search

princeton-ddss / SpeechMLPipeline

SpeechMLPipeline is a complete pipeline to deploy Machine Learning Models to generate labelled and timestamped transcripts from audio inputs
MIT License
0 stars 1 forks source link

FFmpeg returns the empty procssed file after removing the backgroud noise #11

Closed fjying closed 1 year ago

fjying commented 1 year ago

The issue did not occur under test data, but occur under some Citadel Data.

When the processed file is opened, it shows: the error occurs. The file contains no playable stream.

The file is completely empty: 0 bytes.

Sample File: ch01_20170525135534.wav

fjying commented 1 year ago

There are multiple versions of ffmpeg isntalled on Linux: version 6.0 and older versions. The old version 3.4.8 does not support the remove noise filter, which leads to the empty file output. Running Python codes to run ffmpeg shell scripts automatically uses the old ffmpeg version.

fjying commented 1 year ago

Solution:

(1) Ask Ryan to remove the older versoins of ffmpeg (2) Run ffmpeg shell scripts (.sh file) directly on the terminal instead of Python to use the latest verson of ffmpeg

fjying commented 1 year ago

Wait for Ryan to install encoder after removing the older versions

fjying commented 1 year ago

Issue solved after installing the latest version