Add ability to show the luminosity histogram alone and the rgb histograms alone in Affinity Photo

[KLE-France]

Ante-scriptum: Google-fu and AP-search-fu dutifully done. Didn't find this topic. But surely missed it because I doubt I'm the first to ask. Setting is Windows 10 Home, Affinity Photo Version 2.5.5.

The histogram's channel dropdown currently only offers "All Channels," "Red Channel," "Green Channel," and "Blue Channel."

Users should be able to:

1. show the luminosity histogram alone, without the rgb channels, and inversely
2. show the rgb channels alone, without the luminosity channel.

I know nothing about programming, but it seems that it would be relatively easy to add these choices, no?

I can only speak for the two other programs I use: both RawTherapee and XnView MP have this ability. Here's what this looks like in e.g. XnView MP:

 

 

If one were to be prioritized it would be the ability to show the luminosity histogram by itself.

Thanks for your consideration!

[GarryP]

I hardly ever use the Histogram Panel but this sounds like an interesting suggestion.

Can you give me some basic examples of circumstances where what you have suggested would be useful?

For instance, when might I like to see a luminosity histogram and what would I do with it if I could see it?

[fde101]

Personally I couldn't care less about the histogram, it is the most useless of the available scopes.

I would rather they spent time improving on the other scopes and really on the accessibility of all of the scopes:

• The Scope tab can only display one scope at a time and only one instance of the tab can be opened.  Ideally you should be able to display at least two scopes at once (the histogram again doesn't count.  It is just another scope and there is no reason to treat it differently than the others as is currently done - it should be merged in with the other scopes and the option exist to have multiple scopes displayed).  For myself I would normally want to have the option to see the RGB parade and the Vectorscope together, as that generally strikes me as the most useful combination.
• The scopes are always displayed at a fixed size, and it is WAY too small.  When working with Resolve I will sometimes break the scopes off and fill a screen with them.  I would like to be able to do that in Affinity Photo as well.

[Gripsholm Lion]

Anyone who has to produce sets of related images for websites will understand why consistent perceptual brightness matters. Inconsistency is amateurish. It is easier and quicker to achieve in this context with a luminance-only histogram, something which several other software companies (DxO, for example) appear to understand. Incidentally, they also understand that fewer clicks tend towards better UX, by using buttons in the same space as Affinity's dropdown.

[KLE-France]

55 minutes ago, GarryP said:

when might I like to see a luminosity histogram and what would I do with it if I could see it?

When you don't want the semantic interference of the red, green and blue ones. Being able to see the luminance channel alone, for me at least, is particularly valuable when preparing to print an image. Cambridge in Color has a good article that underlines the difference of the luminance channel.

31 minutes ago, fde101 said:

on the other scopes and really on the accessibility of all of the scopes

100% agree, if you can believe me! The scopes too can be greatly improved... but that's another subject. Do post your suggestions to the forum and you'll surely have a "+1" on my part.

[fde101]

1 hour ago, Gripsholm Lion said:

It is easier and quicker to achieve in this context with a luminance-only histogram

1 hour ago, KLE-France said:

Being able to see the luminance channel alone, for me at least, is particularly valuable when preparing to print an image

There is an Intensity Waveform available under Scopes which could theoretically be used for this purpose, but Serif botched the scopes implementation (as I explained above) by making them far too small and not allowing them to be enlarged, making it quite a bit harder to use effectively.

A waveform scope reads left-to-right with the image (left side of the scope reflects the left side oide of the scope reflects the right side of the image), and bottom-to-top with intensity (of luma in the case of the Intensity Waveform or of each color channel for the RGB waveform); the brightness of the trace at a point within that grid reflects how many pixels within the column of the image have that intensity - in other words, if the trace is particularly bright at the top of the scope along the left side and in the middle of the scope along the right side, this means you have very bright pixels along the left side of the image and a lot of mid-bright pixels along the right side of the image.

 

Here is an example of an intensity waveform display from Resolve (the ones in Photo are too small to make it out all that well):

Here I can see that the image is relatively dark on the left side as the brightest part of the traces are clustered around the lower end of the scope; however, if you look a bit more closely you can see that there are scattered bits of brighter data that rise well above the midpoint, up to somewhere between 640 and 768, just so scattered that they look like noise on the scope, so the intensity is not clustered around any one area.  The intensity gradually rises as you move toward the right side of the image, with roughly the right third of the image having some very bright areas, in this case from a blown-out sky.

