clipping colors

a few years ago i was at a congress and saw these colourful chairs. Now i found these pics on my harddisk and came back to try an edit. Applying the neutral profile, there are no clippings visible, but whatever i try while editing, i find some clipping colors in the histogram. Would like to see, how you solve this.


DSC03272.ARW (23,7 MB)
DSC03272.jpg.out.pp3 (14,3 KB)

License: ![Creative Commons License](upload://g1CVFPFbqzp4kTnOJRXL2OdbDiR.png)
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

1 Like

This is what I end up with if I push it to the point of just not showing any clipping:


DSC03272.jpg.pp3 (15.9 KB) RawTherapee 5.8 development

I do think this rendition is too saturated, but pushing it without clipping was the goal here, not an end-result that looks pleasing…

1 Like

great, seems the L-curve in Lab* does the trick. IIRC the scene was some brighter, and this caused me struggle with the clipping…

Yeah, L*a*b is good if you want/need to work with just the lightness or colour components of an image.

You mention the actual scene being brighter: Another way to approach this is to use the Color appearance(Cam16 & Jzcz) tool in the Local Adjustments module as base and build upon that one. Here’s a (quick and pushed) example:


DSC03272.02.jpg.pp3 (19.2 KB) RawTherapee 5.8 Development

Everything is a lot brighter/lighter but you do loose some of the warm and deep colours. This is in line with the much lighter edit, though. I guess the sweet-spot is somewhere in between my first and second example.

1 Like

Your post was specfic to RT however I thought I would try and see how this one faired in DT with all the gamut mapping built in to the scene referred modules.

The new UCS saturation formula in the rgb colorbalance works nicely in DT. Its perceptual nature really helps with these sorts of images…when you toggle back to the JzHzCz mode previously used it really darkens and saturates the shadows… putting much more of the image out of gamut… With UCS selected you can get a nice light “airy” edit or make it quite a bit richer in color and still keep almost everything in gamut…

Exactly that was the point, when i decided to ask you for a play raw :slightly_smiling_face:

Thanks for your answer, unfortunately i dont have any experience with DT, thats why i asked for RT edits. But seems i should spend some time and investigate in DT.

That is why I didn’t say much more and frankly your image what was prompted me to play around a bit…its quite dramatic in DT when you alter just that setting… Here is a basic scene referred workflow with essentially all defaults and then running the autotune for filmic…

I have added rgb colorbalance with the basic add color preset and in the default UCS mode…

Now if you go back to the older mode…

Much darker using this colorspace model for saturation…

EDIT: Just because I killed the light in the edits above this is without the tone eq compression I had added

UCS

JZAzBz

1 Like

Great results! What kind of histogram is shown there? Looks different from RT histograms and how come, it doesnt touch the right side?

DT has a profile called the working profile and it has a histogram profile. I have them both set to Linear rec2020. If I am in gamut there then I allow the output profile to map the colors to srgb… for sure if I proof it directly as is to srgb I will be out of gamut…I can do that with the one gamut checker in DT… it will use the softproof profile to show out of gamut so I can set that to srgb. DT has an over exposure warning but it can be set to monitor exposure saturation or full gamut. I set it to full gamut and it will use the histogram profile as a reference… that is the one that I check on to be sure I am good. But here it is against sRGB

JZAZBZ against sRGB

UCS

1 Like

Ic , thanks for the explanations :+1:

What you asked for? Certainly, not. :person_shrugging:t4:
Sorry but I couldn’t resist…

My ART sidecar:
DSC03272.ARW.arp (24.6 KB)
Thanks for tolerating this monochrome offering.

:wink:

I’ve done 2 versions in RT5.8 dev, based on your edit.

1st one using CC and CL curves. Lately I’ve really been liking using the CL curve in this general shape to desaturate the shadows and highlights.
DSC03272.ARW.pp3 (15.1 KB)

2nd one using a/b curves to target the colors individually resulting in less saturated blues but more saturated yellows and oranges. Also with this method, hues will shift.


DSC03272-1.jpg.out.pp3 (15.1 KB)

Maybe with a combination of all of those (plus L curve) you might get what you’re looking for. :slight_smile:

1 Like

Wow, the second one looks already very good. I will investigate deeper in your edits.