The middle gray inside Quick Mask Mode seems to depend on the document color space but for what reason?
yes, it does.
something has to determine values. A grayscale file is managed the same way as a mask.
Should it?
no
Is it fixable.
not sure.
Ask CCox.
Tom,
The hue and tone of a QM dot have no relevance other than demarking the masked area. The only reason for adjusting the QM hue and tone is to enhance the user’s visibility of the selected area, especially when the image’s color is too similar to be easily distinguished from the mask. If the mask’s tone were to indicate degree of selection, you would then forfeit the ability to use tone for differentiating between masked and unmasked areas—indeed a most undesirable deficiency. Besides, you know the degree of selection from the foreground gray level, which you set on the toolbar.
George
thomas, if it doesn’t give 50% selected, what does it correspond to?
George,
Are you saying that the foreground color (or tone) you choose to paint with inside QM doesn’t have any influence on the degree of selection when you leave QM?
Pierre,
If my document space is Adobe RGB then R:147, G:147, B:147 seems to be the middle gray because if I paint with a brush set to R:147, G:147, B:147, I can see the selection edges when I leave QM. If I set the foreground color to R:148, G:148, B:148, the selection edges isn’t visible when I leave QM. It must mean that R:147, G:147, B:147 is the middle gray because Photoshop doesn’t show the selection edges on selections, where all pixels are less than 50 % selected.
If the document space is Apple RGB for instance, the middle gray seems to move down to R:130, G:130, B:130. (RGB:130 gives visible selection edges, RGB:131 doesn’t).
Now what do I do if I want to make a selection with QM where every pixel is exactly 50 % selected ? Not even the opacity slider for the brush tool seems to do what I want (50 % opacity doesn’t seem to mean 50 % selected).
—
Regards
Madsen
Tom,
"…Are you saying that the foreground color (or tone) you choose to paint with inside QM doesn’t have any influence on the degree of selection when you leave QM…"
No. Just the opposite. The foreground tone you set AFTER invoking QM determines the degree of selection when you revert to normal mode.
George
George,
No. Just the opposite. The foreground tone you set AFTER invoking QM determines the degree of selection when you revert to normal mode.
Okay and that’s what I’m talking about. I can’t understand why a foreground color set to R:128, G:128, B:128 after invoking QM doesn’t give me a selection with 50 % selected pixels, when I revert to normal (or standard) mode.
—
Regards
Madsen
Not even the opacity slider for the brush tool seems to do what I want (50 % opacity doesn’t seem to mean 50 %
selected).
That is wrong because 50 % opacity does indeed mean 50 % selected in any color space. I forgot to set the foreground color to black when I tested the opacity thing. D’OH!
Well, I’ll use the opacity setting instead.
—
Regards
Madsen
Okay and that’s what I’m talking about. I can’t understand why a foreground color set to R:128, G:128, B:128 after invoking QM
doesn’t give me a selection with 50 % selected pixels, when I revert to normal (or standard) mode.
BECAUSE…….
THE ASSIGNED COLOR SPACE AFFECTS THE BRUSH IN MASKS!
When dealing with RGB files, with masks, and you punch in equal amounts of R_G_B such as 128 / 128 / 128, you "think" you are creating a 50% value in a mask.
You are if your opacity is set to 100%, and the color space in which describes your masks is not a twisted train wreck along with the color engine selection.
If you set your opacity to 50% and use what you "think" is a color designation of , 192 / 192 / 192, you’d think that you would get 25%
RIGHT?
Logical assumption?
NO.
again, the assigned color space affects the masks.
This is just another reason why prepress people refuse to work in RGB……………….Chris.
Thomas needs to come to terms with this.
We need device N color designations, for color space conversions, for non color managed objects, as well as masking.
BECAUSE…….
THE ASSIGNED COLOR SPACE AFFECTS THE BRUSH IN MASKS!
Let me rephrase. I can understand that it depends on the color space but I can’t understand why it is so. It shouldn’t be IMHO.
When dealing with RGB files, with masks, and you punch in equal amounts of R_G_B such as 128 / 128 / 128, you "think" you are creating a 50% value in a mask.
Yes.
You are if your opacity is set to 100%, and the color space in which describes your masks is not a twisted train wreck along with the color engine selection.
If you look at it that way, then Adobe RGB must be a twisted train wreck.
If you set your opacity to 50% and use what you "think" is a color designation of , 192 / 192 / 192, you’d think that you would get 25%
RIGHT?
Right.
Logical assumption?
NO.
I think it’s a logical assumption but it confuses me that it doesn’t work that way in masks.
This is just another reason why prepress people refuse to work in RGB……………….Chris.
So you’re saying that it works in CMYK or what do you mean by that?
