// the find
DevonCrawford/Timelapse-Auto-Ramp-Photoshop-Plugin
Analyze RAW images from a timelapse, and auto - ramp the exposure for manual changes of camera settings. The best way to achieve amazing results in difficult lighting.
A command-line Java tool that reads XMP sidecar files from a timelapse sequence, detects manual exposure changes between frames, and smooths them out mathematically so the transition doesn't look like a jump cut. Aimed at photographers who shoot day-to-night or sunrise timelapses manually. Solves a real, specific pain point.
The core algorithm idea is sound — detecting delta between two exposure settings and distributing it across N leading frames is the right approach to the holy grail problem in timelapse photography. Writing back to XMP sidecars rather than destructively modifying RAWs is the correct design choice. It handles shutter, aperture, and ISO independently via separate change classes (Change, ExposureChange, TemperatureChange), which suggests the math accounts for different stop relationships. The example GIF in the README actually shows a noticeable improvement, which is more than most hobbyist tools bother to provide.
Abandoned in 2020 with no releases beyond v1.0 — you're shipping a .jar with no build tooling, no pom.xml, no Gradle file, just raw .java files and a pre-built binary. No GUI and a CLI that requires knowing the exact file path of the first image is a meaningful friction point for the photographer audience this targets. No handling for edge cases like mixed camera bodies, corrupted XMP, or sequences with more than one manual change cluster — the debugging tip is literally 'comment out a method.' Temperature/color shift ramping is mentioned as a class but the README says nothing about it working, which suggests it may be half-finished.