Caution

You're reading an old version of this documentation. If you want up-to-date information, please have a look at 0.9.1.

librosa.time_to_frames¶

librosa.time_to_frames(times, sr=22050, hop_length=512, n_fft=None)[source]¶

Converts time stamps into STFT frames.

Parameters

timesnp.ndarray [shape=(n,)]: time (in seconds) or vector of time values
srnumber > 0 [scalar]: audio sampling rate
hop_lengthint > 0 [scalar]: number of samples between successive frames
n_fftNone or int > 0 [scalar]: Optional: length of the FFT window. If given, time conversion will include an offset of - n_fft // 2 to counteract windowing effects in STFT.

Note

This may result in negative frame indices.

Returns

framesnp.ndarray [shape=(n,), dtype=int]

Frame numbers corresponding to the given times:

frames[i] = floor( times[i] * sr / hop_length )

See also

frames_to_time: convert frame indices to time values
time_to_samples: convert time values to sample indices

Examples

Get the frame numbers for every 100ms

>>> librosa.time_to_frames(np.arange(0, 1, 0.1),
...                         sr=22050, hop_length=512)
array([ 0,  4,  8, 12, 17, 21, 25, 30, 34, 38])