We need device N color designations, for color space
conversions, for non color managed objects, as well as
masking.
Does anyone know where I can learn more about this Device N thing? (preferably a simple explanation). I’ve heard of it many times but can’t understand what it means and what it’s good for.
—
Regards
Madsen
"…the assigned color space affects the masks…This is just another reason why prepress people refuse to work in RGB………Chris"
Mike,
You present this apparently as a quote from someone named Chris. The only Chris we all know is Chris Cox. Are you quoting him?
I don’t see RGB as the culprit here. Wouldn’t you have the same scenario in any other device-dependent space (like ?
George
I’m not quoting Chris Cox, I’m telling him what needs to be changed.
geez.
OK, Mike, I see that you were addressing Chris, not quoting him. But you left unanswered what it is about RGB that would not also apply if CMYK were used in this situation. Not meaning to badger you—just trying to understand. 🙂
George
The issue here, is that the working space, "Affects" masking in RGB or CMYK.
It shouldn’t.
From an engineering perspective, I understand why it behaves as such.
From a user point of view, it’s a kludge.
Dunno, Mike. I thought Tom brought closure to this (now phantom) problem, conceding pilot error rather than a quirk in PS or a color space limitation. Tom simply neglected to control BOTH foreground AND opacity when forming his quick mask, and when he corrected that oversight the selection process upon exiting QM worked as expected.
George
it’s a well documneted issue that the color space affects the masking.
"…it’s a well documneted issue that the color space affects the masking…"
1. If it’s so well documented, why is it obscure enough to escape most people’s attention, mine included?
2. If there is an effect, I’ll bet it is minor and nowhere near the magnitude that Tom thought he was experiencing. At this point, in fact, it would be the magnitude that needs to be shown, not the mere existence of a perhaps secondary effect.
George
Hmmm. Assuming anybody out there is still interested in this issue, I see now that it has not been put to bed.
WHEN USING A LAPTOP with LCD display, Adobe 1998 color space, and 100% brush opacity,I have obtained the same apparent 50% selection point(147/148) that Tom reported. That is, upon returning from QM mode to normal mode, the marching ants appear around areas masked at a luminosity of 147 or under and not around those masked at 148 or above (with "Color Indicates" set to "masked areas")
With a DESKTOP computer, also with Adobe 1998 and 100% brush opacity but with CRT DISPLAY, the 50% selection point is 127/128—just where one would expect it to be.
This finding re-opens the question for me. Is it attributable to LCD/CRT differences or to some other color setting differences I have not yet nailed down?—-Intriguing—.
Tom, were you using an LCD display?
George
Is it attributable to LCD/CRT differences or to some other color setting differences I have not yet nailed down?<
only if your image is assigned a monitor profile.
Mike,
Sorry, I’m not up to snuff on color management. How do I determine if my image has been assigned a monitor profile?
I can say that for the image I used in testing, I did not consciously assign any profile, but that does not preclude a possible automatic assignment.
George
In the picture window, bottom center, left-ish side, there is a sideways triangle. Click on it and select document profile. This tells you what profile is being assigned to the image at that time.
If is says "untagged RGB", than the working space in the color settings is determining the color space.
George wrote:
Tom, were you using an LCD display?
Yes my primary display is an LCD but the middle gray for QM, channels or layer masks is still RGB:147 when I drag Photoshop over on my secondary CRT.
I have even tried to disable my LCD and attached the CRT to output one of the graphic card instead of the LCD (and changed the monitor profile to the profile describing the CRT because that’s not happening automatically) but RGB:147 is still the middle point. I would be even more confused if it wasn’t the same on the LCD and the CRT.
Are you absolutely sure that your document space was set to Adobe RGB when you tested it on your LCD? It sounds like you were using your monitor profile as the document space.
I have used Adobe RGB as the document space both on the LCD and the CRT.
I can see that if I assign my monitor profile to the document, the middle gray for masks move up to RGB:148, meaning that the selection edges is visible at RGB:148 but not at RGB:149. That’s only one number higher than Adobe RGB but maybe your monitor profile on your laptop is completely different than mine. Maybe it’s a standard canned sRGB monitor profile where the middle point is around 128?
—
Regards
Madsen
It sounds like you were using your monitor profile as the document space.
Or simply that the document space and/or RGB working space on your laptop was set to sRGB which is the default setting, as far as I remember.
If it’s an untagged document, the document color space is the same as your working space. As Mike says, you can see whether the document is tagged or untagged in the lower left corner.
—
Regards
Madsen
Thank you for that, Mike. It reads "untagged" and my working space is Adobe 1998.
I have been barking up the wrong tree in looking for CRT/LCD behavior differences.
