Flashcut

A simple Flask web app that analyzes an uploaded audio file for onsets and generates cut segments; optionally renders a video timeline using FFmpeg.

Deploying to Railway

• This repo is ready for Railway out of the box (Nixpacks). It uses a Procfile and binds to $PORT.
• Python version is pinned via runtime.txt.
• Dependencies are pinned in requirements.txt for reliable builds on Python 3.11.

FFmpeg (optional, for rendering)

Rendering videos/images requires FFmpeg. If it is not available, the app still works for analysis; rendering will be skipped with a friendly message.

To include FFmpeg on Railway, enable it in the Nixpacks settings in the Railway dashboard (Install Packages → ffmpeg).

Environment

Optional variables:

• FLASK_SECRET — Flask session secret (defaults to "dev-secret").
• Python version is pinned via runtime.txt (3.11.9). No extra nixpacks.toml is required.

Health check

• GET /health → { "ok": true }

Notes

• Max upload size is 500 MB by default.
• Gunicorn is configured with a single threaded worker suitable for CPU-bound analysis; increase if you expect higher concurrency.