Audio buffer stream

[alnitak]

Description

close #144

This new type of AudioSource is created by subclassing the SoLoud C++ AudioSource class.

It works by initializing it with its parameters, then adding PCM audio data, and finally signaling that no more data is incoming:

• setBufferStream returns the new AudioSource. Using its parameters, you can set the PCM audio format, the maximum memory usage, the buffering time required to resume playback (when paused to wait for more data), and a callback to handle buffering events.
• addAudioDataStream* adds PCM audio data to the stream. Helper methods are available to add signed/unsigned 8-bit integers, signed 16- and 32-bit integers, and 32-bit floats.
• setDataIsEnded marks the end of the data stream.

The returned AudioSource is compatible with filters, oscillators and all other tools.

Use-case

• streaming PCM audio data from a websocket and playing it while receiving data: example/lib/buffer_stream/websocket.dart is used for testing. It receives PCM data from a local websocket, provided by the websocketd tool. The Flutter app uses ffmpeg to stream PCM data over the websocket via websocketd, allowing you to simulate slow download speeds and buffering behavior.
• example/lib/generate.dart generates audio and fills it into the AudioBuffer.
• the Cartesia TTS Flutter project demonstrates how to use Cartesia AI APIs to send text and receive voice audio data.
• you can also use flutter_recorder to capture PCM audio from the microphone and load it into the AudioSource for flutter_soloud.

Type of Change

• ✨ New feature (non-breaking change which adds functionality)
• 🛠️ Bug fix (non-breaking change which fixes an issue)
• ❌ Breaking change (fix or feature that would cause existing functionality to change)
• 🧹 Code refactor
• ✅ Build configuration change
• 📝 Documentation
• 🗑️ Chore

[filiph]

I have some suggestions but this looks great, thank you!

[alnitak]

I thought it would be interesting to have a new method to reset the BufferStream. Something like
void resetBufferStream(AudioSource source)
which removes the ended flag and frees up the old audio data allowing the developer to add new audio data from scratch?

[filiph]

I thought it would be interesting to have a new method to reset the BufferStream. Something like void resetBufferStream(AudioSource source) which removes the ended flag and frees up the old audio data allowing the developer to add new audio data from scratch?

To be honest, I'm not the intended audience (I don't foresee needing buffer stream) so I'm not the right person to ask. Maybe the folks who're asking for the feature might tell you? Otherwise, in general, I'm a proponent of YAGNI.

[prateekmedia]

@alnitak The resetBufferStream method can be useful when audio stream needs to be interuppted due to any reason and new audio stream needs to feed.

[alnitak]

Thanks @prateekmedia,

I was thinking of the resetBufferStream because I was trying to figure out how to do a walky_talky like app or any other workflow like this:

1. setup an AudioSource with setBufferStream
2. receive data from the net and add them to the AudioSource while playing it
3. when data is ended call resetBufferStream and go back to 2 instead of disposing of the AudioSource and going to 1

Anyway, let's flush this PR and shortly we will think better about what will be useful to add.