Something I haven't been able to figure out

[Mike Buckley]

Mike - some of us are interested as to how and why things work so that we can try to find a solution.

Completely understandable. However, my point is that if we can't perceive an unwanted visual change in the image, looking for a solution is looking for a solution for which there is no problem.

This is often said but also 100% incorrect.

I suppose it's just another one of those myths that caught on on the Internet. I have never tested it myself and instead assumed that so many people couldn't be so wrong. Having said that, I never tested it because, as mentioned in my previous post, when I can't perceive a change in the image nothing else matters to me in the slightest.

EDIT: See my next post.

[Mike Buckley]

I decided to test the saving of a file for a reason I won't discuss other than to mention that it has nothing to do with photography. I made an extremely slight change to a JPEG and saved it. It became smaller, exactly as I have repeatedly read since I first got into digital photography would happen.

This is my last post on the subject, primarily because I don't want Kim's thread to become hijacked and secondarily because I have nothing to prove to anyone.

[purplehaze]

Kim,

An Affinity forum actually pointed to this Photoshop tutorial on Blend If: https://youtu.be/3GKDBXrpV8s

The Blend Options in Affinity are said to be even better.

Now, all I have to is translate this into Capture One

[DanK]

Mike,

I agree that there is no payoff in making tweaks that aren't visible, but the change in saturation from tonal adjustments is often noticeable and sometimes problematic. I first encountered this as a problem when increasing contrast in images of highly saturated red flowers. I also found it a problem when trying to add some dynamic range to landscapes taken under foggy conditions. They became badly oversaturated because the contrast adjustment was so large. There are several factors that lead me to move some images from Lightroom to Photoshop, and this is one of them. Lightroom offers no way to increase contrast without increasing saturation.

Re JPEGs: I'm wondering whether the issue is whether the file is changed between saves.

[Mike Buckley]

They became badly oversaturated because the contrast adjustment was so large.

My hunch is that that's the key, that you were making a dramatic change.

This brings me back to my point that making a change to the luminosity doesn't necessarily make a perceivable change to the saturation. I rarely make dramatic changes and the few times that I do, I can very easily revert the saturation to what it was before I changed the luminosity. That's done by simply applying the same mask to a new edit step that reduces the saturation.

[Cantab]

Kim,

An Affinity forum actually pointed to this Photoshop tutorial on Blend If: https://youtu.be/3GKDBXrpV8s

The Blend Options in Affinity are said to be even better.

Now, all I have to is translate this into Capture One

Janis, thanks for posting the Youtube link -- very interesting.

[KimC]

Thank you for the added information.

Just to be complete, I thought I would show what the dodge curve looks like:

In both cases, these are good starting points, but I will sometimes place with the midpoint on the curves (both dodge and burn) to see how things look. Moving the centre point further away intensifies the effect and bringing it closer makes it more subtle.

[KimC]

Thanks for pointing me to this Janis, it was very helpful. I watched both videos - the Affinity one, and the one below. Rather overwhelming how much there is to learn... but one thing at a time.

Kim,

An Affinity forum actually pointed to this Photoshop tutorial on Blend If: https://youtu.be/3GKDBXrpV8s

The Blend Options in Affinity are said to be even better.

Now, all I have to is translate this into Capture One

[KimC]

Thanks to all those who shared information is this thread; it's very much appreciated and gives me hours of experimentation. Always so appreciative of the people willing to help on this site, as learning photography and all that comes with it, can be overwhelming at times when you are flying solo.

[jisner]

Could someone lend me some help as to how she enhances the horse's coat? Thank you

As others here have suggested, dodging and burning is the way to do what you want. Not using the tools, but by painting on on layers. To dodge, paint in white on new layer in Overlay mode. To burn, paint in black on a new layer in soft light mode. Here's the key: don't paint freehand, but rather paint through a luminosity selection. Luminosity selections are highly feathered, which means you don't have be particularly accurate with your brush to avoid halos and other artifacts.

This is a technique that Tony Kuyper calls "luminosity painting." Luminosity selections can be created manually, but it's a chore. It's easier to use Tony's TK5 panel, a Photoshop extension.

