@wpferguson I used the Negadoctor.dtstyle that you kindly shared , it resulted in strange and very dark colors. I finally tracked it down to the setting for “raw black/white point” module settings. When I changed it to passthrough color were came inline and I could continue with the editing. This raised a question for me. What is the role of this module within the workflow for the negative films? I played with it and seems the sliders more or less do same thing as the Color Calibration, and negadoctor modules. I only can guess that its uses are for the initial backlight at the time of scanning the negatives. In my case I am using a dedicated negative film scanner which randomly and annoyingly message pops out saying “OpticFilm 8100 color matrix not found” ( OpticFilm 8100 is the name of my scanner ) it causes no issues that I can see and then disappears. I also liked the input profile to linear Rec2020 RGB in “input color profile” module. Colors seem to be more rich and saturated and more room to tame it later
All the negatives I tested on were created using a camera, therefore the raw black white point is required to get the correct colors. I didn’t think about a scanner. In that case pass through would be appropriate.
Thank you.
I spend a lot time doing trial and error and comparing your method. I am very impressed. I had this negative image that was very difficult and was producing strange background cast. I do not know why it is an old negative and do not remember the parameters that took it under. Your method with linear REC2020 given me more latitude to correct it as much I could, still there but much less than before.
I have not tried this before and surly will do. I am surprised that you recommending it BEFORE negdoctor.
Your method is unorthodox but pleasantly produces much better results
and is simpler. One just had to become more comfortable with using negdoctor.
For sure is an eye opener.
Thank you again for sharing it.
I’m glad that you liked the results with REC2020.
Yep, it’s important that the RGB primaries modules is placed before negadoctor. I use it because if REC2020 is way smoother as an input profile compared to the standard matrix, especially in the greens, it can introduce some hues as well. But you can do without the RGB Primaries, and I don’t know if the values that I gave you with the screenshot, that apply to my A7III, will give pleasant results with your scanner.
RGB primaries doesn’t have the same effect if placed after in the pipeline. First of all, if it’s put after negadoctor, obviously the color changes of RGB primaries would be reversed because of the inversion that negadoctor does. And anyway, I don’t like the effect of some modules (RGB Primaries, color calibration) placed after negadoctor, it just doesn’t behave the same.
I’m not the most technical guy around, but it seems to me that the files coming from negadoctor behave more like a tiff when using the modules placed after negadoctor in the pipeline. So I try to make the most corrections possible in negadoctor, either colors (shadow cast and the very important highlights white balance) and tonality (print properties). For people who scan with a scanner, you’re maybe dealing with a tiff file anyway ?
Why do I sometimes use AGX to finish : after negadoctor, if I’m not completely happy with the exposure, an AGX with autolevels doesn’t move the general tonality in a weird way despite being a tonemapper (once again, I say that by experience, not pure technical knowledge) : it allows me to set and secure the white and black points, and I can use the pivot relative exposure as an exposure compensation without burning highlights or creating black areas in the shadows. It’s more effective in my opinion than an exposure module placed after negadoctor. If I use an exposure module, then I have to correct again the white and blacks, etc.
Try for yourself guys with the file that I posted here or one of your owns. Activate AGX with autolevels, maybe increase the white point a little bit (it can be sometimes slightly conservative) then move the pivot relative exposure. It’s fast and efficient.
This workflow may not be perfect, but it’s the way I scan thousands of files. It gives me consistent results so far, even if, like you, I always look for more speed and consistency.
Sure will do.
luckily my scanner produces raw “dng” files, but I understand what you mean.
The only difficulty sometimes that I have, which I think with practice may become easier is the sliders in negdocter module are very sensitive.
I have bookmarked your post and will use it as I come up with a work routine for the upcoming conversions. It is a redo project since the 1st time around over 20 years ago did not know much about digital conversion. As I get better at it, would like to pull out my old beloved camera and lenses and shoot some with the film.
@Ayoul I have been using and testing your method and consistently get the excellent results. My only minor change is that I use filmic with auto tune levels and quickly get the right color mood reminsent of film colors. Your method is now my goto default method for my negatives. Quick and easy.
Thank you.
I’m glad that my method was useful. You can indeed replace Agx with filmic, it may work better or worse depending on the picture. I forgot to mention a few things about Agx :
- shoulder and toe power should be at 1 by default
- contrast at 2.8
- you can use the pivot target output to change exposure instead of the pivot relative exposure, it can lead to softer results