LYRICS TO XML

This project provides tools to generate subtitles from audio files using OpenAI's Whisper model, convert subtitle files from SRT format to a custom BaseXML format, and run the entire process sequentially.

Prerequisites

• Python 3.7 or higher
• Install required Python packages:

  pip install -r requirements.txt

• The Whisper model will be downloaded automatically when running the scripts.

Folder Structure

• generate_subtitles.py: Script to transcribe audio and generate subtitles in SRT or XML-like format.
• srt_to_basexml.py: Script to convert SRT subtitle files to BaseXML format used by the project.
• run_generate_and_convert.py: Script to run the subtitle generation and conversion sequentially.
• subtitle/: Folder where generated subtitle files (SRT/XML) are saved.
• Preset/: Folder where BaseXML files are saved.
• base.xml: Example or base XML file.
• music/: Folder for audio files (not included in this repo).
• subtitle/: Folder for subtitle files.

Usage

1. Generate Subtitles from Audio

Run the generate_subtitles.py script with the path to your audio file:

python generate_subtitles.py path/to/audio.mp3 [--format srt|xml] [--offset seconds] [--model tiny|base|small|medium|large]

• --format: Output subtitle format, either srt (default) or xml.
• --offset: Optional time offset in seconds to adjust subtitle timestamps.
• --model: Whisper model size to use (default is base).

Example:

python generate_subtitles.py music/song.mp3 --format srt --model small

The generated subtitle file will be saved in the subtitle/ folder.

2. Convert SRT to BaseXML

Convert an existing SRT subtitle file to BaseXML format using:

python srt_to_basexml.py path/to/subtitle.srt [output.xml]

• If output.xml is not provided, the output will be saved as Preset/{subtitle_basename}.xml.

Example:

python srt_to_basexml.py subtitle/song.srt

3. Run Subtitle Generation and Conversion Sequentially

Use the run_generate_and_convert.py script to generate subtitles from audio and convert them to BaseXML in one step:

python run_generate_and_convert.py path/to/audio.mp3 [--model tiny|base|small|medium|large] [--offset seconds]

Example:

python run_generate_and_convert.py music/song.mp3 --model base --offset 0.5

Notes

• Ensure your audio files are placed in the music/ folder or provide the correct path.
• Generated subtitles are saved in the subtitle/ folder.
• Converted BaseXML files are saved in the Preset/ folder.
• The Whisper model will be downloaded automatically on first run.

License

This project is provided as-is without warranty.