[I tried searching for info regarding this, but couldn’t come up with something (which I find very surprising, as I cannot be the only one having this problem, can I?).]
I read a lot about DT and am well aware of its new scene-referred workflow. My concern is from the usability side: Opening a RAW file has an initial setup that is very flat and dull looking - nothing like e.g. the in-camera JPG.
My workflow is usually “get it right incam 98% of the time”, so only 2% need to be touched in a RAW tool (but I still export everything from RAWs to have it consistent).
I tried getting the image as close as possible to the incam JPG (I thought, I could then save that edit into a style and apply everywhere from there on), but it’s not only exposure and constrast, but also the colors are slightly off. Filmic’s “middle tone saturation” helps a little, but still, after 30…40mins of fiddling around, the incam JPG is definitely still preferable.
Files in Lightroom/DPP start off very close to the incam JPG. So I thought “ok - maybe that’s some secret proprietary large corp sauce”. Yet, I now tried ART (another RAW therapee) and the initial look of the RAW file is almost indistinguishable from the incam JPG.
So I clearly must be doing something wrong in DT here…
Any help would be greatly appreciated, as DT seems to be much more flexible in editing.
That would be a very weird decision, as having a good point to start from saves a tremendous amount of time. If from 100 images 10 are not a 100% to my linking, I would only need to touch those 10 (and those would be already close), where now I have to completely ‘redo’ 100 images for no additional benefit (and as said, even after trying this for 40mins I couldn’t even match the ‘default’ JPG).
I’m not going to rehash the arguments here - suffice to say that people will never agree on what a “good” starting point is. Variations on this question are asked frequently on these forums so you can search/browse for such discussions. If you’re intention is to reproduce the JPEG that will always take a long time (it will always be “not quite” the same) but with experience most people can get good edits out of darktable in under 5 minutes (I think my average is 2-3 mins per photo).
Well, if the incam JPG isn’t a “good starting” point, one should probably consider buying a camera from a company that knows what it is doing.
But if with DT I need to edit 100 pics instead of 10, this would take several hours longer compared to “initial startup is similiar to JPG” (even if you cannot match that 100% - but obviously ART/Lightroom & Co can do exactly that to maybe 98% - so ‘good enough’).
Starting that far off of something ‘usable’ seems unnecessarily labour-intensive - or is there a benefit to this that I do not see?
We all have been there.
DT, like all photo/video editing SW, has a steep leaning curve.
Initially, I struggled a lot to even get on par with the OOC jpeg. Then, slowly, I started to learn and now my edits are so much better than camera jpegs that I don’t even shoot RAW+JPG anymore.
And it doesn’t take too long either: with the new DT scene referred workflow, I can get a better result than jpeg in minutes. Moreover, in a photoshoot I normally edit one picture for each change of light and for the rest of the picture it’s just matter of copy/paste history and final touch-up of exposure / crop.
BTW, have a look to the PlayRAW section of the forum to see how editing can improve a pic
One difference between the camera jpeg and dt/filmic is that my camera “throws away” about 1 stop in the highlights when creating its jpeg… That means that I get one stop more leeway in the highlights, quite nice with landscapes, and enough reason for me to prefer dt/filmic…
And, depending on your images, you can have a series with similar lighting and exposition. In that case, you can copy the treatment of the first of a series to the others to get at least a good starting point. Or you can save (part of) the treatment as a style for future reuse. See the manual for the mechanics.
I had a look at the “let’s create an SOOC-style” thread. At one point someone only used the ‘Color Contrast’ module (filmic disabled) to create a good match.
This confuses me. I thought, DT now operates in scene-linear? So I thought “filmic” does the conversion from scene-linear to RGB? Therefore I am surprised, that one can get a decent looking result without that transform at all.
Getting a “nice” result in DT is indeed rather quick. That’s why I am using it so far (according to a quick scan, I have used DT to edit ~20k images so far).
Yet, I still struggle to get colors that are “as nice” as in the SOOC jpgs. Dynamics are definitely better, but colors not so much.
That’s why I was searching for a “more like SOOC jpg start” to have the colors dialed in and be able to start from there.
You can edit images in Darktable bloody fast, if you use shortcuts, autoapplied presets and custom styles. I can edit up to 360 images per hour, if I don’t have to do any local edits (retouching etc). I have written about it here
In the bad old days when we only had the display referred workflow (basecurve + shadows & highlights module), the auto-applied preset for the basecurve got me pretty close to the in-camera jpeg of my Nikon D750. But if I wanted to do something more than just crop the image and adjust the exposure, I had to really fight with the program. But if that is OK with you, you can set up the latest Darktable to work like before the scene-referred workflow was introduced in version 2.7. But I strongly recommend learning to use Darktable the modern way.
So one thing I would say is think of your Canon camera as a guy named Canon. Your jpg is not the ground truth but the effort of a skilled guy named Canon but with bias and color judgements made with the scene lighting to produce an image. There is another guy out there named Sony a guy named Nikon etc …each guy will have a skilled edit but not the ground truth… Chasing the edit of your camera thinking it is the ground truth will likely always lead to a lot of unnecessary effort. You will see it said over and over…if you want the jpg then just use the jpg but if you want to take that raw data and see what you can make with it…trust yourself and free yourself from the comparison and just make your edit…You have paid for a guy named Canon to edit your photos and spit out jpgs so use them if they are good but IMO chasing the jpg in raw editing is not productive and it has been debated over and over from many angles…to somewhat mimic the way an image is initially presented in ART you would need to use the base curve in DT as a starting point.
While you are absolutely right on “the overall edit”, I would argue that there IS a ground truth part on an image - e.g. if a red flower becomes a violet flower.
And while I would like to be able to influence e.g. the overall dynamic range/contrast/a.s.o. of my image, I would also like to stay color-accurate as much as possible while doing so and not have someone with a red t-shirt ending up with a blue one.
So while there is no ‘correct’ contrast, there is definitively ‘correct colors’ - e.g. looking at a photograph of my color chart on my calibrated monitor.