Computing multiple spectrograms with the Core API

Contents

Computing multiple spectrograms with the Core API [1]#

Create an OSEkit AudioDataset from the files on disk, by directly specifying the time-related requirements in the constructor.

An Instrument can be provided to the AudioDataset for the WAV data to be converted in pressure units. This will lead the resulting spectra to be expressed in dB SPL (rather than in dB FS):

We will only use the folder in which the files are located: we don’t have to dig up to the file level.

from pathlib import Path

audio_folder = Path(r"_static/sample_audio")

from osekit.core_api.audio_dataset import AudioDataset
from osekit.core_api.instrument import Instrument
from pandas import Timestamp, Timedelta

audio_dataset = AudioDataset.from_folder(
    folder=audio_folder,
    strptime_format="%y%m%d_%H%M%S",
    begin=Timestamp("2022-09-25 22:35:15"),
    end=Timestamp("2022-09-25 22:36:25"),
    data_duration=Timedelta(seconds=5),
    instrument=Instrument(end_to_end_db=150.0),
)

The AudioDataset object contains all the to-be-exported AudioData:

print(f"{' AUDIO DATASET ':#^60}")
print(f"{'Begin:':<30}{str(audio_dataset.begin):>30}")
print(f"{'End:':<30}{str(audio_dataset.end):>30}")
print(f"{'Sample rate:':<30}{str(audio_dataset.sample_rate):>30}")
print(f"{'Nb of audio data:':<30}{str(len(audio_dataset.data)):>30}")

###################### AUDIO DATASET #######################
Begin:                                   2022-09-25 22:35:15
End:                                     2022-09-25 22:36:25
Sample rate:                                           48000
Nb of audio data:                                         14

We also wanted to skip the AudioData that are in the gap between recordings. Such AudioData have no linked file, thus their is_empty property should be True.

print(f"{' BEFORE FILTERING ':#^60}")
print(
    f"{'Nb of Empty data:':<30}{str(len([ad for ad in audio_dataset.data if ad.is_empty])):>30}\n"
)

# Remove the empty data by using the default AudioDataset constructor:
audio_dataset = AudioDataset([ad for ad in audio_dataset.data if not ad.is_empty])

##################### BEFORE FILTERING #####################
Nb of Empty data:                                          4

The AudioData should now only contain non-empty AudioData:

print(f"{' AFTER FILTERING ':#^60}")
print(f"{'Nb of audio data:':<30}{str(len(audio_dataset.data)):>30}")
print(
    f"{'Nb of Empty data:':<30}{str(len([ad for ad in audio_dataset.data if ad.is_empty])):>30}\n"
)

##################### AFTER FILTERING ######################
Nb of audio data:                                         10
Nb of Empty data:                                          0

Instantiate a scipy.signal.ShortTimeFFT FFT object with the required parameters:

from scipy.signal import ShortTimeFFT
from scipy.signal.windows import hamming

sft = ShortTimeFFT(
    win=hamming(1024),
    hop=128,
    fs=audio_dataset.sample_rate,
)

Create an OSEkit SpectroDataset from the AudioDataset and the ShortTimeFFT objects:

from osekit.core_api.spectro_dataset import SpectroDataset

spectro_dataset = SpectroDataset.from_audio_dataset(
    audio_dataset=audio_dataset,
    fft=sft,
    v_lim=(0.0, 150.0),  # Boundaries of the spectrograms
    colormap="viridis",  # Default value
)

We can plot a SpectroData object if we want to glance at the output before computing all spectrograms:

import matplotlib.pyplot as plt

spectro_dataset.data[0].plot()
plt.show()
_images/00007f5869527a51c1d40793bc04f3d6d3ead93c0eb55da7c003d580f624c560.png

We are now ready to export the spectrograms and matrices:

# Export all spectrograms
spectro_dataset.save_spectrogram(folder=audio_folder / "spectrograms")

# Export all NPZ matrices
spectro_dataset.write(folder=audio_folder / "matrices")

PSD estimates#

We can also export Power Spectral Density estimates using the welch method:

spectro_dataset.write_welch(folder=audio_folder / "welch")
Contents