Struct ReferenceAudio 

pub struct ReferenceAudio {
    pub samples: Vec<f32>,
    pub sample_rate: u32,
}

Raw audio data used as a reference for voice cloning.

The model will extract speaker characteristics from this audio and use them to condition speech synthesis. For best results:

• Use 3–10 seconds of clean speech (single speaker, no background noise)
• Match the model’s native sample rate (e.g. 24 kHz for Kokoro) or the library will resample automatically

Example

use any_tts::ReferenceAudio;

let audio = ReferenceAudio::new(vec![0.0f32; 24000], 24000);
assert_eq!(audio.duration_secs(), 1.0);

Fields

samples: Vec<f32>

Raw f32 PCM audio samples in [-1.0, 1.0].

sample_rate: u32

Sample rate of the audio in Hz.

Implementations

impl ReferenceAudio

pub fn new(samples: Vec<f32>, sample_rate: u32) -> Self

Create a new reference audio from raw samples.

pub fn duration_secs(&self) -> f32

Duration of the reference audio in seconds.

pub fn is_empty(&self) -> bool

Whether the audio is empty.

Trait Implementations

impl Clone for ReferenceAudio

fn clone(&self) -> ReferenceAudio

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for ReferenceAudio

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.