Transcrybe

Transcrybe is a small command-line tool that transcribes MP3 audio files into text using Google Speech Recognition. It splits long audio into smaller chunks, processes each chunk, and writes the result to transcription.txt.

Features

• Transcribes MP3 files to text.
• Splits long audio into smaller chunks automatically.
• Supports a configurable language code.
• Cleans up temporary chunk files after processing.

Requirements

• Python 3.7 or higher
• FFmpeg
• Python packages listed in requirements.txt

Installation

1. Install FFmpeg.

For macOS:

brew install ffmpeg

For Ubuntu or Debian:

sudo apt-get install ffmpeg

For Windows:

Install FFmpeg and make sure it is available on your PATH.

2. Install the Python dependencies:

pip install -r requirements.txt

Usage

Run the script from the command line:

python transcriber.py <mp3_file_path> [--language <language_code>]

Arguments

• <mp3_file_path>: Path to the MP3 file to be transcribed.
• --language: Language code for transcription. Defaults to en-US.

Example

python transcriber.py my_audio.mp3 --language en-US

Output

The transcribed text is saved to transcription.txt in the current directory.

Notes

• Transcription quality depends on the clarity of the audio.
• If a chunk cannot be understood, the script writes [Unintelligible].
• The script uses Google Speech Recognition through the SpeechRecognition Python package.

License

This project is licensed under the MIT License. See LICENSE for details.