I might just add another thought - trying to get the image to look like the OOC jpeg without sliding sliders in darktable is a bit like… (searching for metaphor) trying to make a cake sweet without sugar! (not implying that you were doing this 
) It can be done but not… not naturally, because dt is designed to give a really neutral starting point.
The style I made (for example) effectively ‘presets’ a bunch of sliders, which (to my understanding) is more the preferred option.
This is just my point of view by the way… don’t take it as anything more than that!
1 Like
Yeah, that makes perfect sense. I’m not opposed to adjusting sliders or anything! But since the primary purpose of a LUT is to modify colors to take you from point A to point B in a predictable and repeatable way, I figured that was a good place to start.
I feel that I’m getting close enough now, that any differences would be down to personal taste.
Really, I just didn’t want to have to make a choice between RAW with bad colors but more dynamic range, or JPEG with good colors but less dynamic range. If I can get the RAW to look like the JPEG, then it’s the best of both worlds.
2 Likes
Absolutely! And in the end, all that matters is that it looks how you want it to.
I didn’t mean any criticism… rereading my last post I think it seemed a little grumpy. Wasn’t meant to be, and thanks for sharing the images too!
1 Like
@123sg Thank you for sharing that Darktable style link. I plan to study that to see how you arrived at that point. I feel this will be educational for me. It’s also nice that Darktable makes it easy to concisely share a sequence of edits like that.
1 Like
One thing I tripped over initially was fuji’s DR modes. In DR400, the raw file is underexposed one stop, and in DR800 it’s two stops. My Lua script detects this using exiftool, and adjusts the tone equalizer accordingly. This usually gets me pretty close to the JPEG’s exposure. One could probably do a similar thing to implement in-camera white balance shifts.
Film simulation colors however are a different matter. Actually, three different matters:
The first is white balance, which is implemented differently by Fuji than in darktable. That’s what you discovered with your Color Correction change. The problem is that this change varies with color temperature, so it can’t be consistently replicated by a LUT alone.
Second is the tone curve. My LUTs are intended to be applied after filmic. This means, however, that they rely on filmic’s default tone curve. In 4.2.0, filmic’s defaults changed, so my LUTs are probably no longer correct. Stuart Sowerby’s LUTs have a strong tone curve built-in, and are intended to replace filmic. But of course they don’t implement the scene-to-display mapping, so you should probably put them after a zero-contrast filmic.
The third issue is colors. It seems to me that my LUTs try to partly “undo” filmic’s color preservation, which leads to issues with saturated highlights. I should try to redo the LUTs for 50% sigmoid, which should fit better. I suspect Stuart Sowerby’s LUTs to have similar mismatch with filmic’s color preservation modes. So maybe they’ll work better after a zero-contrast sigmoid, too.
Personally, I’ve found LUTs ultimately didn’t quite work out for me. I’m currently experimenting with a separate rgb tone curve after sigmoid, and then a lookup table instance to correct colors. Splitting the film simulations thus into a tonal part and a color part seems to work more robustly.
Other raw developers seem to solve these issues by providing camera-specific tone curves and LUTs (ICC profiles) to get them all to a common baseline, and then implement film simulations as a second-level LUT from there. Since darktable lacks this common baseline, a LUT-based approach is bound to be difficult. Especially since darktable’s default rendering changed subtly version-to-version.
However, I must say that most of my color rendering issues have been resolved by switching from filmic to sigmoid. This is highly personal, and perhaps mathematically misguided, but I just prefer the look of per-channel, hue-twisted sigmoid highlights.
6 Likes
Thanks for all the info!
@bastibe , would you like to try the style I posted above? I’m sure it’s far from perfect, however, I’d be very interested to see what you think as you’ve done a LOT more of this kind of thing. And I don’t have a Fuji…
It’s basically using color calibration as a channel mixer to shift the colours around, and sigmoid to add contrast and saturation. I tried to match the Fuji X-T3 Provia that @ratherlargerobot kindly posted.
I’ve tried it on a few raws I was able to download, and it seems better than I expected, which has sort of piqued my interest.
Two mistakes… one is it need more highlight saturation - would be easy to fix in color balance rgb, and two, I accidentally included the lens correction as a fixed setting so it’s wrong for any other lens.
I could redo it but not much point.
I am not sure if the post of @herbert-50 was already mentioned or not? But with his styles, I got pretty good results in replicating fuji-jpegs. What I like about his styles, he uses
- filmic RGB as the base for the style
- a tone curve for contrast
- a color lookup table to match the color with the simulation
so you can also, for example, use only the ClassicChrome colors in your edits. What I quite like.
He also has a lengthy blog post how he created the styles, but only in german.
I hope this helps you 
2 Likes
While it obviously takes longer, I do think it makes more sense to first learn to use Darktable “correctly”. You can then add LUTs and other tricks later. Without a proper understanding of the workflow, you’re likely to run into frustrations regularly.
1 Like
This is sort of what DT chart was for when it was working…ie match a JPG to the raw. It made a clut and tone curve instance…there is also a small script that can be used to match a JPG and raw using Argyll… creates an ICC. GitHub - pmjdebruijn/colormatch: ColorMatch
There’s also the color mapping module in darktable, which you can use to copy the look of a jpeg to a raw. But it is far from an exact match and can produce quite wonky results.
