Layers and adjustment layers

After reading quite a few PS books, I still have a mental block about the difference between working in layers and in adjustment layers. Would appreciate some explanations such as:

- What are all the operations that can only be done in layers but not in adjustment layers? And vice versa?

- What are all the operations that should only be done in layers, but not in adjustment layers? And vice versa? And why?

- Should layers and/or adjustment layers be merged before applying the clone tool, history brush, healing brush, etc. to remove dust and scratches?

- Is it possible to applying sharpening to the L channel without merging the layers first?

Thanks.

wrote:

After reading quite a few PS books, I still have a mental block about the difference between working in layers and in adjustment layers. Would appreciate some explanations such as:

- What are all the operations that can only be done in layers but not in adjustment layers? And vice versa?

You haven't understood anything about Adjustment layers. Layers are like transparent sheets, that contain image elements. That way, you can add a picture of say a bird to a landscape picture. The bird is on a separate layer, so you can still move it around and work on it without touching the background picture of the landscape.

Adjustment layers are layers that do not contain image elements, but adjustment commands like 'Levels' or 'Curves'.

- What are all the operations that should only be done in layers, but not in adjustment layers? And vice versa? And why?

See above.

- Should layers and/or adjustment layers be merged before applying the clone tool, history brush, healing brush, etc. to remove dust and scratches?

That depends on what to want to achieve. Filters and tools only work on one layer at a time (the selected layer). If you want your filter or tool to act on everything together, you'll have to flatten first.

- Is it possible to applying sharpening to the L channel without merging the layers first?

Yes, but you will only sharpen the L-channel of that one layer.

--
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

Adjustment layers are just for: color change, like hue contrast Layers are use for all kind of thing. It is hard to explain at least u see it in action

I use layer and mask all the time. Threat layers as from bottom up like transparency stock of photograph to make one solid image.

You don’t have to merge anything. The only time you can do this is when you are sure 100% that the image you did doesn’t have to be change anymore.

If you need specific help I’m doing web site. Please post question to my forum and I will graphically show how to do it. Visit: www.ujstudio.com

- What are all the operations that can only be done in layers but not in adjustment layers? And vice versa?

A layer contains a picture--for example, you may build an image of people at the beach by taking a picture of the beach, putting it on one layer, taking a picture of your dad, putting it on another layer, and so on.

And adjustment layer does not contain a picture. An adjustment layer is a COMMAND: "Make everything below this adjustment layer darker," "make everything below this adjustment layer blue," and so on.

--
Biohazard? Radiation hazard? SO last-century.
Nanohazard T-shirts now available! http://www.villaintees.com Art, literature, shareware, polyamory, kink, and more:
http://www.xeromag.com/franklin.html

Layers stymied me for a long time when I started with PhotoShop. I finally had a white light experience in the middle of the night and came up with some physical analogies.

Tlhe simplest explanation, I think, is the idea that an Adjustment Layer acts just like putting a filter in front of your camera lens. You have color filters, contrast filters, neutral density filters, graduated filters, etc. They generally have an overall effect on the entire film frame.

Layers have images painted on clear, glass sheets that you place between yourself and the image you are viewing. The images on the layers can vary in density so they may entirely block what is on the image immediately below them, or they may be more faint so you partially see through them to the image below. When you finally flatten the layers, the upper layers are dropped on the bottom (original) layer and are merged into and become part of the bottom image.

The big boys may object to this analogy but it helped get me over the hurdle.

Good luck! . . . . patrick

"Tacit" wrote in message

- What are all the operations that can only be done in layers but not in adjustment layers? And vice versa?

A layer contains a picture--for example, you may build an image of people

at

the beach by taking a picture of the beach, putting it on one layer,

taking a

picture of your dad, putting it on another layer, and so on.
And adjustment layer does not contain a picture. An adjustment layer is a COMMAND: "Make everything below this adjustment layer darker," "make

everything

below this adjustment layer blue," and so on.

--
Biohazard? Radiation hazard? SO last-century.
Nanohazard T-shirts now available! http://www.villaintees.com Art, literature, shareware, polyamory, kink, and more:
http://www.xeromag.com/franklin.html

From:

- Is it possible to applying sharpening to the L channel without merging the layers first?

