As soon as I switch it on (e.g. landscapes with snow scenes) the picture gets flat and dull. I don´t manage to get back the original look which seems much better to me.
So do I have to use these tone compressors on every picture or will my picture also be fine without them?
I personally use sigmoid as my default tone mapper as it gives nice results straight out of the box. Sometimes I go for filmic with high dynamic range lighting, but then color needs fixing in the color balance rgb modules. The link kofa provided would be a good read and I demonstrate three different methods for the same picture.
I think It really depends on the specifics of the particular scene, both composition and lighting. I’ve had a very few images where I didn’t think I needed any tone curve, mostly dark scenes with a subject lit by some shaft of light. I think the important thing is to get “white to white”, usually with some exposure adjustment, then decide how you want to distribute the rest of the tones in the scene with some sort of non-linear transfer function (curve).
Do note that you’ll always have at least one tone curve in the processing chain, the one that is applied when the image is displayed or exported to a file for rendition. ‘sRGB’ is really two things, a color mapping and a roughly 2.2 gamma tone curve. What you’re doing with filmic, sigmoid, or other tone curve mechanism is messing with the data between linear and that output transform.
I get that, but for my workflow that could push resolved highlights into saturation. If I anchor to white, I can then lift the mid-tones with the curve.
That is not what I am saying. I personally always use a tone mapper. The old school tone mapper of base curve is appropriate, but I do feel the filmic and sigmoid modules are a more modern take on tone mapping. In my view sigmoid is easier for a new user of DT as often it just works out of the box and doesn’t need tweaking. Filmic is a more complex module which doesn’t in itself make it superior to Sigmoid. For me most edits are based on using sigmoid, but there are times I decide to compare the result from filmic and pick the best way to go after a comparison.
I feel if you are wanting to avoid a tone mapper you are using an approach that I would not be supportive of. Tone mappers have an important role to play.
Ya but I think that is more of a display referred mentality… In DT when you set midtones many images are going to blow the highlights until you then tone map. The compression can at times kill the details but there are also ways to manage that… If the image has lower DNR then you don’t need to rely on tone mapping for the image but this is the way DT scene referred editing is intended as scene white is often thought of as infinite…
Try this , if using sigmoid go to the exposure instance and click the autopicker…try the whole image which is default or select an area that should be in the midtones… Then tweak skew and maybe contrast. Note the ev setting …you might find its higher than what you would normally use… If filmic then do the same but use also the autopickers on filmic ie for white then black…is that similar or different to what you get… This contrasts say having an image where the brightest pixels are at 2/3 of display so you add a bit of exposure to set white and then you try to lift midtones to tone the image…often it’s hard to get nice midtones editing this way
Thank you for all your replies. Kofi asked me which files I am processing and maybe that was the solution.
My workflow.
AWR file to dxo pure raw (raw to tif)
resulting TIFF File to darktable
Now when I process the raw file directly with darktable I don´t run into these flat image results with the tone mappers. Or at least I can tweak them nicely.
Using filmic or sigmoid with the tif files from dxo pure raw makes the whites look grey (I get the flat look which I do not manage to get rid of).
So could it be that dxo pure raw uses a kind of tonemapping when processing them to tif files? Then I would not need to tonemap them in darktable anymore?
I do need dxo pureraw because of its superior denoising with my high iso images.
Now I also tried to convert the raw files (with dxo pure raw) to dng instead of tif files. This also seems to make it easier. Maybe that is the route for me to go. I always thought tiff was superior to dng (more data).
Thanks again
simon
In my view just process the raw files in DT. A tif file is no where near as good to edit. I doubt that dxo is worth the effort. DT has excellent noise reduction and sharpening options. If dxo was so good it would do the processing from start to finish in my view.