sendspin_auracast

Receive Sendspin PCM audio and broadcast it over Bluetooth LE Audio (Auracast) using Bumble.

Installation

pip install sendspin_auracast

For local development:

python -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install -e ".[dev]"

Usage

Broadcast audio from a Sendspin WebSocket server:

sendspin-auracast ws://localhost:8927/sendspin

The client advertises itself as a Sendspin player, receives PCM audio chunks, encodes them as LC3, and broadcasts them as an Auracast stream.

Use a custom Bumble transport or broadcast name:

sendspin-auracast ws://localhost:8927/sendspin \
  --transport serial:/dev/ttyACM0,1000000 \
  --name "Sendspin Auracast"

Broadcast with encryption:

sendspin-auracast ws://localhost:8927/sendspin --code "my-password"

Add manufacturer-specific BLE advertising data:

sendspin-auracast ws://localhost:8927/sendspin --manufacturer-data 0x0059:010203

Tune startup latency:

sendspin-auracast ws://localhost:8927/sendspin \
  --sendspin-buffer-ms 50 \
  --presentation-delay-us 20000 \
  --max-transport-latency-ms 40

The bridge defaults to a 100 ms advertised Sendspin player buffer. Smaller values can reduce the delay after pressing play, while larger values may be more reliable on busy hosts or noisy Bluetooth links.

For debugging, print received chunks instead of broadcasting:

sendspin-auracast ws://localhost:8927/sendspin --print

To write raw audio bytes to stdout instead:

sendspin-auracast ws://localhost:8927/sendspin --raw

Development

Run tests:

pytest

Run linting:

ruff check .

Build the package:

python -m build

Prepare a release locally:

python scripts/prepare_release.py 0.2.0

Publish to PyPI:

twine check dist/*
twine upload dist/*

Releases

Use the GitHub Actions workflow Prepare Release to create a PyPI release. It:

• updates pyproject.toml to the requested version
• runs lint, tests, python -m build, and twine check
• commits the version bump and creates a v* tag
• pushes the commit and tag
• creates a GitHub Release

Run the Publish to PyPI workflow manually when you want to publish. Provide the release tag, such as v0.2.0, and it will publish that exact version to PyPI. Make sure the repository's PyPI trusted publishing settings are configured for the release environment before using it.