If you don't want to merge (and usually you don't) then just go to the top of the layer stack and make a new blank layer, then hold down the Alt key (PC) and keep it down while doing Layer -> Merge Visible and you'll get a new composite layer on top with all the underlying layer structure unchanged below. Now you can sharpen the L channel on this composite layer (or run USM and do Edit > Fade and change the mode to Luminosity to do something similar without the mode changes).

Bill

patrick wrote:

Layers stymied me for a long time when I started with PhotoShop. I finally had a white light experience in the middle of the night and came up with some physical analogies.

Tlhe simplest explanation, I think, is the idea that an Adjustment Layer acts just like putting a filter in front of your camera lens. You have color filters, contrast filters, neutral density filters, graduated filters, etc. They generally have an overall effect on the entire film frame.
Layers have images painted on clear, glass sheets that you place between yourself and the image you are viewing. The images on the layers can vary in density so they may entirely block what is on the image immediately below them, or they may be more faint so you partially see through them to the image below. When you finally flatten the layers, the upper layers are dropped on the bottom (original) layer and are merged into and become part of the bottom image.

The big boys may object to this analogy but it helped get me over the hurdle.

Good luck! . . . . patrick

This is a great explanation, especially for someone with a knowledge of traditional photography. Sometimes the big boys are unable to use analogy to simplify an explanation.

One minor point though. Instead of describing layers as "clear, glass sheets", I would describe them as "sheets with a variable opacity". A layer with 100% opacity will completely block the layers below it. A layer with <100% opacity will let all the layers below it to show through partially. The adjustment layers also have a variable opacity.

One mistake I kept making was not associating adjustment layers with separate layers.

Bill Hilton wrote:

From:

- Is it possible to applying sharpening to the L channel without merging the layers first?

If you don't want to merge (and usually you don't) then just go to the top of the layer stack and make a new blank layer, then hold down the Alt key (PC) and keep it down while doing Layer -> Merge Visible and you'll get a new composite layer on top with all the underlying layer structure unchanged below. Now you can sharpen the L channel on this composite layer (or run USM and do Edit > Fade and change the mode to Luminosity to do something similar without the mode changes).

Bill

This will probably work if the image is already in LAB mode. But I work in RGB mode and then convert to LAB mode before sharpening the L channel. When I convert to LAB mode, there is a message asking if I want to merge the layers first (or something like that). Perhaps my question should be: is it possible to convert modes without flattening the layers?

wrote:

This will probably work if the image is already in LAB mode. But I work in RGB mode and then convert to LAB mode before sharpening the L channel. When I convert to LAB mode, there is a message asking if I want to merge the layers first (or something like that). Perhaps my question should be: is it possible to convert modes without flattening the layers?

You get a question if you want to flatten the layers first. Guess what happens if you say "Don't flatten"...

--
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

If you don't want to merge (and usually you don't) then just go to the top of the layer stack and make a new blank layer, then hold down the Alt key (PC) and keep it down while doing Layer -> Merge Visible and you'll get a new composite layer on top ...

From:

This will probably work if the image is already in LAB mode. But I work in RGB mode and then convert to LAB mode before sharpening the L channel.

Don't convert the entire file, just copy the top composite layer into a new file and convert *it* to LAB.

When I convert to LAB mode, there is a message asking if I want to merge the layers first (or something like that). Perhaps my question should be: is it possible to convert modes without flattening the layers?

One of the edge sharpening actions I use assumes you're in LAB mode, so what I do is complete the composite merge step described above in RGB mode, then Select > All and Edit > Copy on the composite layer, open a new file in LAB mode and paste the data in, then run my action in LAB mode. When done working in LAB mode you can move the layer back to the original RGB file if you wish as a new top layer (the conversion to RGB is done automatically and all the underlying layers are still intact below) or just print with in it LAB mode, or whatever. I usually rename the new top layer to indicate it was edge sharpened and include the parameters in the name, something like "SOE r 1.0" to indicate the radius (since I always use the same threshold and amount when edge sharpening).

