Comparing filmic color science v5/v6

My tonemapping method is designed to keep the saturation in the highlights while in the same time not doing overflows in the working rgb profile used, actually it slightly darken the “luminance” in the highlights

This is usefull to avoid orange shifting to salmon hue, blue shifting to purple, red shifting to pink,…

1 Like

Does that mean you chose to move “out-of-gamut” to lower luminance, where filmic moves those colours to lower chrominance? Because you are still doing some form of “gamut-clipping” in your tone mapping.
As you have to, when not allowing out-of-gamut colours in the module’s output.

Of course, if there’s a way to prevent such colours entering your tone mapping, that might be even better. Not sure how feasable that is.

It seems that filmic is doing a rec2020 ->srgb-> gamut clipping to srgb->rec2020 conversion inside the module itself.

So after the module there no reason anymore to work with a wide gamut color space

Nope, I’m not doing any gamut mapping inside the tone mapping

That statement doesn’t look right to me, simply because AP’s gamut shift diagram, as referred to just above, says “Kirk/Filmlight Ych space” is used.

I was wondering how big Kirk/Filmlight Ych space is in relation to rec2020. Does Filmic look at your Export Profile and take that into account when gamut mapping, perhaps? I’m surprised if anything is fixed on sRGB because I thought the philosophy of Filmic was that it’s somewhat future-proof, in that when we have big-gamut screens, we’ll be able to make use of them without completely re-editing our photos. I’m not sure how that will work in practice, anybody? I just had a quick look at Filmic in the manual, Display tab, nothing leaped out at me.

Maybe this, however it has not sense to do it inside filmic (this should be done in the color space conversion module or in a new module) and it has not sense to force users to do it always without the possibility to disable this gamut mapping.

I’m open-minded about this (I think!) Hopefully you’re aware of rendering intent when images are output. I just posted about this here -

1 Like

See

1 Like

That was a very interesting and insightful read for me. Some questions I had were answered there. Thanks for linking it @kofa .

My attempt with dt 4.0.0, which I’m still getting used to.


DSC09445.arw.xmp (8.1 KB)

2 Likes

filmic is a pixel-wise operation, so halos are unlikely, most of the time. CA can result in halos; the other possibility is highlight recovery. What happens if you apply (just as an experiment) very strong CA correction, and crank up the HL recovery threshold to +6 EV? ‘Of course’, the norms can also play funny on harsh transitions.

HR makes no difference. It does not go up to 6 EV but turning it off should do as well.
Massive CA reduction appears to help at first but after applying +2 EV exposure (which the example image certainly needs to resemble reality) it’s obvious that the problem does not lie there because the result looks very artificial.
So this is CA on full, +2 EV, no HR:

Preserve chrominance: MaxRGB

Preserve chrominance: No

What I’d expect from the effect applied would be smooth transition from blue sky to green branches.
I’ve been experiencing this since using Filmic RGB. It’s not a new thing, it’s just more obvious with MaxRGB. And I was always stomped here that I use it wrong, too much CA, etc etc, you name it. :slight_smile:

Oh well. Do any of the other norms give better results?
And/or, if, instead of boosting exposure everywhere, you use tone equalizer to only shift midtones or to tone down the highlights?

Some do, some don’t. Yes, there’s several ways around this (color balance RGB for example). However the starting point when working with Filmic has always been to fix the exposure first and then deal with the consequences.
For me it’s easiest to set the preservation to no. :slight_smile: After all, in reality the sky isn’t usually that blue and if I want to make it so, I can use different tools.

I also use no and luminance Y quite often with v6.

My render of the very image:


P1220719.RW2 (18.8 MB)
P1220719.RW2.xmp (7.3 KB)
License: Creative Commons, By-Attribution, Share-Alike.

At this level of detail, the demosaic algorithm probably can’t produce accurate colour information.

1 Like


P1220719_01.RW2.xmp (13.7 KB)

This is with maxRGB. I think Eucledian looks decent (more believable):

Which color managed viewer do you recommend?

I’ve been using gThumb for a while now for viewing and importing.
If there is an embedded profile an the image, it will let you toggle between that and the system profile.

1 Like

I use geeqie

2 Likes