[Manfred M]

As others here have suggested, dodging and burning is the way to do what you want. Not using the tools, but by painting on on layers. To dodge, paint in white on new layer in Overlay mode. To burn, paint in black on a new layer in soft light mode. Here's the key: don't paint freehand, but rather paint through a luminosity selection. Luminosity selections are highly feathered, which means you don't have be particularly accurate with your brush to avoid halos and other artifacts.

This is a technique that Tony Kuyper calls "luminosity painting." Luminosity selections can be created manually, but it's a chore. It's easier to use Tony's TK5 panel, a Photoshop extension.

Either the overlay or soft light blending modes can be used interchangeably for the technique you are writing about. The soft light approach tends to be a bit more aggressive. The problem with both approaches is that they lead to the saturation issues I referred to in #9 and are harder to correct for as this has to be dealt with separately without being able to use a nice clean solution like a clipping mask.

I've even seen some people dodge and burn on the same layer by just changing between the black and white brushes.


When it comes to creating a luminosity mask, I've never understood the big deal some people seem to have in creating them. Getting the pre-made approach seems to be more complicated than making a custom on from an alpha channel and then tweaking it. I find these useful for extracting clean details, but rarely for something else. I find that painting, with a stylus is my preferred way to dodge and burn as techniques like this one and using Layer Styles "blend if" don't necessarily let me dodge and burn where I want to place the effect. I use these techniques for contouring and no level of automation can beat building that up by hand.

[jisner]

Either the overlay or soft light blending modes can be used interchangeably for the technique you are writing about. The soft light approach tends to be a bit more aggressive. The problem with both approaches is that they lead to the saturation issues I referred to in #9 and are harder to correct for as this has to be dealt with separately without being able to use a nice clean solution like a clipping mask.

I've even seen some people dodge and burn on the same layer by just changing between the black and white brushes.


When it comes to creating a luminosity mask, I've never understood the big deal some people seem to have in creating them. Getting the pre-made approach seems to be more complicated than making a custom on from an alpha channel and then tweaking it. I find these useful for extracting clean details, but rarely for something else. I find that painting, with a stylus is my preferred way to dodge and burn as techniques like this one and using Layer Styles "blend if" don't necessarily let me dodge and burn where I want to place the effect. I use these techniques for contouring and no level of automation can beat building that up by hand.

If you look at the curve equivalent of soft light, it's actually less aggressive than overlay. I agree with comments by others that for small changes (not radical changes), concern about color shift is unwarranted. Unless someone enjoys channel calculations, they'll use actions to generate their luminosity masks as a bare minimum. If that qualifies as "automation" or a "pre-made approach" then mea culpa. Finally, luminosity selections are useless for extracting details because they're so highly feathered. For example a luminosity selection made by ctrl-clicking [cmd-clicking] the RGB channel (usually called a Lights-1) extends all the way to 0, although it starts feathering at 128.

The nice thing about painting on separate layers (transparent layers, not 50% gray as some advocate) is that you can later ctrl-click [cmd-click] the layer and get back the original selection. That lets you use the same selection in a later adjustment layer, e.g., a Hue/Saturation.

[Mike Buckley]

It would be helpful if the people explaining how layers, masks, blending modes and the like work, that you begin each post with the exact name of the software you are referring to. It would be especially nice if you would edit your previous posts to include that information.

[Manfred M]

It would be helpful if the people explaining how layers, masks, blending modes and the like work, that you begin each post with the exact name of the software you are referring to. It would be especially nice if you would edit your previous posts to include that information.

Photoshop CC 2017

[DanK]

I am a bit lost by this debate about luminance masks. I have used them infrequently, so my level of expertise with them is very low. However, I don't think they would be useful for many images for which I need burning or dodging. I have used luminance masks when I want to make an adjustment to a specific tonal range. That isn't what comes to mind when I think about images in which I have dodged or burned. More often, I have a specific part of the image I want to alter, without making any change to other areas of the image with similar luminance. To take a simple example, I recently edited a few candids of an 18-month-old child. In some cases, the light on her face was either to low or too high, so I dodged or burned around the face. It seems to me that luminance masks would be useless for this, unless I wanted to adjust all other parts of the image with similar luminance (which I did not want to do).

