Retouching in LAB with the Clone Stamp Tool works great. But painting with the Brush Tool doesn’t work. When picking up a color with the Eye Dropper and painting the color is too bright in all 3 Channels. Is there any way to get the correct color (formula)?
Well I’m talking about only working on one channel in LAB, no layers. Have you tried it lately or do you just assume it should work? I’m using CS3. When I sample a color (even if the area is flat and clean and the pixel radius of the picker is 1×1) and paint, it becomes lighter. Has always been like that for me in LAB.
"Lab Mode, Gray Working Space & Painting in Lab Channels:
Photoshop uses the Gray Working Space found in Color Settings to display single channels in full colour mode images (not in the composite channel gamma as some users expect). There appears to be a bug when working in Lab mode where painting in individual channels with the brush tool results in tones that are lighter or darker than the sampled tone (the clone tool is not affected). Simply using Grayscale Gamma 2.2 in the Gray Working Space will correct this issue. Another work around that does not involve changing the Gray Working Space is to convert the Lab document to Multichannel mode, editing the separate channels and then returning to Lab mode."
Ok now I get you, painting on A and B CHANNELS in LAB does seem to create this behaviour, taking a value about 10% higher than it should. Seems to be still Ok on the L channel though. Stephen has given you some great help there.
Forgot to ask, are there any pixel changes when going from LAB to Multichannel and/or back (such from RGB to LAB)? I know they make practically no differnce, just curious. Thanks everybody!
To the best of my knowledge, going to Multichannel mode does not involve a conversion as in the case of one flavour of RGB to RGB or RGB to Lab or CMYK. The channels from the original should just be "split" into separate channels with no composite preview channel as found in the previously mentioned standard colour modes. I would be surprised if there was rounding errors as there is no translation/conversion taking place when using the image/mode command to go in and out of Multichannel mode.
I only mentioned Multichannel mode as it is another option to overcome this bug, if you don’t wish to use gray 2.2 working space in colour settings.
Off topic for the OP, a similar thing if one simply targets a single channel in a colour file and then invokes the mode/grayscale command. There should be no conversion, the single channel exactly the same as the source should result.
P.S. Sorry for not making it clear that it was the two colour channels that are affected by this bug and not the Lightness channel.
I’m using PsCS3, Mac OS 10.4.11, some custom monitor profile, that someone else made, and it does occur in every channel of LAB. And of course the paint mode is normal etc.
What BinaryFX/Stephen writes is correct, as far as going to Multichannel mode.
Any misunderstanding here was due to my misconstruing of the ambiguous phrasing of Peter’s post re: are there any pixel changes when going from LAB to Multichannel and/or back (such from RGB to LAB)?
It has been years since I was first made aware of this issue and tried to find a solution, back then it was only the colour channels being edited so I did not check the L channel, I will have to test that, thanks!
I would simply use Gray 2.2 gamma in grayscale working space colour settings to overcome this Lab channel issue (or clone stamp), unless this messed up your workflow (then I would use MC mode or perhaps try to remember to change in/out of gray 2.2 gamma for these edits while keeping the original working space setting for all other work).
Thanks, B. Great site by the way. I have "15% Dot Gain", not Gamma 1.8 or 2.2 for Grayscale. It’s the standard in Germany. Not sure why or what difference it makes.
