scanplan: A Python repository from Peetz0r

This is a messy set of messy script that I never really intended to put in git until now.

It assumes your scanner is a Canon Canoscan 9000F mk2 (which is not available anymore). It's SANE driver name is hardcoded and there are a bunch of hardcoded sizes and offsets for it's 35mm film holder.

Also the script are a messh of shell and python. The shell scripts depend on fish for no good reason other than that it's my prefered shell and I copypasten them from interactive sessions. They should someday become posix sh scripts. Especially since they're mostly scanimage oneliners.

You should collect 16 bit tiff samples of unexposed bits of each type of film stock in a directory and run the film-base-to-txt script there. It'll create text files containing the average color of those.

Every album directory should contain a file film-base which should be a symlink to a txt file created by film-base-to-txt. This will be used to properly compensate for the film base when inverting, without relying on auto(mis)detecting said colors.

When scanning regular 35mm film, you should:

create a new directory
run 35mm-neg-scan for strips of 4 frames , or 35mm-neg-scan-6p for strips of 6 frames
use the button on the scanner to scan the strips of this roll
if the film stock isn't in your film base catalog, then you should
- probably while scanning the 2nd "page", aka 3rd and 4th strip: open out1.tif in GIMP or similar
- crop out an unexposed bit of the film, export that as a tiff file
- run film-base-to-txt
- you should also do this for BW film, and you should make variations for pushed/pulled/expired/etc film
create a symlink named film-base to the film base txt file
if it's BW film, also create an empty file called bw (lack of this file tells the scripts that it's color film)
when the physical scanning is done, you should run 35mm-neg-scans-to-pos-strips
look at the strip*.tiff files with an image viewer
then run 35mm-strip-to-frame
now look at frame*.tiff and see the mess you made

I'm still working on improving the (embarrassingly simple) code that cuts out the frames from the strips, and the color correction, and everything else really.

I have scanned around 60 rolls of film and most of it is usable, but a lot of them have issues woith cropping and/or colors. Also, those that are mis-cropped are also usually mis-exposure-corrected because the black film base should (but itn't) cropped away.

However, since I always keep the out*.tiff files from scanimage around, I can declare the "scanning" finished.

I have also scanned around 240 slides with the oneliners in dias.txt ('dia' is Dutch for 'slide') but since those used much simpler hardcoded cropping and didn't need any color correction (only basic exposure and gamma corrections) I regard those as a seperate (and finished) project. But I included my notes anyway.

This is a picture of the scanning process. The picture itself is taken with a Yashica FX-3, Tokina RMC 25-50mm F/4.0 zoom lens (probably at 50mm and F/4.0), on Fuji C200 film, and has also been trough this same process before ending up here. The gamma correction used was (0.50, 0.42, 0.40) which will probably change at some point.

Peetz0r/scanplan