p.2 #1 · Convert scanned color negatives to positive pics?
Well, I've been playing around with software and learned a few things. First, my issues with Darktable's negadoctor. It seems to be picking up the base film color wrong. It looks too dark. When I open an image with MS Paint and select the color around the sprocket holes I don't get the same ratios between RGB as the percentages that Darktable reports. If I try to compute RGB from the NDr percentages based on 0-255, they don't come out correct. Maybe Darktable is assuming more than 8-bits per color. Regardless of why, that would explain why the end result has an overall cast, typically blue.
I tried Film.Scan.Converter.exe and it seems to work correctly, and runs on Windows instead of Linux. Fast, accurate, much more realistic result. The base color pickup agrees with what MS Paint says. I haven't played with all its features yet but it will be my go-to negative translation program for now.
p.2 #2 · Convert scanned color negatives to positive pics?
A quick photo essay. I put this collection together for a friend who asked how it was done. I'm sure most of you know all this already, but I didn't when I decided to awaken my film camera. Maybe this will be useful for whoever else does the same. And if not, it is at least slightly entertaining.
A dedicated scanner will do all this for you. I didn't want to invest in one because I wasn't sure how much I'd use it. Instead I found a convenient rig and a used macro lens at my local Hunts photo store. scanner by James Muller, on Flickr
The scanner's light source is an even white and the lens has an image ratio of 1:1. So the camera can take full frame pictures of film. (My original goal was scanning old slides, not color negative film. A friend 3-D-printed a slide holder for me.)
Shooting color-positive film is easy, you just take a picture of the picture, no other processing necessary. Color-negative requires converting to positive. I had software with a "negative" tool so I thought it would be easy. What I didn't know was that the base color of the film had to be removed, something my software didn't do. I had to ask here to learn this. (I am grateful for the info!) For yucks, here are some results, with one final interesting lesson:
The original color negative. The base color is visible in the strip on the right edge, the negative of unexposed film which one might have naively guessed to be negative white, positive black. When I first scanned the film I tried to eliminate the edges, not knowing that they would be necessary to find the base color. I also didn't realize at first that the camera's white balance should be that of the light source, not auto, lest it try to correct for a perceived white balance from the image. (Surprisingly, the edge darkness isn't constant top to bottom. That may be because I used a jpg instead of a raw image, and the camera had done some processing. Or maybe the developed film was just like that. It doesn't seem to have affected the final result, but it means the base color has to be picked from the edge carefully.) ColorNegativeScan by James Muller, on Flickr
By playing tricks with the software I saved this negative image with just the base color removed. The right edge is now neutral white/gray, with the brightness varying slightly top to bottom. ColorNegativeBaseColorRemoved by James Muller, on Flickr
When both processes are applied, the final image appears normal, as if by magic: FInalResult by James Muller, on Flickr
Reversing the order of the two processes, base color removal and negating, didn't make any difference. The end result was the same.
Yeah, I know, this is all obvious for most of you. But I had fun learning it.