@bastibe Interesting to read that you find sigmoid has solved a lot of your colour rendering issues. My experience so far has been the opposite - that sigmoid introduces a very warm colour shift as well as losing a lot of highlight detail. Admittedly I haven’t don’t much in-depth experimentation with it yet, but I was initially turned off by this extra warmth. On the other hand, the new defaults in filmic RGB seem to have really helped make that module much better. I especially like to turn off preserve chrominance for the desaturated highlights, which I love from some of my Fuji film sims like classic neg.
1 Like
The new HLR methods are also rescuing filmic from being blamed for the blue cyan and other situational artifacts or side effects that it got blamed for…
The thing is, I very much like a slow highlight rolloff that fades smoothly into white, like negative film does. This I find somewhat difficult to achieve with filmic, as it tries to preserve too much color for this kind of look. Sigmoid with a strong negative skew, however, does exactly what I want.
I think I understand what filmic’s color perfection modes are trying to do, and use them occasionally for that purpose. It’s just not what I typically want to achieve.
1 Like
I fully understand that, and I like it too. I find that setting “preserve chrominance” to “no” in filmic gives this nice rolloff look while preserving the highlight detail better than sigmoid. I have done quite a few A/B comparisons with filmic and sigmoid, and I find that I usually prefer the look that filmic provides in fewer steps. This is because I often need to adjust color balance, tone EQ and sometimes exposure after turning on sigmoid.
Ultimately, they both can produce almost the same results after tweaking the settings and other modules, so choosing the one to use is simply a personal preference. I am very glad that sigmoid has been developed and included for several reasons:
1 - It sometimes gives the better look in certain images (particularly where smooth gradients are wanted, like sunsets)
2 - More options are always welcome in toolbox software like darktable
3 - It has pushed the development and tweaking of filmic for the better
4 - @jandren has said he intends to continue developing sigmoid based on feedback, so I’m looking forward to future versions (it may yet become my favourite!)
4 Likes
I actually think v5 is pretty good for this and offers a lot of control for that but many people were also wanting the opposite and so AP came up with v6 which people keep using as it was the new shinny version but personally I find v5 a lot more configurable and you are not fighting the gamut control put in v6… I agree with a few others that it is nice to have both sigmoid and filmic and that it seems some images work better using one or the other and some there if really not much difference… Sigmoid might require less tweaking when it works than v5 but I have found when I really want control the span a placement of the latitude coupled with saturation slider allows for a range of images from near monochrome to oversaturated beyond reason…
Oh yes I’m watching!
The most obvious update is maybe extremely small but I had liked to figure out what the best default value for hue preservation in sigmoid would be. In terms of making it the least likely users want to change it.
@bastibe just added a vote for 50%
Haven’t used luts much so interesting to see that application together with the sigmoid module! Please share if there are any specific that make you stumble with this combination!
As for the compressed highlights in sigmoid, I’m a fan of the look in most cases. But it’s also an interesting thing to think about solutions.
6 Likes
Maybe a bit of a tangent, but this is what got me thinking that LUTs might be a good approach in the first place. I know, apples and oranges…
I purchased the X-T3 LUTs from colorizer.net, which convert Fujifilm F-Log movie footage into your choice of the Fujifilm picture styles, including variations for DR200, DR400, and a couple different contrast/saturation combinations. When you apply one to F-Log footage, the translation appears to be perfect. I shot side by side tests at one point, and they matched really well.
Having said all that, I think that learning more about the fundamentals of raw processing and Darktable will be time well spent. You folks have proved to me that I can get the sort of results that I’m looking for.
3 Likes
I think you’ll find it hard to reach a consensus. In most scenarios I prefer 100% hue preservation, with slightly positive skew. All comes down to personal taste, and for that we have presets.
The problem with LUTS can be if you didn’t make it then it is not often clear what it does… If it has just color or color and tone or whatever. Since its basically just a map then the farther you are from the conditions in which it was created then you are less likely for it to look as intended. Depending on these elements it might not work at the current place in the pipeline or even with the tonemapping of filmic or sigmoid applied so I hope you find a solution but you might just find that the luts give you this match here and there but not consistently…
This is a very nice breakdown of the issues requirements and uses of LUTs…
1 Like
Thanks Jakob! It’s up to you whether you want to add more features to sigmoid and increase the complexity (its simplicity compared to filmic is certainly one of its strengths), or whether you are happy for other modules to perform what sigmoid does not. As I mentioned before, you can get almost identical results by playing with the sliders and sometimes involving other modules. As for me, I’m going to stop making grand statements about sigmoid because every day I change my mind about something. I have just played with some forest photos I had, and sigmoid gave me the most pleasing results. And on another beach photo, I got some lovely cloud detail by playing with the contrast and skew alone, which I had previously never managed to have much success at.
As I’ve gone somewhat off topic now, I want to go back to the general area of what this thread is about and ask you if you have noticed any difference with different raws and whether Fujifilm in particular plays differently with sigmoid because of the x-trans sensor?
I assume Fujifilm supply some raw conversion software with the camera - maybe you could use this to export a TIFF with as high bit depth as possible then edit that. I’m not sure, but this might be a halfway house between editing a jpg and a raw. Just a thought.
They do. And it can export to TIFF as well.
But the software is only available for Win and MacThings.
Have fun!
Claes in Lund, Sweden