Photography

Introduction

I only started photography as a hobby about a year ago, and already I have about 50gb in my photo library. I shoot RAW, and evidently RAW photos take up a lot of space. I decided to test whether it was possible to compress these files to save space. Specifically, I wanted to know if I could use a compressed archive, like a ZIP file, to reduce the size on disk of my photo library.

If you would like to know the procedure used to get these results, see part 1 (Procedure).

Reintroduction

I’ve heard rumors that developing color film (C41) requires holding the chemical temperature constant ±1℉ for fear the color shift demons will recolor your film in strange and unusual ways. I developed slices of the same shot at different chemical temperatures to test that claim.

Results

All pictures were taken at f/8, 1/1000s, ISO400, including the digital control. ISO was forced by the film speed I had, aperture was selected arbitrarily, and shutter speed was set by using the recommended meter reading of the FE2 film camera. The digital control histogram is scaled to the image minimum and maximum. The analog histograms are scaled to the minimum and maximum supported density of the scanner. This means the analog histograms can be compared with each other fairly but the digital histogram has a different scale. All analog pictures were scanned with auto-exposure. The scanner auto-mapped each RGB channel film densities to image minimum and maximum.

If you aren’t interested in the procedure and just want to see the results, skip to part 2 (Results).

Introduction

It seems to be common consensus that when processing C41 (color negative) film at home, you need to control the temperature of the chemistry very carefully. I’ve heard that if your developer is off by 1℉, You’re going to get some bad color shifts. Other posts claim 5℉ is the threshold. Looking around online, I’ve found many claims about color shifts, but no experimental results. One post claimed that the color shifts are correctable in a digital workflow, but otherwise if optically printing will be uncorrectable due to highlight and shadow color shift being independent.