Bill

Bill Hilton wrote:

If you don't want to merge (and usually you don't) then just go to the top of the layer stack and make a new blank layer, then hold down the Alt key (PC) and keep it down while doing Layer -> Merge Visible and you'll get a new composite layer on top ...

From:

This will probably work if the image is already in LAB mode. But I work in RGB mode and then convert to LAB mode before sharpening the L channel.

Don't convert the entire file, just copy the top composite layer into a new file and convert *it* to LAB.

When I convert to LAB mode, there is a message asking if I want to merge the layers first (or something like that). Perhaps my question should be: is it possible to convert modes without flattening the layers?

One of the edge sharpening actions I use assumes you're in LAB mode, so what I do is complete the composite merge step described above in RGB mode, then Select > All and Edit > Copy on the composite layer, open a new file in LAB mode and paste the data in, then run my action in LAB mode. When done working in LAB mode you can move the layer back to the original RGB file if you wish as a new top layer (the conversion to RGB is done automatically and all the underlying layers are still intact below) or just print with in it LAB mode, or whatever. I usually rename the new top layer to indicate it was edge sharpened and include the parameters in the name, something like "SOE r 1.0" to indicate the radius (since I always use the same threshold and amount when edge sharpening).

Bill

Thank you so much for this trick!

"Johan W. Elzenga" wrote:

wrote:

This will probably work if the image is already in LAB mode. But I work in RGB mode and then convert to LAB mode before sharpening the L channel. When I convert to LAB mode, there is a message asking if I want to merge the layers first (or something like that). Perhaps my question should be: is it possible to convert modes without flattening the layers?

You get a question if you want to flatten the layers first. Guess what happens if you say "Don't flatten"...

If you convert without flattening, the result will *keep* the layers, but will *lose* the adjustment layers. Not what I want, not sure if that's what you want. If you really want to help in this ng, try to understand the questions and provide relevant answers, like Bill did. Or stay on the sideline.

wrote:

This will probably work if the image is already in LAB mode. But I work in RGB mode and then convert to LAB mode before sharpening the L channel. When I convert to LAB mode, there is a message asking if I want to merge the layers first (or something like that). Perhaps my question should be: is it possible to convert modes without flattening the layers?

You get a question if you want to flatten the layers first. Guess what happens if you say "Don't flatten"...

If you convert without flattening, the result will *keep* the layers, but will *lose* the adjustment layers. Not what I want, not sure if that's what you want. If you really want to help in this ng, try to understand the questions and provide relevant answers, like Bill did. Or stay on the sideline.

Perhaps you should read the original question again before making comments on my answer. The question mentions specifically "a message asking if I want to merge the layers first (or something like that)". That will ONLY happen if you are talking about NORMAL layers. The question reads: "Changing modes will affect layer compositioning. Flatten image before mode change?" with buttons "Don't flatten", "Cancel" and "Flatten". It's clear you do indeed have a choice here.

If your document contains adjustment layers, you do NOT get a question if you WANT to flatten or not. In that case you get a different message, that reads: "Changing modes will discard an adjustment layer; change mode anyway?". Clearly, this is not a question if you WANT to flatten first. You have no choice: change mode and you'll loose your adjustment layers.

So, in my opinion, the original question was clearly about normal layers only, and so there is nothing wrong with my answer.

--
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

wrote:

"Johan W. Elzenga" wrote:

wrote:

This will probably work if the image is already in LAB mode. But I work in RGB mode and then convert to LAB mode before sharpening the L channel. When I convert to LAB mode, there is a message asking if I want to merge the layers first (or something like that). Perhaps my question should be: is it possible to convert modes without flattening the layers?

You get a question if you want to flatten the layers first. Guess what happens if you say "Don't flatten"...

If you convert without flattening, the result will *keep* the layers, but will *lose* the adjustment layers. Not what I want, not sure if that's what you want. If you really want to help in this ng, try to understand the questions and provide relevant answers, like Bill did. Or stay on the sideline.

First, read my other answer. I do believe my answer was correct, even if you didn't like it.

