This came up for me while working on an image in LAB so I'll use that. I copied the original and saved a channel as a selection into a new document. The image in the new document looks like it's been inverted. When I load that selection into the original file it goes back to looking as it did when I selected it from the copy. Can someone tell me why it looks inverted when I save it into the new document? I'm guessing that it has to do with the new document being in grayscale instead of LAB but it still doesn't make any sense to me.

One other quick question on another topic. I hear of a rubber stamp tool mentioned in books and it's also mentioned in the photoshop CS2 help file. However, I can't find it anywhere and it doesn't show it's existence in the photoshop help file although the name of the tool is mentioned in other places. Am I seeing things or is this another name for a tool like the clone tool.

These have to go down as the dumbest two questions every asked on this forum so get a good laugh out of it while you can.

Hi Isaac,

This came up for me while working on an image in LAB so I'll use that. I copied the original and saved a channel as a selection into a new document. The image in the new document looks like it's been inverted. When I load that selection into the original file it goes back to looking as it did when I selected it from the copy. Can someone tell me why it looks inverted when I save it into the new document? I'm guessing that it has to do with the new document being in grayscale instead of LAB but it still doesn't make any sense to me.1) 1) whenever you load a channel as a selection (Ctrl+click on it), you actually load the luminosity of that channel only, meaning, the white/lightest parts of your channel will be 'strongly' selected...while the strenght of your selection will fade with the darker gradation of grey to no selection at all in the black/darkest parts of your channel.

2) When you save that (luminosity) selection to a new document, the look of it will be strongly influenced by the 'background colour' of your new document (default white) .... hoewer light a grey might be, it is always darker than white, thus, the optical illusion of inversion.... (You could try to change the 'background colour' of your new document to black or transparent and you will see the difference!!)

3) If you want to copy and see your whole/unchanged channel to a new document, you should select its canvas (Ctrl+A), copy it (Ctrl+C) and paste in into the new document without worrying about its background colour.

BTW, it works exactly the same in RGB.


One other quick question on another topic. I hear of a rubber stamp tool mentioned in books and it's also mentioned in the photoshop CS2 help file. However, I can't find it anywhere and it doesn't show it's existence in the photoshop help file although the name of the tool is mentioned in other places. Am I seeing things or is this another name for a tool like the clone tool. As far as I know, it is exactly just another name for the Clone Tool.

These have to go down as the dumbest two questions every asked on this forum so get a good laugh out of it while you can. No dumb questions at all!!! :happy: :happy:

Flora,

I respect you a great deal. You really seem to know your stuff. Add to that the fact that your nice and it's a complete package.

Regarding the saving of a selection. The problem that makes this a little more complex that in say, RGB, is the fact that I am doing it in LAB. I am focusing on white spots in order to remove them from an image. The image has got a yellow cast so I am using the B channel. In the B channel the white spots are actually darker than the yellow of the image. When I save the channel as a selection and, as you say, it saves the luminosity of the channel, it does in fact, invert. The white is now lighter and the yellow darker. This would not happen in an RGB or CMYK image if I have figured this out correctly. I didn't check it. My problem didn't have anything to do with the background color.

Now, why it does it this way is another question all together which I don't need answered. I'm sure it is valuable to be able to save the luminosity like this when you get the original look once you stick it back into the colorspace you were working with and that is why you have the option to save it into a new document.

Regarding the Rubber Stamp Tool. Are you sure of this? You seemed iffy about it. I have heard of this Rubber Stamp Tool before as I said, but haven't actually seen it. Why would they call the Clone Tool the Rubber Stamp Tool? Especially a book that is in Photoshop CS2? Do you know if that is what the Clone Stamp was called at some point in it's history? I'm kind of interested in this even though it is a rather unimportant issue. I especially don't understand why the Photoshop help file would bring it up if it is no where to be found within the program.

I'd love to hear your response to my logic with the saving of a selection. Of course, I'd like to hear if you have anything more for me in the Rubber Stamp area. I'm surprised more people have not responded.

Hi Isaac,

...me again...

thank you so much for your kind words!!! :pleased: :blush:

I wasn't 100% sure about it, but now I am... I found out that until Photoshop 5, or 5.5, the Clone Tool was actually called the Rubber Stamp.... So, those who started with Photoshop way, way back, probably, keep on calling it that.. :happy:

Here are a couple of links:

From http://www.lonestardigital.com/rubber_stamp.htm

Adobe Photoshop's Rubber Stamp Tool
(Also known as the Clone Tool)

The Rubber Stamp tool is a sophisticated copy & paste operation. Simply put, it allows you to copy an existing part of a picture and then paste it on top of another area. The Rubber Stamp does its job without looking like an unsightly band-aid patch. The Rubber Stamp tool is also called the Clone Tool. Removing Unwanted Elements with Photoshop's Rubber Stamp / Clone Tool (http://www.betterphoto.com/digital/cloning.asp)

As for the save as a selection, after reading your first post, I actually made my experiments in LAB mode (just like you had explained) ... and repeated them in RGB with the same result ...
Maybe, if you posted a cropped section of the problem area in the picture you are working on, I could try it...

