🚀 The feature
On Android, replace LiveKit’s default microphone input with a custom audio pipeline that:
• Captures microphone audio using native APIs (AudioRecord / Oboe)
• Processes audio buffers using the shared JUCE DSP engine
• Publishes DSP-processed audio to LiveKit as the audio track
LiveKit’s transport, encoding, and room logic remain unchanged.
Motivation, pitch
Android currently publishes raw microphone audio directly to LiveKit. To support client-side voice processing, we need to intercept audio before publishing.
Implementing this natively ensures low latency and real-time safety, while JUCE guarantees consistent DSP behaviour across platforms. This change is limited in scope to input replacement and does not modify LiveKit internals or server-side behaviour.