Getting started with UliGesing / Darktable-Initial-Workflow-Module

I think Sigmoid should be easier for a beginner with DT 5.

I have edited thousands of photos with a dt 2-x-version and I was happy with the result most times. So I know a little bit about editing, on the other hand I am very new beginner.

I will have a look at the recommended vidoes.

Yes, I like his videos a lot.

Sigmoid and Filmic are both the equivalent of Basecurve. They are all tone mappers that do fundamentally the same thing, just in slightly different ways.

At the point where he uses Filmic, you can just use Sigmoid, if you prefer.

Please keep in mind, a beginner can make a photo with a few edits a lot better as the original. A pro will still see a lot things what can be done. I am happy for the moment to get the photo a little bit better.

I have thousands of mobile phone jpg photos with plant documentations. The photo shows how the plant changes and that is why the photo was made, not taken to be a ā€œbeautifulā€ photo. If I would like to have an excellent photo I would have used a DSLR. But there are some ā€œniceā€ photos whcih I want to improve a little bit.

Since the original is jpg from a mobile phone, I don’t expect too much.

In that case, probably don’t use Sigmoid or Filmic. They are intended for processing raw files, not JPEGs. Basecurve is probably not ideal either. For basic color and contrast adjustments, Color Balance RGB should be your main tool.

Keep in mind that almost all tutorials assume that you’re editing raw files, which in some ways is very different to editing JPEGs.

1 Like

Thanks, to learn Dt 5 I am playing with mobile phone raw-files at the moment. Is sigmoid a bad idea with mobile phone raw files (dng)?

Please see attached exifdata. For me there are strange entries, like Noise Profile

xiaomi-dng-exif.data.txt (6.3 KB)

If you use one of the standard workflows, darktable should not activate filmic/sigmoid/basecurve on a jpeg image, only on raw images.

If your phone dngs are actually raw images, you’ll probably want to use sigmoid. But then you also will have ā€œdemosaicā€ activated. If ā€œdemosaicā€ is not automatically activated, there’s a good chance your image is not a raw image…

1 Like

If you want, at dt_samples – Google Диск you will find some examples. There is always a dng file and a jpg file, taken at the same time. This is an option of Androids Proshot.

No need to download all, you have an online preview. Click on the folder name. Probably you see a file listing. On the right upper side there is a symbol for image preview.

These are examples which I want to optimize. It should get better, than the jpg of the camera, which is available for comparison.

I don’t want. For one, a preview will not tell me if a dng is a raw file.
For another, I hope you read the part you did not quote…

1 Like

Of course, I read everything. That doesn’t mean I understand everything at the moment.

Here is a screenshot with demosaic.

This is the same photo created as jpg.

Tried the screenshot jpg with sigmoid. I like what sigmoid does with the jpg, not perfect, but better than the original.

Interesting I like more what sigmoid does with the jpg than with the dng. Let’s say something in the middle of both would be fine.

Note the jpg-photo is already optimized by the phone, so both files look different.

I am playing to learn dt 5. Normally it is useless to take dng with a mobile phone. 50MB for a mobile-phone raw-file doesn’t make sense normally. This is to learn only, if it makes sense to take shots in an emergency case as dng with a phone.

Sigmoid is a tonemapper and while it will impart changes on a jpg, the jpg is an 8bit file that is already tonemapped… in your first image if you look at the wave form you need to drop the black level substantially and tweak the skew and slope in sigmoid and then see where you are. From there you might want to do local color or contrast edits and then finish with some denoise and sharpening again if needed… all of that is done in camera on your jpg…

1 Like

I thought I cannot use sigmoid with jpg-files. I did nothing else then (de)activating it with default values and compared.

This example is a backlight photo with my phone. Backlight is always horrible with the phone. Actually this is a better backlight photo

You wrote this…sorry if I misunderstood…

Anyway I am not on a PC with DT at the moment …but in ART which has ported the Sigmoid module and I do have access to that…here is your DNG with just default sigmoid in ART

So a decent result is possible…

1 Like

Thanks a lot. Strange your DNG looks closer to my jog.

This is my dng with simply activating sigmoid.

This is the jpg with sigmoid on.

I did it in ART…I would have to go back and check if anything else was enabled…I think capture sharpening and maybe local contrast might have been added and its a different editor/pipeline but if I had DT it would be easy to get the same result…just adding exposure in say multiply blend mode would likely make it look good… :slight_smile:

If you look at the waveform its a side effect of dropping the black point…go into rgb color balance and drop the global luminance in the 4 way tab to get a nice black…you will see a big improvement out of the gate…

1 Like

At the moment I am not so interested how I get an optimal result by tweaking the parameters for a special file.

I want to know which modules / options are recommended for my example files.

At the end jpg file editing from the phone is more important. If I want a good photo I am using my DSLR.

To make it clear, I want to spend only a few seconds when editing phone photos. With DSLR raw DNG editing takes as long as it takes, could be 15min. that I am playing, normally a photo is edited within 1 minute.

Well you have a high volume and no one size fits all is going to work so I think the best approach is edit one image to your satisfaction using it as a sort of key frame and then copy this to all similar images…move to the next key frame image edit and repeat. Phone images are highly processed already but whatever you do will vary with different sets of images so you need some flexibity

That’s exactly what I want to do. But first I have to learn which modules could be helpful in my case.

Is it possible with jpg to adjust sigmoid automatically with default values?

I am little bit confused, because I didn’t know that treatment of jpg is limited,

I am still learning by trying the things which https://youtu.be/ZUc6LOzg_Nk explains.