That's what I thought with the Rubber Stamp tool. At least that's the only thing I could come up with. I used the Clone Stamp "instead" and got pretty similar results as to what was being described.

As far as the saving selection deal goes, I don't need any help with the actual image, I was just interested in why I was getting what appeared to be an inverse of the image when I copied a channel. This just came to mind but if you were to do the same thing to an image with a strong blue color cast then your saved selection would appear correct. I just had this idea that I would get the same image that the channel was when in fact you are getting a luminosity mask as you said.

If you want to see the image I am dealing with and how it is explained then just go to Margulis' LAB book. I am going to assume you have it. It starts on Pg. 226. If not then let me know and I'll post it if you really want to see it.

BTW, I really not this lame when it comes to Photoshop.

Hi Isaac, perhaps I can explain your photoshop optical illusion. On first glance the image looks like it is inverted but may not look quite right to you. If you run your mouse over the image you will see that the If you run your mouse over the image and examine the info palette you will notice that the white parts of your saved selection are still white but the black parts (or dark parts) are inverted either mathematically or close to it. So now you are really scratching your head, right? OK, if now turn off your white background layer leaving only your pasted selection visible. The white areas will still look white, the dark areas will be almost invisible or faint gray. Now if you run your mouse over the image and view the info palette, you will see that the white areas have high values but low and behold the dark areas have low values (close to black) but you can not see much. And the answer to the mystery is that when you selected the luminosity, and pasted it in, the black areas maintained thier color but were reduced significantly in Opacity. Pasting onto a white background causes a blend which shows up as white because the opacity of the background significantly overpowers the pasted layer where the opacity of the dark areas is low. You can verify this by creating a new blank document w/ a white background and a black layer and adj the opacity, or just play with the image you were working on.
Regards, Murray

This is just getting more confusing with every post. :dizzy:

If you have the Margulis LAB book then just go with that image cause that is exactly where the question came up with me. The image has a heavy yellow cast. So, when I Ctrl-Alt-3 to select the b channel and then go to Select:Save Selection and choose New for the Document to save the selection into. I'm not pasting anything manually.

I then take the original document that I selected the b channel from and put it next to the new document with that b channel in it. What I get is an exact inverse. The areas of the image in the original which Margulis says are white show up being 44% in K or about 67 in L. The A and B are both 0 obviously.

In the New document created with the selected B channel of the original I get 56% in K or about 56 in the L. Now this shows a pretty absolute inverse if I am not mistaken. It tells me that the Mid of the L channel is actually 61.5 and not the 60 that I believed it to be but I'm not sure that's important because the K wraps around 50% perfectly. No matter where I go with the mouse, I am getting different shades of gray on the info palette. It does seem to get worse for lights than darks or visa versa. There are no blacks or whites as we are dealing with an LAB channel and colors that fit in a gamut.

That's the best I think I can describe what my situation is without you actually looking at the image yourself. Let's pray that someone has the LAB book. Oh, and BTW, like I said if I were to load that new doc into a LAB image, it would show up just as the original had before I saved the selection into a new document. Also, if I were just to copy the channel it would copy fine within the same document.

Isaac, I also have the book and I made the same selection of Great Grandma, however in my case the new grayscale document is exactly matching the B channel of the originlal image. There is no inversion. And I get the same result before and after auto-leveling the B channel. I have no idea why your new document is being inverted. Am also using CS2.
Regards, Murray

Well, this is odd then. I have set my LAB settings according to the book so everything should be kosher. I will post my B channel and new document in a little bit so you can see the difference. This is very very odd to me although Flora explanation of the luminosity of the selection made sense. Thanks a lot.

BTW, if you know anyone here that you think may know what is causing me problems then send them over here to take a look.

Thanks again Murray and Flora
Isaac

Isaac, I have experimented with a number of preferences and settings and other scenarios with the image in question and can not duplicate your inversion. Assuming that you have not done anything to invert the selection prior to saving, then there is only one last thing I can suggest to you.
I have found over time that certain events can alter parameters of Photoshop's Preferences file. When that happens, the very strangest things tend to happen and there appears to be no trend or explanation. So as a last resort, you may want to reset those preferences by holding down the Ctrl+Shift+Alt keys down at the same time while restarting Photoshop. A dialog box will appear asking you if you want to reset the preferences to the default factory settings. PLS NOTE: Prior to doing this, you should save your Workspace settings so that you do not need to spend the 5 minutes changing your File Preferences and layout of the Palettes etc if you have moved them around.
Regards, Murray

Here are the two things I get. One is the b channel and one is the saved selection of the b channel into a new document.

