This is an appreciation post for Darktable, and this great community
Joining this forum made me dive deeper into processing than I’d anticipated, and I feel like I’ve made real progress in a few weeks. Here is an illustrative example.
I have a photo I am rather proud of that I’ve initially struggled to develop with AgX: highlights were too harsh and I’ve set too much contrast. Using filmic, I felt like the highlights were better preserved and they had a nice warm tone to them.
Then after a few weeks on this forum, I took a new look on it. I realized the “natural” warm tone was, I believe, the infamous filmic-salmon problem. In the meantime I learned to handle AgX better, discovered some tricks with contrast equalizer, and generally grew up in my photo appreciation.
Here is a zoomed comparison screenshot between the second version with filmic (left) and the third version with AgX (right). On the right version, we can see better color, noise and sharpening handling.
I find it interesting that the filmic version has harsher desaturation in the highlights, but that’s an issue we normally see with agx. Not to say that agx is necessarily wrong, but real world specular highlights are already naturally desaturated so processing them in agx causes a double whammy. Did you use tone equalizer to tame the highlights prior to tone mapping?
And that’s a very nice photo! I like that it’s technically good while also being a look into a day in the life of a tiny marvel of biology.
If you’d like to test it yourself, take a raw photo of a plant. Make sure it’s not overexposed so no channel is blown out and it renders in the RGB range–lower the exposure if needed. Turn off the tone mapper so what you are seeing is as close as you can get to the recorded pixels. Use the color picker to check the pixel values, or just zoom in to 800%. The pixels are not brighter green, but whiter green. We all know our eyes can trick us, but they aren’t in this case.
And somewhere in your memories, you do know what a bright colored light source looks like. It is not even a little bit white. Think of car brake lights at night. The idea of very bright light desaturating is an imitation of film and an homage to fire and heat, but it’s not a real optical effect.
Exactly. If you take a look at the sun (please don’t look long), it appears yellow, not white. Although space photos usually show it as a deep, saturated orange, I don’t know whether that’s the “true” color, or just something they’ve done to make it look jazzy.
I think that’s because they use a really narrow bandwidth filter to show more detail - I can’t remember what the wavelength is though.
EDIT: 656.28 nm or the Hydrogen-Alpha line apparently. That is a deep red, but I bet a little clipping and hue twisting will make it a nice orange?
EDIT again: Thinking further, the Sun is actually white by definition - it’s sunlight which is basically what we base our definition of white light on.
OK. I was getting over fixated on the highlight part and missed the specular part. So, specular reflections may be a different colour to the object they’re on and as they may reflect the light source, are often desaturated (though not if the light is coloured). Highlights are orthogonal to this as the effect is caused by the type of reflection, not the intensity.
.