Yeah, I’ve seen that too George. I created an alpha channel out of the selection to look at it, and even deleted, filled, and otherwise fiddled with the selection and its inverse. I don’t get it either.
Its what used to be called "the claw" it basically makes a selection based on a greyscale of the VISIBLE image.
If you move to the channels palette and control click on the RGB composite channel, this does exactly the same thing.
It has many uses for example you can use it to lift the highlights for hazing, or applying a curve adjustment to bring out shadow detail. After inverting the selection obviously
Try going into quick mask (Q) straight after if you want to Levels adjust the composite for more control
Kewl…the CLAW!
That’s got a nice ring to it. Kinda like "I’m a Photoshop Power User…I’ve got the CLAW!"
It’s an old ICQ habit I haven’t gotten rid of.
No George – its not feathering.
The marching ants are just a crude visual representation of your selection. If a selection is graduated the marching ants choose the midpoint. You have spotted it by noticing that 128 is the point in this case.
Use quick mask (your Q key) to see any selection represented as a channel. if you then go into the channels palette you will be able see the quickmask channel at the bottom, which you can view on its own by clicking the eyeball icons. You are now seeing a greyscale version of your selection.
This is basic to understanding how photoshop selections work. Channels and selections are interchangeable, within photoshop they are both held as 256 levels of greyscale or opacity.
VISIBLE pixels means that the claw is taking a composite or flat view of the visible image. If you turn some layers on or off, you will get a different selection when you use the claw. Its like a greyscale snapshot of the visible image.
Mark,
You are right about the feathering. Rather than being feathered, it appears that the selection comes with an opacity at each pixel site derived from the luminosity at that site. Thus, a luminosity of 128 would give an opacity of 50%, a luminosity of 192 an opacity of 75%, etc. While this seems to hold, the reason or purpose or usefulness of it remains obscure (to be truthful, I can’t conjure up any circumstance where I would invoke it).
Your "Claw" designator—I’ve never seen it before. Where did it come from? Is it supposed to be descriptive in some way?
George
It can come in really handy sometimes.
In fact, I’ve known about it for a while, I think it was YOU Jason who posted it oh-so-long ago. I just never knew it as the claw. I wasn’t sure what it did until this thread – that’s when I really started playing with it.
George/Tony:
(Note that I only read the first couple of posts, not the entire thread, so appologies if this is "out of place".)
If understand correctly, you’re asking "what is the point of the composite selection feature?"
Well, I’ve found it very handy for a few things. One example is: let’s say you take an digital photo that has too much specual highlighting, you can load the composite selection, feather (to your liking – or not) and then use a Curves to correct the image.
Another example is: load the composite selection, feather and then use a Curves to add a soft focus/glow to the entire image.
Thanks Trevor. I think that’s what I was saying in post #10. I know you didn’t read the whole thread, but if I understand you right, that’s kind of how I used it.
That’s what I love about this forum. I learn something new all the dern time.
Peace,
Tony
Trevor,
Great to have you chime in. I can’t seem to come up with the rationale that relates your stated results with the mechanics of the action—all pixels being partially selected with the degree of selection proportional to the luminosity, thus biasing the selection in favor of the most luminous and rendering the least luminous relatively inert.
George
George – all of these techniques are really about separating the highlights from the shadows of the image, and then performing your action on one or the other. If you invert the selection (control shifty I) it operates on the shadows, left alone and not inverted it operates on the highlights.
Trevors two techniques use the highlights creating an adjustment layer straight after the claw will automatically create a layer mask which can then be Levels adjusted to further refine the result. To see this luminosity layer mask on its own Press (contrrol \) followed by (control tilde~).
I’m not sure where the ‘claw’ term came from originaly, its been around for years.
Lets say your highlights have a yellow-ish hue, while the rest of the image looks OK. This could easily be a result from intense yellow lights.
(I just had a tuna salad image that suffered from this same exact problem)
Select composite, bring down the yellow in curves – will fix, without affecting the rest of the image.
I just had a tuna salad image that suffered from this same exact problem
Public Service Announcement: Don’t eat yellow tuna.
Let me try to summarize "the Claw" and, by all means correct me where I may go astray.
The claw performs a graded selection of the image in which, at every pixel site, the depth of selection (the opacity) is proportional to the luminosity (composite grayscale value). Thus, fully luminous pixels are fully opaque, black pixels are fully transparent.All other pixels are partially opaque, the opacity increasing with original luminosity. The selection’s marching ants appear at the 50% opacity level
Using the claw on an adjustment layer mutes the adjustment more and more as the opacity wanes. The more luminous the pixel, the more intense the adjustment. Or, to say the same thing, the less luminous the pixel, the weaker the adjustment’s effect.
By inverting the selection, the adjustment’s effectivity can be switched from the high-luminosity pixels to the low-luminosity pixels. Thus, highlights or shadows can be singled out for maximum effect of the adjustment, whatever it may be.
Let the flak fly !!
George
Well, I guess the first flak is from friendly fire…I have confirmed that inverting the claw’s selection does indeed invert the opacity gradation with the most luminous pixels becoming lowest in opacity and the least luminous highest in opacity. OOOOOHHHHH LA LA. I like it!
George
By "George", I think he’s got it ;-).
(I’ve read the whole thread now…interesting…)
Thanks for providing comfort, Trevor.
A smiley face would follow if I knew how to invoke it!
George
Actually, I guess the forum does it automatically now:
:- ) = 🙂
;- ) = 😉
etc.
Trevor,
Come to think of it, since the Forum automatically converts the colon/dash/parenthesis to the icon, how were you able to show BOTH the unconverted punctuation and the converted icon?
George
:- ) = 🙂
;- ) = 😉
Like this, you mean?
Dennis
Looks like you put a space before the parenthesis to kill the conversion on the left of the equal sign. I’ll try that below.
:- ) = 🙂
;- ) = 😉
EDIT; Yep!! 🙂 😉 🙂 🙂 😉