Hi Isaac,

As far as the saving selection deal goes, I don't need any help with the actual imageI hadn't asked because I thought you needed help with the picture ... I simply wanted to replicate the procedure you were following to see if I had the same problem ... for this reason I thought it would be better if we were experimenting on the same picture ... :happy:

If you want to see the image I am dealing with and how it is explained then just go to Margulis' LAB book. I am going to assume you have it. No, I don't have it ...

Murray,

I made the same selection of Great Grandma, however in my case the new grayscale document is exactly matching the B channel of the originlal image. There is no inversion.Just curious... How did you select the B channel? ...

Was it:

1) by Ctrl+A select canvas? (this way you select all the pixels in the image and this selection will always look exactly as the original indipendently of the colour of the background of the document or Layer it will be pasted on)


2) by Ctrl+click on it or by clicking on the 'Load Channel as a Selection' button at the bottom of the channels Palette?... (this is the way that could give you the 'inverted look' if pasted on a 'New document' with a white background ... Could I, please, ask you to try this? ... Could you then change the colour of the background of the 'New document' to black and to transparent and paste the same selection on them?

I'd really like to know if your results will also 'look' very different from each other .... but still, as you already said, having exactly the same values since you never changed anything in your selection.

Thank you in advance.. :happy:

Hi Flora,
Firstly, it is worth noting that everything in the B channel is above the midpoint of its range (no Blue). In a LAB image the mid point for Luminosity selection is 50 (B Value = 0 to +128). When I select the luminosity of the B Channel with any of the following:
- Ctrl + Alt + 3
- Crtl + Click on B
- Load Layer as selection button (with B selected),
it appears as if the whole channel is being selected which it is except at reduced opacity proportional to the luminance. As you point out, this is different from selecting B and doing Ctrl + A.

Now with the B Channel still active I copy the selection and paste it inot a new document with a white background, it will look almost all white but it is not. If you look carefully, you will see a reduced opacity version of the B channel which is the Luminosity. Because the background is white, it is difficult to see the pattern but it is there. In fact if you turn off the white background layer and place an eyedropper target on her hat and her hand, and look at the Lightness Value values in the info window, you will see they are exactly the same as in the B channel (about 72 on the hat and about 67 on the hand).
If you paste the selection onto a black background, the image gets darker but it is not inverted. Moreover if you turn off the black background, and view the eyedropper tagets, they have the same values as in the B channel.
The same is true for pasting the selection onto a transparent background where it will be really difficult for you to see the image but it is there.

Using the method Dan uses in his book - Save Selection as New Document creates a multichannel document which basically pastes the luminosity selection onto a 50% gray background causing it to look exactly like the original B channel.

So far, there are no inversions anywhere. The images that Isaac uploaded are definitely inverses of each other. If you apply them in subtract mode, the result is black. I have not been able to duplicate his inversion.

Best Regards, Murray

Flora,

Sorry about that. Here is the image untouched other than it was switched to RGB so that I could get it to you. Did you see what I posted with my results?

Murray,

I'm doing it verbatim to what Dan describes in the book. (Yes, we're on a first name basis) I do the control+alt+3 and then save as new selection and what I attached to my post is what I get. The funny thing is this is just the new document. If I move the image in that new document to a the original image then it turns back into the b channel I wanted to see in the new document. So, for some reason it inverses it's way into a new doc but remembers its original values for when it is placed back into a LAB image. I'm clueless.

Isaac, after you have selected the luminosity, and you Select > Save Selection, there are 2 choices in the Pull-Down Menu of the dialog box. If you leave the default and you save the selection as a new channel, then your image will be inverted and saves as a new channel in the same image that you are working on. If you select New Document from the pull down menu, then Photoshop will create a new image with a singe layer (mode will be Multichannel)and it will look identical to the the B channel and it will not be inverted. So are you actually saving the selection to a channel or to a new document? Or you are not doing a Copy and Paste of the selection?
Regards, Murray

Hi,

Murray,

thank you so much for taking the time to experiment with the copy+paste I thought was at the origin of Isaac's problem! :pleased: ... Needles to say, your results were exactly the same as mine!

Isaac,

thank you for the picture!

I've been trying to find something on the Internet, particularly on the Adobe site and to reproduce your definite 'inversion' in this couple of days... but like Murray, I couldn't duplicate it by simply saving the selection as a New Document.

Attachment 1 show the path I followed to save the selection as a New Document.

Attachment 2 shows the New multichannel Document. (like for Murray, identical to the LAB B channel).

Attachment 3 ... the only way I was able to duplicate your inversion was by inverting Ctrl+I either the B Channel immediately, or the New Document after saving...