Optical Audio to Digital WAV

Optical audio on film has been a critical part of the presentation of motion pictures since the 1920s when variable-area/variable-density was first introduced. Converting scans of film can be difficult, and extracting audio moreso. There has been a very good tool for the past decade, AEO-Light, that is popular to use to convert the audio. However, the codebase is in C++, and it is difficult to understand how it works, and to add new features to. As a result, this project is to create a modern Optical to Digital conversion tool, written in python, that can convert optical audio of all shapes and sizes (possibly other formats too like Dolby-Digital or SDDS) and Dolby A/SR easily and effectivly. Also, another important goal is that anyone can understand how the audio conversions work and can extend the functionality into their own projects too.

Useful Links

• Sound-on-film https://en.wikipedia.org/wiki/Sound-on-film
• Dolby Stereo https://en.wikipedia.org/wiki/Dolby_Stereo
• Dolby SR https://en.wikipedia.org/wiki/Dolby_SR
• Dolby Digital https://en.wikipedia.org/wiki/Dolby_Digital

Progress

A basic web-based application now exists that converts individual images that contain Optical Audio to a .wav that is subsequently downloaded. More features are being added. Also, the KylesOpticalDecoder.py can be used standalone in CLI, allowing for other programs/scripting to run it in a pipeline.

Kyle's Optical Decoder

Python CLI tool that can extract film optical audio

Prerequisites

• Python 3.10+
• Node.js 18+

Setup (normal local)

First time: start-dev-fresh.sh Afterwards: start-dev.sh

Setup (development)

Backend (Python):

python -m venv .venv
source .venv/bin/activate   # macOS/Linux
# .venv\Scripts\activate    # Windows

pip install opencv-python numpy natsort scipy fastapi uvicorn pydantic

Frontend (React):

cd frontend
npm install

Running

Start the backend server:

python server.py

Start the frontend dev server (in a separate terminal):

cd frontend
npm run dev

Then open http://localhost:5173 in your browser.

CLI Usage

KylesOpticalDecoder.py can also be used standalone from the command line:

python KylesOpticalDecoder.py --help

Theory of Operation

TODO

License

This project is licensed under the GNU General Public License v3.0. See LICENSE for the full text.

Contributors

• Kyle Mikolajczyk
• Will Dirkschka
• Ben Peters
• Thomas Piccicone (35mm Scan Examples)