Lyricoder

Lyricoder is a script that generates a JSON file of song lyrics with timestamps.

🚀 Concept

Lyricoder is a script that generates a JSON file of song lyrics with timestamps. The script takes in a .mp3 file and the name of the song and outputs a JSON file with the lyrics and timestamps. The script uses OpenAI Whisper to transcribe the song.

graph TB
    A([Start]) --> B[Read Input Parameters]
    B --> C[Open MP3 File]
    C --> D(Call OpenAI Whisper API)
    D --> E[Receive and Parse Response]
    E --> F[Is response SRT Format?]
    F -- Yes --> G[Convert SRT to JSON]
    F -- No --> H[Print Error - Invalid response format]
    G --> I[Export JSON to File]
    I --> J[Print Success Message]
    J --> K((End))

This script is useful for a music player app that could display the lyrics of the song as it is playing.

🌈 Installation

Clone the repository to your local machine and navigate into the directory:

git clone https://github.com/360macky/lyricoder.git
cd lyricoder

You will need to make the script executable and move it to a directory in your PATH. You can do this by running the following commands:

chmod +x lyricoder.py
sudo mv lyricoder.py /usr/local/bin/lyricoder

The chmod command makes the file executable and the mv command moves the file. /usr/local/bin is a common place to put user programs on macOS, and should be in your PATH.

Now, you should be able to run the script from anywhere using the lyricoder command:

lyricoder --input ./song.mp3 --name "Song Name" --output ./song.json

🪐 Usage

python lyricoder.py --input ./song.mp3 --name "Song Name" --output ./song.json

🛠 Core Development

The script takes in three command-line arguments: --input (the path to the .mp3 file), --name (the name of the song), and --output (the path to the output .json file).

The song is transcribed using OpenAI Whisper through a POST request to the Whisper API. The ASR system returns the transcribed text in SRT (SubRip Subtitle) format along with timestamps. The response is then parsed and converted into a JSON format.

The final output is a JSON file which includes each line of lyrics along with its start and end times. The JSON is structured as follows:

{
  "Song Name": [
    {
      "id": 1,
      "start_time": "00:00:03",
      "end_time": "00:00:04",
      "text": "First line of lyrics"
    },
    {
      "id": 2,
      "start_time": "00:00:05",
      "end_time": "00:00:09",
      "text": "Second line of lyrics"
    },
    ...
  ]
}

This script includes error handling for failed transcription attempts and unexpected errors.

🤲 Contributing

Lyricoder is an open source project.

If you want to be the author of a new feature, fix a bug or contribute with something new.

Fork the repository and make changes as you like. Pull requests are warmly welcome.

📃 License

Distributed under the MIT License. See LICENSE for more information.

360macky / lyricoder