meowsuite

This is the repository for a family of quick scripts and aides for manipulating audio files and populating their metadata.

They were created to aid me on my quest to preserve content from, but not limited to, the Vietnamese diaspora and Sinosphere.

Suite of applications and their descriptions

• meowsplit splits a .wav file into tracks based on the intervals of silence it can detect.
• meowtag applies the MP3 ID tags that you populate from the CSV created by meowsplit
• meowconcat concatenates any n-number of wav files together, in case it is needed
• meowlabel queries OpenAI to transcribe any album art to get a track list into CSV format for you. This is handy in case information is not available on Discogs or other sources.
• meowmvgen collects all the mp3's in a working directory and makes a video out of them

A use case on how you could use these applications

This is my workflow:

1. Backup my media which gets put into .aea or .wav format, depending on my method of backing up. (I also scan the album art.) ffmpeg -i "my_audio.aea" "my_audio.wav"
2. Use meowsplit to process and create segments from the media output. meowsplit "my_audio.wav" & 'resources\Convert-MP3FFMPEG.ps1';
3. Identify the content on the segments, either manually or with computer vision, then use Visual Studio Code to populate the .csv that gets created with the relevant metadata. meowlabel path/to/album_art_scan.jpg
4. Use meowtag to populate the relevant .mp3 files with the metadata content, including album art if applicable. meowtag "my_audio_INFO.csv"

The end result is a reduced workflow tedium, leaving only the method of backing up the only factor dependent on equipment and audio playback, and only identifying the content on the segments limited by how fast I can query ChatGPT to transcribe scanned album art for me.