vis-core is the core of visualizer2. It contains the thread orchestration and a few analyzation tools. These include:
In vis-core audio input happens using the recorder. You can implement recorders yourself if the one you need does not exist yet. The interface is dead simple:
pub trait Recorder: std::fmt::Debug {
/// Return the sample buffer where this recorder pushes data into
fn sample_buffer<'a>(&'a self) -> &'a analyzer::SampleBuffer;
/// Synchronize sample buffer for this time stamp
///
/// Returns true as long as new samples are available
///
/// Async recorders (eg. pulse) will always return true
/// and ignore this call otherwise
fn sync(&mut self, time: f32) -> bool;
}As an example, take a look at the pulseaudio recorder.