Struct AudioFrame 

pub struct AudioFrame<'a> { /* private fields */ }

A frame of audio samples with associated sample rate.

AudioFrame is the standard audio input type across the WaveKat ecosystem. It stores samples as f32 normalized to [-1.0, 1.0], regardless of the original input format.

Construct via AudioFrame::new, which accepts both &[f32] (zero-copy) and &[i16] (converts once) through the IntoSamples trait.

Examples

use wavekat_core::AudioFrame;

// f32 input — zero-copy via Cow::Borrowed
let samples = [0.1f32, -0.2, 0.3];
let frame = AudioFrame::new(&samples, 16000);
assert_eq!(frame.samples(), &[0.1, -0.2, 0.3]);

// i16 input — normalized to f32 [-1.0, 1.0]
let samples = [i16::MAX, 0, i16::MIN];
let frame = AudioFrame::new(&samples, 16000);
assert!((frame.samples()[0] - 1.0).abs() < 0.001);

Implementations

impl<'a> AudioFrame<'a>

pub fn new(samples: impl IntoSamples<'a>, sample_rate: u32) -> Self

Create a new audio frame from any supported sample type.

Accepts &[f32] (zero-copy) or &[i16] (converts to normalized f32).

pub fn samples(&self) -> &[f32]

The audio samples as f32 normalized to [-1.0, 1.0].

pub fn sample_rate(&self) -> u32

Sample rate in Hz (e.g. 16000).

pub fn len(&self) -> usize

Number of samples in the frame.

pub fn is_empty(&self) -> bool

Returns true if the frame contains no samples.

pub fn duration_secs(&self) -> f64

Duration of this frame in seconds.

pub fn into_owned(self) -> AudioFrame<'static>

Consume the frame and return the owned samples.

impl AudioFrame<'static>

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

Construct an owned frame directly from a Vec<f32>.

Zero-copy — wraps the vec as Cow::Owned without cloning. Intended for audio producers (TTS, ASR) that generate owned data.

Example

use wavekat_core::AudioFrame;

let samples = vec![0.5f32, -0.5, 0.3];
let frame = AudioFrame::from_vec(samples, 24000);
assert_eq!(frame.sample_rate(), 24000);
assert_eq!(frame.len(), 3);

impl AudioFrame<'_>

pub fn resample( &self, target_rate: u32, ) -> Result<AudioFrame<'static>, CoreError>

Resample this frame to a different sample rate.

Returns a new owned AudioFrame at target_rate. If the frame is already at the target rate, returns a clone without touching the resampler.

Uses high-quality sinc interpolation via rubato.

Errors

Returns [CoreError::Audio] if the resampler cannot be constructed (e.g. zero sample rate) or if processing fails.

Example

use wavekat_core::AudioFrame;

let frame = AudioFrame::from_vec(vec![0.0f32; 4410], 44100);
let resampled = frame.resample(16000).unwrap();
assert_eq!(resampled.sample_rate(), 16000);

impl AudioFrame<'_>

pub fn write_wav(&self, path: impl AsRef<Path>) -> Result<(), CoreError>

Write this frame to a WAV file at path.

Always writes mono f32 PCM at the frame’s native sample rate.

Example

use wavekat_core::AudioFrame;

let frame = AudioFrame::from_vec(vec![0.0f32; 16000], 16000);
frame.write_wav("output.wav").unwrap();

impl AudioFrame<'static>

pub fn from_wav(path: impl AsRef<Path>) -> Result<Self, CoreError>

Read a mono WAV file and return an owned AudioFrame.

Accepts both f32 and i16 WAV files. i16 samples are normalised to [-1.0, 1.0] (divided by 32768).

Example

use wavekat_core::AudioFrame;

let frame = AudioFrame::from_wav("input.wav").unwrap();
println!("{} Hz, {} samples", frame.sample_rate(), frame.len());

Trait Implementations

impl<'a> Clone for AudioFrame<'a>

fn clone(&self) -> AudioFrame<'a>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<'a> Debug for AudioFrame<'a>

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

Formats the value using the given formatter.