The nice thing about painting on separate layers (transparent layers, not 50% gray as some advocate) is that you can later ctrl-click [cmd-click] the layer and get back the original selection. That lets you use the same selection in a later adjustment layer, e.g., a Hue/Saturation.

That's new to me and seems quite useful. Thanks for posting it. I will play with this.

[jisner]

I am a bit lost by this debate about luminance masks. I have used them infrequently, so my level of expertise with them is very low. However, I don't think they would be useful for many images for which I need burning or dodging. I have used luminance masks when I want to make an adjustment to a specific tonal range. That isn't what comes to mind when I think about images in which I have dodged or burned. More often, I have a specific part of the image I want to alter, without making any change to other areas of the image with similar luminance. To take a simple example, I recently edited a few candids of an 18-month-old child. In some cases, the light on her face was either to low or too high, so I dodged or burned around the face. It seems to me that luminance masks would be useless for this, unless I wanted to adjust all other parts of the image with similar luminance (which I did not want to do).



That's new to me and seems quite useful. Thanks for posting it. I will play with this.

Here is an 8 minute video that demonstrates the luminosity painting technique. In the video, the goal is to increase local contrast by lightening (i.e. dodging) certain lights and darkening (i.e. burning) certain darks.

The dog in the OP is a good subject for this technique. The streak of reflected light on the dog's coat was already there. You can enhance it by dodging through a luminosity selection which selects the streak. To exaggerate the effect, you could then select for the surrounding dark fur and burn through that.

An example that would not be appropriate for luminosity painting would be a portrait with the face evenly illuminated by flash. Say you wanted to create the effect of split lighting (one side darker than the other). Here, luminosity painting would not help you because there is no mask that would be selective for one side or the other. However if the original portrait had been taken under split lighting, you could use luminosity painting to exaggerate (or lessen) the effect.

[DanK]

Here is an 8 minute video that demonstrates the luminosity painting technique. In the video, the goal is to increase local contrast by lightening (i.e. dodging) certain lights and darkening (i.e. burning) certain darks.

The dog in the OP is a good subject for this technique. The streak of reflected light on the dog's coat was already there. You can enhance it by dodging through a luminosity selection which selects the streak. To exaggerate the effect, you could then select for the surrounding dark fur and burn through that.

An example that would not be appropriate for luminosity painting would be a portrait with the face evenly illuminated by flash. Say you wanted to create the effect of split lighting (one side darker than the other). Here, luminosity painting would not help you because there is no mask that would be selective for one side or the other. However if the original portrait had been taken under split lighting, you could use luminosity painting to exaggerate (or lessen) the effect.

I understand. The example in the video is labeled dodging and burning, but it is essentially a very sophisticated global tonality adjustment--a more sophisticated version of what one can do with a curve--in which one lightens or darkens all areas in an image selected solely based on luminance. That is very different from dodging or burning areas selected for other reasons, as in the example I gave, where one selects based on factors other than luminance. I was simply trying to clarify the difference.

In the old darkroom days, one was pretty much limited to the latter. One dodged by taking an object--I used a flat cardboard circle on a stick--and waving it around over a given area during part of the exposure of the paper. One burned by waving a large sheet of cardboard with a hole cut out. You could approximate the technique you describe only by doing that over every area with similar luminance, which would take some agility.

[jisner]

The example in the video is labeled dodging and burning, but it is essentially a very sophisticated global tonality adjustment....That is very different from dodging or burning areas selected for other reasons, as in the example I gave, where one selects based on factors other than luminance. I was simply trying to clarify the difference.

The dodging and burning in the video are local adjustments, not global. He painted through a selection so that only those elements of the scene that were selected were affected. Notice that he uses a large brush, yet the effect of the brush does not spill over into areas that are not selected.

Dodging and burning are primarily used for tonality adjustments, but I see no reason why they can't be used for "factors other than luminance." TK5 (and probably some of its its competitors) allow you to select based on other factors: component color channels (in RGB or Lab), saturation, color, and Color Range. From these you can make Lights, Darks, Midtones, and Zone masks for any of these channels, and you can "infinitize" them as he illustrated in the video (which works by applying a Levels adjustment directly to the channel).