Luminosity Masking

[revnart]

Hi, after "some" time I comeback with question for all of you

What methods do you use/know to create luminosity map (mask) - full one, not only darks, midtones or highlights?

I'm interested especially in ways different than:
• HSB B channel
• HSL L channel
• RGB Desaturate
• RGB Composite channel selection
• RGB Layer in luminosity blend mode
• LAB L channel
• LAB Composite channel selection
• LAB Desaturate
• LAB Layer in luminosity blend mode
• Grayscale mode
• Calculations > gray > normal

I'm making some "private-research" because of completely different results depending on techniques used.
Will be very happy if you could share your techniques/ideas.

Thank you

 

[polarwoc]

Seems like a very interesting question. I have never used luminosity masks before, so I would like to know all that anyone has to share on the topic. Meanwhile, if you @revnart have anything new other than what is on popular Youtube videos, please share.

 

[revnart]

I've already tested methods from my list, and noticed completely different results. Of course sometimes differences are very subtle and visible only in specific color areas but sometimes mask is different in much bigger scale. I understand that there must be some differences between HSL and HSB Luminosity channels but there are also similar but not equal masks in same color mode - example in attached gif

added two more methods to my list:
• Gradient map
• Black&White adjustment layer

 

[IamSam]

What methods do you use/know to create luminosity map (mask) - full one, not only darks, midtones or highlights?

Please define what you mean by a FULL ONE.

 

[revnart]

Examples are in gif from my previous post, but it's just mask based on luminosity values from image, like Lightness channel from LAB. "Full" is maybe not the perfect word

 

[IamSam]

Are you referring to a single method used for the isolation of just one tonal value and then making a slection from that?

 

[revnart]

As English is not my native language I'm afraid that I can misunderstood your post What I'm looking for are methods of converting layer or document containing color image into black and white layer/mask/selection (doesn't matter which one) containing only its luminosity data.

 

[IamSam]

containing only its luminosity data

All images have a luminosity or tonal range from 0 (black), to 255 (white) and everything in between is a shade of grey from dark to light. Are you asking about the entire tonal range from 0 to 255 or are you asking about a method of choosing just one tonal value?

 

[revnart]

Yes, entire tonal range 0-255 (full luminosity mask ). I'm not interested in partial mask - one tonal value, or darks, midtones, highlights etc. As I stated in my first post here, I know several different ways of doing that but I'm curious if someone here have more ideas

 

[IamSam]

I'm not interested in partial mask - one tonal value, or darks, midtones, highlights etc.

I guess I'm not understanding. A full luminosity mask would involve a process that extracts the entire tonal range (0-255) of an image including darks, midtones, and highlights! You can't exclude these. What this means is that the process would produce 255 different channel layers representing the ENTIRE tonal range of any given image.

I don't know of any process to do that. Can you link me to a tutorial that demonstrates what you claim?

 

[revnart]

@IamSam , we are for sure don't understand each other But I think I know where is the problem

There are a lot tutorials which shows you how to select highlights, or only darks etc. - I don't need to separate them.
I just need B&W version of image showing all its luminosity values at once from 0 to 255 (all = darks, mids and highlights) - don't want to exclude anything

Saying I don't want darks, mids or highlights - I meant that I don't need them to be separate

 

[IamSam]

OK, based on my current knowledge and understanding of Luminosity Masking, I guess I won't be of much help until I understand what it is your looking for.

For me, the concept behind luminosity masking is having the ability to isolate the luminous (tonal) values of an image into individual channels. This affords you the ability to select one of the many different tonal values of an image, create a layer mask (hence the term mask), and then make adjustments to that particular tonal range. As I stated in post #10, a full luminosity mask would involve a process that extracts the entire tonal range (0-255) of an image including darks, midtones, and highlights! The process would produce 255 different channel layers representing the ENTIRE tonal range of any given image.

It seems to me that your not asking about luminosity masking, you're asking about creating a single luminosity layer and not a mask, that shows the images 255 different luminous values at the same time.................which I don't think can be done, not to mention that it would be difficult to make selections from such a layer.

I apologize for not understanding.

 

[IamSam]

The best way to do what your describing is to use a Black and White adjustment layer.

 

[revnart]

OK, different approach No need to apologize, its my fault of not being able to explain correctly

Yes, idea behind luminosity masks is being able to catch specific tonal range But what I'm asking is not my finished target It's just first step, later when I will test all the differences I will separate darks/mids/highlights etc. from it.

Example:
If you open photo, and for example you convert it to LAB, L channel contains data I need, same as simple image desaturate - you throw away color, and hue leaving you only with image luminosity.

Channel is just black and white layer, same as mask so it not makes me any difference if the result of some method will be selection or layer or mask or channel since all of them can be transformed back and forth into each of them.

Hope now I made it simpler