BTawaifi / Audio-DeSilencer

An audio processing tool for detecting and removing silence in audio recordings. Create text files for video silence removal using custom-defined thresholds.

Audio DeSilencer

Audio DeSilencer is a powerful Python package for audio processing that empowers you to detect and remove silence in audio recordings. Additionally, it provides functionality to facilitate video silence removal using the generated text files.

Features

• Silence Detection: Identify and segment silent parts of audio based on custom-defined silence thresholds and minimum silence duration.

• Silence Removal: Effortlessly remove detected silence segments to enhance audio quality and eliminate unnecessary gaps.

• Text File Generation: Create text files containing the timeline data of silent and non-silent parts, facilitating further processing for video silence removal or in-depth analysis.

• Command-Line Interface: Seamlessly integrate Audio DeSilencer into your audio and video processing pipelines with an intuitive command-line interface.

Installation

Installation via pip

To install Audio DeSilencer using pip, simply run:

pip install audio-desilencer

Now, you can run Audio DeSilencer from the command line:

audio-desilencer input_audio.mp3 --output_folder output_directory --min_silence_len 100 --threshold -30

Manual Installation

If you prefer not to use pip, you can manually install Audio DeSilencer by following these steps:

1. Clone the Audio DeSilencer repository:

   git clone https://github.com/BTawaifi/Audio-DeSilencer.git

2. Navigate to the project directory:

   cd Audio-DeSilencer

3. Install the required dependencies:

   pip install -r requirements.txt

4. Now, you can run Audio DeSilencer from the command line:

   python audio-desilencer.py input_audio.mp3 --output_folder output_directory --min_silence_len 100 --threshold -30

Command-Line Arguments

• input_audio.mp3: The path to the input audio file.
• --output_folder: (Optional) The folder where the output files and timeline data will be saved (default is "output").
• --min_silence_len: (Optional) The minimum duration of silence (in milliseconds) to be considered as a separate silent part (default is 100 ms).
• --threshold: (Optional) The silence threshold in dBFS (decibels relative to full scale) used to distinguish between silent and non-silent parts (default is -30 dBFS).

Output

Audio DeSilencer will generate the following output:

• interview_silent.mp3: Audio file with silence segments removed.
• interview_non_silent.mp3: Audio file containing only the non-silent parts.
• silent_parts.txt: Timeline data of silent parts.
• non_silent_parts.txt: Timeline data of non-silent parts.

Usage Example

Here's a usage example to get you started:

python audio-desilencer.py input_audio.mp3 --output_folder output_directory --min_silence_len 100 --threshold -30

Or if you have it installed using pip

audio-desilencer input_audio.mp3 --output_folder output_directory --min_silence_len 100 --threshold -30

Contributing

Contributions to Audio DeSilencer are highly encouraged! Whether you'd like to report issues or submit pull requests to enhance the tool, your contributions are invaluable. Please refer to the CONTRIBUTING guide for details on how to contribute.

License

This project is licensed under the MIT License - see the LICENSE file for details.