Next, here's a neat trick. You only want to convert to LAB mode because you want to sharpen the L-channel only, right? Try this: Use Unsharp Mask in RGB mode, and use a setting that is slightly too much. Next, choose "Fade Unsharp Mask". Set the mode to "Luminosity". That is the same as converting to LAB and sharpening the L-channel! Finally, if the effect is still too much, you can lower the slider to decrease the amount.

Like it, or should I stay on the sideline from now on if you ask a question?

--
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/

Johan W. Elzenga wrote:

wrote:

"Johan W. Elzenga" wrote:

wrote:

This will probably work if the image is already in LAB mode. But I work in RGB mode and then convert to LAB mode before sharpening the L channel. When I convert to LAB mode, there is a message asking if I want to merge the layers first (or something like that). Perhaps my question should be: is it possible to convert modes without flattening the layers?

You get a question if you want to flatten the layers first. Guess what happens if you say "Don't flatten"...

If you convert without flattening, the result will *keep* the layers, but will *lose* the adjustment layers. Not what I want, not sure if that's what you want. If you really want to help in this ng, try to understand the questions and provide relevant answers, like Bill did. Or stay on the sideline.

First, read my other answer. I do believe my answer was correct, even if you didn't like it.

Next, here's a neat trick. You only want to convert to LAB mode because you want to sharpen the L-channel only, right? Try this: Use Unsharp Mask in RGB mode, and use a setting that is slightly too much. Next, choose "Fade Unsharp Mask". Set the mode to "Luminosity". That is the same as converting to LAB and sharpening the L-channel! Finally, if the effect is still too much, you can lower the slider to decrease the amount.

Like it, or should I stay on the sideline from now on if you ask a question?

You are exactly right Johan.

--

__________________________________________________
Leo McKenzie
www.solocomputerservices.com/scsgrafx.php

"Johan W. Elzenga" wrote:

wrote:

"Johan W. Elzenga" wrote:

wrote:

This will probably work if the image is already in LAB mode. But I work in RGB mode and then convert to LAB mode before sharpening the L channel. When I convert to LAB mode, there is a message asking if I want to merge the layers first (or something like that). Perhaps my question should be: is it possible to convert modes without flattening the layers?

You get a question if you want to flatten the layers first. Guess what happens if you say "Don't flatten"...

If you convert without flattening, the result will *keep* the layers, but will *lose* the adjustment layers. Not what I want, not sure if that's what you want. If you really want to help in this ng, try to understand the questions and provide relevant answers, like Bill did. Or stay on the sideline.

First, read my other answer. I do believe my answer was correct, even if you didn't like it.

Next, here's a neat trick. You only want to convert to LAB mode because you want to sharpen the L-channel only, right? Try this: Use Unsharp Mask in RGB mode, and use a setting that is slightly too much. Next, choose "Fade Unsharp Mask". Set the mode to "Luminosity". That is the same as converting to LAB and sharpening the L-channel! Finally, if the effect is still too much, you can lower the slider to decrease the amount.

Like it, or should I stay on the sideline from now on if you ask a question?

My apology for my comment and thanks for your help. There are enough Yanks offending the Europeans and there is no need for yet another one.

- Not a Texan

wrote:

If you convert without flattening, the result will *keep* the layers, but will *lose* the adjustment layers. Not what I want, not sure if that's what you want. If you really want to help in this ng, try to understand the questions and provide relevant answers, like Bill did. Or stay on the sideline.

First, read my other answer. I do believe my answer was correct, even if you didn't like it.

Next, here's a neat trick. You only want to convert to LAB mode because you want to sharpen the L-channel only, right? Try this: Use Unsharp Mask in RGB mode, and use a setting that is slightly too much. Next, choose "Fade Unsharp Mask". Set the mode to "Luminosity". That is the same as converting to LAB and sharpening the L-channel! Finally, if the effect is still too much, you can lower the slider to decrease the amount.

Like it, or should I stay on the sideline from now on if you ask a question?

My apology for my comment and thanks for your help. There are enough Yanks offending the Europeans and there is no need for yet another one.
- Not a Texan

No hard feelings.

--
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.nl/