal working in the Lab-Luv color space (and then translating to the RGB/CMYK working spaces you have choosen in the Color Settings) there are some portions of imaginative colors that can愒 be properly translated into Lab.
As I wrote in the opening message in this thread, it is more a kind of academic issue. In practice you should use ProPhoto RGB only with real world input capture (digital cameras, scanners). Their device color spaces are all within the ProPhoto RGB color space and real world colors are properly handled by internal Lab-Luv color calculations.

Ramón_G_Castañeda@adobeforums.com

unread,
Mar 22, 2009, 9:31:33 AM3/22/09
to

The gray shaded plot is ProPhoto RGB, the red outline is L*a*b.

Image © Bruce Lindbloom, from his site <http://www.brucelindbloom.com/index.html?WorkingSpaceInfo.html>.

0 new messages