This scope conveys all of the information that the histogram does and more.

 

The RGB Parade that I personally tend to favor most of the time is basically a set of waveform scopes broken out by color channel - instead of the color channels overlapping as they do with the RGB Waveform, they are displayed side by side.

With this I can see that each color channel follows basically the same pattern of being dark on the left and bright on the right.  I can see that there is a red bias in the middle of the image, as the red traces tend to rise higher than those of the other channels, but the brightest part of the image comes mainly from the green channel.  The blown-out sky does have a slightly greenish tint.

 

 

The vector scope does not convey brightness information but is rather a way to check on the calibration of hue and saturation.  Hue moves around the scope as with a color wheel and saturation increases as you move away from the center.  The brightness of the trace on the scope again reflects the number of pixels that fall into that category, so if you have a bright trace far from the center at a particular angle, you have a lot of pixels in the image which are highly saturated for the hue represented by that angle.  They are very hard to see on the itsy bitty vector scope that the Affinity products provide, but there are small boxes at the key angles representing the different colors, showing what one of the major standards considers to be the maximum "legal" saturation for each of the key hue values; if the traces extend past that, you have a highly oversaturated image which, in the broadcast world, may result in it being rejected by studios.  In addition to the key axes (RGB, CMY) there is also in many cases a skin tone line (it is there in the scopes in Photo), showing the hue that practically all human skin is clustered around.  If you have a closeup of a person's skin (face for example) you should see bright traces right on that line if the color was calibrated accurately.

 

From this vector scope (from the same image as the other two examples) I can see that the image as a whole is reasonably saturated but not overly so (a pure monochrome image would have one really bright dot in the very middle of the scope as there would be no saturation to pull the traces away from the center), but also that there is a fair amount of skin tone color which is very slightly tinted toward yellow as the traces along the diagonal skin tone line are pulled just slightly counter-clockwise, in the direction of the yellow channel (the box marked "Y" shows the saturation target for yellow) and away from the red channel (the box marked "R").

 

 

In Resolve you can pop the scopes out into a window showing up to four at a time, and I have a second monitor I will often pull that window onto and fill the screen with the scopes:

 

The individual scopes can be customized with gain for the brightness of the traces, which color channels to show, which scope to show in which of the four parts of the window when it is split out (you can also do a 3x3 grid for nine scopes with different configurations, or two scopes top and bottom, or just one really big scope), etc.

 

For example, with a 9-up display, you might have a separate RGB waveform along the bottom showing each individual color channel, replicating the parade but with a bit more horizontal space to see things in while sharing the window with various other scopes:

That is the kind of flexibility I would really like to see available in Photo as well, though really the biggest issue by far with the scopes Photo has already is that they are WAY TOO SMALL!!!

[fde101]

1 hour ago, KLE-France said:

The scopes too can be greatly improved... but that's another subject. Do post your suggestions to the forum and you'll surely have a "+1" on my part.

 

[GarryP]

@KLE-France Thanks for the link.

And now I know there’s a difference between ‘luminance’ and ‘luminosity’, which I didn’t think to consider until now.

[KLE-France]

3 hours ago, GarryP said:

there’s a difference between ‘luminance’ and ‘luminosity’,

Oh heavens! That part in the article where they say, "Unfortunately, the terms 'luminance' and 'luminosity' are often used interchangeably," that is 100% me!

Technically speaking, I do know there's a difference, but, luminance, luminosity, pffff—you know! How bright is it! 😄

[KLE-France]

A quick "heads up" for this thread: forum member lepr indicated correctly in another post I made that the "luminosity histogram" isn't one; the middling-blue histogram just indicates the overlap of the red, green and blue histograms, and indeed that is what's indicated in the Affinity Photo Help.

So, what I'm speaking of here is that overlap histogram (the "All Channels" histogram alone then?), and in need of adding is that AP needs thus a true luminosity histogram! That can't be too difficult to add, no?

In the meantime, as indicated by fde101 above, the Intensity waveform scope can act somewhat as a substitute.

[Gripsholm Lion]

Not exactly at-a-glance user friendly, though, is it? This is something like what I had in mind.