SoundFile

class montreal_forced_aligner.db.SoundFile(**kwargs)

Bases: Base

Database class for storing information about sound files

Parameters:

• file_id (int) – Foreign key to File

• file (File) – Root file

• sound_file_path (Path) – Path to the audio file

• format (str) – Format of the audio file (flac, wav, mp3, etc)

• sample_rate (int) – Sample rate of the audio file

• duration (float) – Duration of audio file

• num_channels (int) – Number of channels in the audio file

• sox_string (str) – String that Kaldi will use to process the sound file

load_audio(begin=0, end=None)

Load a normalized waveform for acoustic processing/visualization

Parameters:

• begin (float, optional) – Starting time point to return, defaults to 0

• end (float, optional) – Ending time point to return, defaults to the end of the file

Returns:

• numpy.array – Time points

• numpy.array – Sample values

normalized_waveform(begin=0, end=None)

Load a normalized waveform for acoustic processing/visualization

Parameters:

• begin (float, optional) – Starting time point to return, defaults to 0

• end (float, optional) – Ending time point to return, defaults to the end of the file

Returns:

• numpy.array – Time points

• numpy.array – Sample values