Audio#

Defined in: voxatlas.audio.audio

class voxatlas.audio.audio.Audio(waveform, sample_rate, path=None, channel=None)[source]#

Bases: object

Store waveform data for one VoxAtlas stream.

Parameters:

  • waveform (ndarray) – One-dimensional waveform array.

  • sample_rate (int) – Sampling rate in Hertz.

  • path (str | None) – Optional source path for the waveform.

  • channel (int | None) – Optional channel index when the waveform came from a multichannel recording.

Returns:

Dataclass describing one audio stream.

Return type:

Audio

Notes

The pipeline treats Audio as the canonical acoustic input object for feature extractors.

Examples

>>> import numpy as np
>>> from voxatlas.audio.audio import Audio
>>> audio = Audio(waveform=np.zeros(16000, dtype=np.float32), sample_rate=16000)
>>> audio.duration
1.0
waveform: ndarray#
sample_rate: int#
path: str | None = None#
channel: int | None = None#
property duration: float#

Return the duration of the waveform in seconds.

Returns:

Audio duration in seconds.

Return type:

float

Notes

Duration is computed directly from waveform length and sample rate.

Examples

>>> import numpy as np
>>> from voxatlas.audio.audio import Audio
>>> audio = Audio(waveform=np.zeros(8000, dtype=np.float32), sample_rate=16000)
>>> audio.duration
0.5