Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix MP3 resampling when a dataset's audio files have different sampling rates #3665

Merged
merged 4 commits into from
Feb 2, 2022
Merged

Fix MP3 resampling when a dataset's audio files have different sampling rates #3665

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
d9e7618
fix mp3 resampling
lhoestq Feb 2, 2022
e8ebbaa
add test
lhoestq Feb 2, 2022
13a3cee
style
lhoestq Feb 2, 2022
b357de7
fix test
lhoestq Feb 2, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion src/datasets/features/audio.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -174,7 +174,7 @@ def _decode_mp3(self, path_or_file):

array, sampling_rate = torchaudio.load(path_or_file, format="mp3")
if self.sampling_rate and self.sampling_rate != sampling_rate:
if not hasattr(self, "_resampler"):
if not hasattr(self, "_resampler") or self._resampler.orig_freq != sampling_rate:
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nice workaround, this should minimize the number of required _resampler replacements 👍

self._resampler = T.Resample(sampling_rate, self.sampling_rate)
array = self._resampler(array)
sampling_rate = self.sampling_rate
Expand Down
Binary file added tests/features/data/test_audio_16000.mp3
View file
Open in desktop
Binary file not shown.
20 changes: 20 additions & 0 deletions tests/features/test_audio.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -133,6 +133,26 @@ def test_audio_decode_example_mp3(shared_datadir):
assert decoded_example["sampling_rate"] == 44100


@require_sox
@require_torchaudio
def test_audio_resampling_mp3_different_sampling_rates(shared_datadir):
audio_path = str(shared_datadir / "test_audio_44100.mp3")
audio_path2 = str(shared_datadir / "test_audio_16000.mp3")
audio = Audio(sampling_rate=48000)

decoded_example = audio.decode_example(audio.encode_example(audio_path))
assert decoded_example.keys() == {"path", "array", "sampling_rate"}
assert decoded_example["path"] == audio_path
assert decoded_example["array"].shape == (119119,)
assert decoded_example["sampling_rate"] == 48000

decoded_example = audio.decode_example(audio.encode_example(audio_path2))
assert decoded_example.keys() == {"path", "array", "sampling_rate"}
assert decoded_example["path"] == audio_path2
assert decoded_example["array"].shape == (120960,)
assert decoded_example["sampling_rate"] == 48000


@require_sndfile
def test_dataset_with_audio_feature(shared_datadir):
audio_path = str(shared_datadir / "test_audio_44100.wav")
Expand Down