exadatum/rnnoise repository overview

⁠RNNoise Docker

RNNoise is a noise suppression library based on deep learning. This Docker image provides an easy-to-use setup for applying RNNoise to reduce background noise in raw PCM audio files.

⁠Features

• Converts WAV files to raw PCM format and back.
• Utilizes RNNoise for high-quality noise reduction.
• Handles both raw PCM and WAV files seamlessly.

⁠Usage

⁠Running the Container

For Raw PCM Input and Output

If your input and output files are in raw PCM format:

docker run --rm -v $(pwd):/data rnnoise /data/input.raw /data/output.raw

For WAV Input and Output

To convert WAV files to raw PCM, process with RNNoise, and convert back to WAV:

docker run --rm -v $(pwd):/data rnnoise /data/input.wav /data/output.wav

⁠File Conversion

The container automatically handles file conversion:

• WAV to Raw PCM: If the input is a WAV file, it will be converted to raw PCM before processing.
• Raw PCM to WAV: After processing, the raw PCM file can be converted back to a WAV file if the original input was WAV.