I’ve been trying to switch from Lightroom 6.14 to darktable for about a year and keep stumbling over the default treatment of raw files that look nothing like the OOC JPEGs and nothing like what I see. I can often get close to what I need, but it takes a time and the adjustments don’t carry over well from photo to photo. Since I can take literally hundreds of shots on any day, maybe half or more of which I’d like to keep, it’s painful to work so hard to get what I need. The screenshot shows a photo in Rawtherapee on the left, and then the camera JPEG and raw ORF in darktable on the right. Lightroom does about the same as Rawtherapee. The camera JPEG is maybe a bit too saturated, and the RT version is a little undersaturated but close enough. For this, I’ve enabled the camera basecurve presets, and set the pixel workflow defaults to display-referred, but that really has little effect on the result. I honestly can’t figure out how to make any progress on this.
I’d switch to Rawtherapee, but it’s handling of keyword metadata is pretty weak for my needs.
Cameras: Olympus EM1 mk II and EM5 mk II
Can you share the raw…also note that even the base curve has color preservation modes…I think the default these days is luminance…maybe try no or one of the others??
Edit you are allowed to upload the raw directly using the upload button…just add a statement and cc license…
You can copy and paste from here…
Thanks. The raw is below. To be clear, what I’m looking for is a way that will get close for all images, not just this one. For example, increasing exposure to about 1.3 works for most normally exposed shots. I’m pretty lost at getting colors and highlights to work well easily.
20221018_102126_0194.ORF (13.4 MB) License: CC BY-SA Craig Denson
I walked through the introductory Process doc [again - after updating to 4.0], and discovered the ‘basic colorfulness’ presets. That, along with exposure settings, gets me 80-90% of the way home. Still working on highlights and contrast/sharpness, but am feeling much better.
Ill have to extract the jpg from your raw and use that to compare…but this is what I get with default filmic…there are some new suggested defaults by the author of the module…
Just adding default LC and default preset basic color
Flimic is 0.7ev contrast 1 latitude set to zero and highlights and shadow shoulder settings in filmic at hard…
Keeping the same and using spot exposure which I set to 50 % for a starting point … bumps the image to about 1.2 ev
You then get this
20221018_102126_0194.ORF.xmp (7.2 KB)
Modern WB with Base curve with new suggested EV bump of 0.7 EV and base curve with preservation set to no…you could try the others but this is really close as far as this image goes…
I have an EM-5 II too, so here is my attempt with my usual scene-referred workflow (not trying to replicate the jpg):
Other than filmic and the usual things like WB, LC, CA and denoise, I use local contrast with clarity preset (sometimes up the detail slider a bit), and colorbalance RGB with the basic colorfulness standard or vibrant presets. That gets me good edits for 90% of pictures, unless I need to adjust HDR with tone equalizer.
PS: WB is set “as shot”. You may have to play around with it, depending on the lighting.
Yes, these are huge improvements, thanks.
Where are settings for WB w/Base curve The conflict between using the WB module and the Color Calibration module is confusing at best.
Hi Craig,
I have downloaded your image and will play with it. I have an Olympus TG6 for diving and hostile environments. I have found it very hard to match the colors from the OOC JPG. This only occasionally presents a problem to me as the results I get from DT’s processing of the ORF file usually results in images that are far superior to the OOC Jpegs. However, just some times the OOC Jpeg has a funky blue which is probably not correct, but can be appealing. I just can not replicate it in any program, not just DT. The closest I get is Olympus’s own OM workspace, but even that is not perfect.
My understanding is that you can stick with WB and don’t need color calibration, unless you need to make local changes to the WB, for example in cases where you have two different illumination sources in different parts of the image. The experts can correct me on this one, if I am wrong (I have only used DT since April of this year).
I downloaded it and for sure with some changes I could have gotten it to the jpg but it did seem to come out redder using scene referred. I still liked the edit but I was trying to respond to the question in the post… I think you are showing the same…shifted to red…This may be more of the new color space and gamut handling in the more modern modules…But the basecurve with the right exposure and preservation mode is very close…adding LC and rgb CB make for a nice image but push the saturation and contrast a bit past the JPG of the OP, that’s not saying those aren’t a nice improvement to the image depending on the eyes of the editor…
And if you don’t mind a couple of small adjustments then setting v5 to no and working the saturation you also get a very decent match with the orange…
I have done what is for me a simple, straight forward and fast edit in DT. Does it achieve what you desire? If not why? Could you post your JPG here as a comparison?
20221018_102126_0194.ORF.xmp (6.5 KB)
You can extract the JPG…you will see if you do that the default or defaultish V6 scene setting will yeild an image that is more red or maybe less yellow than the ooc JPG at least on my screen
Hi Craig, I personally don’t feel compelled to use the color calibration module if the white balance module has done a good job. The edit I did used just the WB module.
I don’t think there is any issue using CC if you know the D65 values for your camera are solid…if not then maybe legacy is safer esp if there seems to be an issue that you can for sure tie to WB. I think many people don’t get the link between WB and doing the CAT in CC and have issues…one common one is that the hue chroma sliders get moved in spot color and then people start using the auto picker attempting to wb from the image when as shot is not right and they get weird results since the sliders are persistent image to image and should be at 50 L 0 H 0C for the picker to work…little things like this often go undetected and give people reasons to doubt CC
Todd, can tell where I should look to push the red a bit more orangish?
On a side note, I don’t know how to use the attached .xmp files to figure out what to do.
Okay quick one first…just go to the lighttable view and select your image…create a duplicate… THen go to load sidecar… select the sidecar file in the dialogue and apply it… then you can study the image…In the base curve example I gave you above was the closest match without some work. I think if you try that you will see… the scene edits seem a bit red… THere are a few ways you could make the shift… I will try to create an example…
This is the xmp using filmic v5… the filmic settings with exposure do most of the matching
20221018_102126_0194.ORF.xmp (9.4 KB)
I found the two olympus base curves were good as well if you set the chrom pres to no… The first one needs 0.7 ev to match the second one is a match with no exposure added…
20221018_102126_0194_01.ORF.xmp (10.2 KB)
Many, many thanks! Totally non-obvious to me, but this works great.