What is the reason of disable_ffmpeg_torchaudio_info in common voice recipe

[trunglebka]

I'm preparing common voice dataset and it show the below warning:

2023-11-01 09:09:52 (prepare.sh:97:main) Stage 1: Prepare CommonVoice manifest
Processing CommonVoice languages:   0%|                                                                                                                                                                                 | 0/1 [00:00<?, ?it/s]
2023-11-01 09:09:55,807 INFO [commonvoice.py:286] Language: en
                                                                                                                                                                                                                                             2
023-11-01 09:09:55,807 INFO [commonvoice.py:296] Spliting train                                                                                                                                                         | 0/3 [00:00<?, ?it/s]
2023-11-01 09:09:55,807 INFO [backend.py:44] The user overrided the global setting for whether to use ffmpeg-torchaudio to compute the duration of audio files. Old setting: True. New setting: False.
                                          2023-11-01 09:09:56,849 INFO [commonvoice.py:155] No such file: /workspace/github/icefall/egs/commonvoice/ASR/download/cv-corpus-15.0-2023-09-08/en/clips/path
                                              /opt/anaconda3/envs/k2/lib/python3.11/site-packages/torchaudio/_backend/soundfile_backend.py:71: UserWarning: The MPEG_LAYER_III subtype is unknown to TorchAudio. As a result, the
bits_per_sample attribute will be set to 0. If you are seeing this warning, please report by opening an issue on github (after checking for existing/closed ones). You may otherwise ignore this warning.
  warnings.warn(ks: 606it [00:00, 2437.13it/s]
/opt/anaconda3/envs/k2/lib/python3.11/site-packages/torchaudio/_backend/soundfile_backend.py:71: UserWarning: The MPEG_LAYER_III subtype is unknown to TorchAudio. As a result, the bits_per_sample attribute will be set to 0. If
 you are seeing this warning, please report by opening an issue on github (after checking for existing/closed ones). You may otherwise ignore this warning.
...
<repetitions of `...The MPEG_LAYER_III subtype is unknown to TorchAudio`>

And after searching the recipe, I can't get arround why ffmpeg backend being disabled here and how can I prepare it correctly?

Here is the commit: 3ee82c1

[desh2608]

This discussion may be relevant: #1025

[trunglebka]

Thank you @desh2608.
Regarding the problem: I was able to run the recipe without that warning by install libsndfile and ffmpeg in conda env (not the OS package)