Testing the defaults won’t do any good. The 0.5EV bump is a suggested starting point. If you don’t like it, then change it.
Obviously, I was merely saying that posting random images saying sigmoid vs filmic is no way to compare the two tone mappers. It had nothing to do with any criticism of the limit merely that it was a variable to be considered when you are comparing….for example if the best performance in general for sigmoid was 2 ev and we were using it at .5 to compare to to default filmic then obviously this would have an impact. I know that it is just a rough starting point and usually needs to change image to image……My main point was to compare conditions should be normalized what ever that looks like……
Yes taste is the main thing. It seems there are many here who prefer crushed blacks, which is something I avoid religiously as I like shadow detail. This can lead to my images being lower contrast, which is also to my taste as I like the old master paintings. This opposite to the high contrast modern approach.
How we view images makes a large difference to our response too, and I expect most view this thread, perhaps on a phone screen, or at the least not viewing images full size. Why? Again, partly the contrast, but also the detail. Some images getting the most likes are found to have harsh noise, clear banding or mask edges, demosaic artifacts, etc. These would not look good printed large, but are not a problem viewed small. Playraw is an interesting melting pot of not just different tools, but different aesthetics.
As one mentioned above, I often wonder how much monitor plays a part as well. Am I seeing others work how they see it, and vice versa? In some cases, almost certainly not.
You could equally argue that revealing the shadows in a photo like this can only take away from the sky which I identify as the most important thing so in this case very dark shadows even black would be fine to me. Also is there anything worth showing in the highlights, often you can lighten them to reveal some better detail but at what sacrifice to the rest of the image. We all go to the sky naturally and why not it is amazing but there are reflections on the water to play with and the light cast on the boat. It really is all about taste, however for sure there is one thing I think you should decide and that is what is your target, ie where do you want people to look or what do you want to emphasize as the focus of the image. For me that should be the main goal not whether you can lift the shadows or reel in the highlights. We use shallow depth of field to blur the details in the background and focus on the subject and sometimes it is the case shadows… I think your intent for the image should guide you and I suppose that could be to always lift the shadows for a look but IMO that is not always warranted. All my comments come from my bias of how I look at the scene and how I perceive the light and color and I would not impose my bias on anyone. Clearly lots of variability in perception from all the sample images…
3 Likes
Yes, I agree with you about intent, so it is simply a taste difference. There is very rarely a time when I find crushed black pleasing to the eye, which is why I lift the shadows. I don’t do it just because I can. In sunset shots, crushed blacks remind me of what I can get on my phone, or from any camera with limited dynamic range and inferior raw processor. It is also a matter of influence. The old masters always had detail in their shadows, and my eyes very rarely perceive pitch black either, except perhaps at night time without lights - there is always some detail to perceive. My intent with this image was something quite dreamy, hence the lighter more pastel colours and softer background. I’m not saying any of this to suggest my way is the only way - the beauty of playraw is the various interpretations - but to expand upon Andrea’s comment that differences are more between taste than tools. The tools are only an apple to apple comparison if the same artist with the same intent tries them both, as opposed to various artists with different intents.
3 Likes
The main difference between sigmoid and filmic from the perspective of a user imho lies in their fundamentally different approach. While filmic tries to provide as many parameters as possible which enable the user to make changes, sigmoid restricts itself to a few adjusting screws. Sigmoid shifts the finetuning to a greater extent to other tools such as tone equalizer, exposure etc.
With both approaches, results can be achieved that are so close to one another that they can hardly be distinguished. Even if the same image is viewed on different displays, the differences are often greater than the results of the tone mapper viewed on the same display.
The way to the result is a bit different for the two approaches. While filmic provides a low-contrast image as a starting point in the standard settings, sigmoid is set to be more contrasty. As a user, the way to a high-contrast and clear image with sigmoid is therefore easier, whereas filmic is more suitable when detailed shadows or foggy moods are the goal.
I would say that sigmoids approch is closer to the workflow found in commercial editing software like lr. This would make it easier for new users to switch to darktable.
Both approaches have their advantages and limitations, but both make sense as they support different “ways to rome”.
4 Likes
For this image I’m absolutely team don’t-reveal-too-many-details-in-the-shadows. I think that way you can make the boat and the ducks stand out more. But as Todd said I think the most important part is to have a vision beforehand and to edit accordingly. After all there is no objective ‘wrong’ in arts. There are also lots of images from the old masters that play with harsh contrasts with barely any details in the shadows, btw. Fan Ho comes to mind.
Anyway, nice discussion. 
RT5.8 dev build…
4 Likes
In the case of this image I have lived for 40 years on a lake in the summer and been out fishing until dark and it would be my experience that looking in that direction with the backlight from that distance there would be very little detail in the shadows. It is often only until you bring the boat into the shadows that you can even gauge how far you are from the shore. I spent a lot of time retrieving lures from trees and the shore as my brother-in-law seemed to have poor depth perception in this lighting
I think reflecting one’s perception or achieving a certain perception is in the end is what we are striving to do and that will be different for everyone to some degree or another.
I came across a quote that might reflect what Tim’s philosophy might be…only guessing…
“For landscapes especially I think photography needs to deal with the difference in the way a scene is viewed compared to the way we do so with our eyes. You look at a dark area of a scene in front of you and your eyes are going to adjust to it and it’s not going to seem as dark, hence the level of contrast might seem less strong than it actually is. It’s often that perception of lower contrast that the photograph actually needs to reflect rather than the reality of the scene.”
2 Likes
I was referring to painters. You see deep blacks a lot more in photography, and in my opinion if there is one style it can work effectively its black and white, which is what Fan Ho shot.
Wow, you did a great job! Maintained the pretty colours and achieved high contrast without crushing them.
1 Like
Ok fair enough, I misunderstood.
And thanks, I was fiddling quite a lot on this one…
I prefer a look of “realism”, even if it cannot be perfectly captured in a photo. So, when I did my edit, I decided that if I were actually viewing the scene, my eyes would see some details on the bank across the waterway, so I lifted the blacks some. Maybe not enough, looking back at my edit.
PS - Looking back, nah, I think I still like my edit.
3 Likes
Not sure what you compare here, so I will give you the step-by-step…
- Filmic + color calibration + CA correction + denoise + exposure: (white balance made on the boat because I have no clue)
- +Tone EQ :
- +Color balance RGB
Full stack :
DSCF9668.RAF.xmp (33,0 Ko)
Tone EQ is mandatory here because we have clearly-defined luminance areas to balance against each other, and no 1D tone curve will let us (that falls back to dodging and burning).
Also color balance RGB is mandatory to recreate color in shadows, because we have hit the limit of detection of the sensor there and pushed the DR really far, so all we have is an aggravated cold black-ish on trees.
10 Likes
I’ve created also two versions of this very nice image:
One using the sigmoid module:
DSCF9668_01.RAF.xmp (10.5 KB)
and one using filmic:
DSCF9668_02.RAF.xmp (11.0 KB)
With filmic it’s much more easy to get details out of the shadows while with sigmoid I find it more easy to get a base image which looks quite decent out of the box.
2 Likes
A completely different interpretion of the scene. Nice mood. The more I looked at it, the more I liked it. Could not resist to give it a try with sigmoid:
DSCF9668_06.RAF.xmp (12.2 KB)
main settings, different from yours: sigmoid (contrast 0.76, skew -0.2, color processing: RGB power norm), slightly different color balance rbg, different handling of shadows and highlights in tone equalizer.
Thanks for explaining your workflow, @anon41087856.
1 Like