I now can associate the 147/148 luminosity mask giving 50% selection with using CS with one set of color settings and the 127/128 luminosity mask with using PS 7 with a different set of color settings—all on the same machine with the same CRT display. That seems to eliminate the LCD as the variant and points to either PSCS vs PS7 or the color setting differences.
The color settings which differ are;
Gray: Gray Gamma 2.2 with PS7; Dot Gain 20% with PSCS
Color Management Policies: RGB, CMYK, and Gray all off in PS7; RGB and CMYK set to "Preserve Embedded Profile" and Gray off in PSCS
Rendering Intent: Relative Colorimetric in PS7; Perceptual in PSCS
I guess my next step should be to set the PSCS color settings to match the settings used in PS7 and then vice versa. That should either eliminate or identify PSCS/PS7 differences as the behavioral swinger.
Thank you again for the help, Mike.
George
Intending to change just one color setting at a time until I found one that changed the 50% selection point, I lucked out.
The very first change I tried did it—I changed GRAY from "Gray Gamma 2.2" to "Dot Gain 20%" in My PS7 settings, to match the PSCS GRAY setting.
Then I exited PS, rebooted to be sure, and restarted PS7. With this one setting change, the 50% point moved to 147/148 from where it had been (127/128).
Having (maybe) resolved that question, now about the zillion more questions generated? Like, Wha happened???? 🙂
George
Tom
"…Are you absolutely sure that your document space was set to Adobe RGB…"
Yes, but since the GRAY color setting now seems to be at the root of the phenomenon, I’ve lost incentive to check the effect of every other setting. (I trimmed the hedges in front of my house instead).
Since mid-gray at 127/128 makes more sense and predicting resultant colors becomes quite precise in that case, I want now to know if there is any good reason not to stick with the GRAY setting of Gray Gamma 2.2, and I’d also like to know just what that setting and its alternative settings mean.
George
George,
It seems you’ve hit the spot.
Adobe RGB has a gamma value of 2.2 and therefore one has to set the gray gamma working space to 2.2 also to get a mask where the middle is 127/128. It also seems quite logical if one thinks about it because a mask is just an 8-bit grayscale channel.
My gray working space was originally set to a dot gain of 20 % and the result was a middle point at around 147/148 when working in a gamma 2.2 color space. A dot gain of 15 % moves it up to 158/159 and a dot gain of 30 % moves it down to 122/123 and I see the same in both Photoshop 6, 7 and CS.
By setting the gray working space to a gamma of 2.2 you’ll have the middle at 127/128 in document spaces with a gamma of 2.2. Both sRGB and Adobe RGB(1998) has a gamma of 2.2, as far as I know, but a color space as Apple RGB has a gamma of 1.8 so with a document in Apple RGB and the gray working space set to Gray Gamma 2.2, you’ll have the midpoint at 109/110. If you set the gray working space to Gray Gamma 1.8, the midpoint is moved up to 127/128.
I’ve now changed my gray working space to a gamma of 2.2 too (because I mostly work in gamma 2.2 spaces) and saved it as a new color setting. I’m not sure whether there are any pitfalls by using that setting for the gray working space though. I practically never work with gray scale images so I don’t think so in my case. I work in RGB and CMYK.
By the way, if you haven’t noticed. You can save your color settings with the Save button inside the color settings dialog box of either Photoshop 7 or CS. Give the color setting a name and save it in the folder that Photoshop defaults to (Program Files\Common Files\Adobe\Color\Settings in my case). Now you can load that color setting in most of your Adobe programs like other installed versions of Photoshop, InDesign, Illustrator and Acrobat. That’s a great feature.
—
Regards
Madsen
Tom,
Thanks for elaborating. Setting GRAY to Gamma 2.2 is making sense. Makes one ask when, if ever, it would be preferable to use a dot gain size setting in a gamma 2.2 system. Ah—so much to learn!!
George
George,
Makes one ask when, if ever, it would be preferable to use a dot gain size setting in a gamma 2.2 system.
When you’re preparing grayscale images for press, the dot gain matters. You can read more about it here:
<
http://www.imagingrevue.com/articles/?article=000291>.
I don’t do that much and if I do, I can always change the gray working space or even better just use Convert to Profile inside Image > Mode where it’s possible to choose between all the different dot gains, no matter what the gray working space is set to.
Ah—so much to learn!!
Oh yes. There’s learning material for a lifetime (and probably much longer) in that program. 🙂
—
Regards
Madsen
Guys,
The reason why it behaves as such is how that particular color,
ie: for example. RGB = 164 / 135 / 29, olive / brown color, in Greyscale, maps to a value that may or may not be relative to a preconceived logical value. (rendering intent and engines included)
Everything has to reference a color space, even masks. The issue here is that a working space SHOULD NOT affect masking, but it does.
I’ve beat myself into a froth over this a few years back.
