SUNSETSEJONG UNIVERSITY
← Back to projects
P-07 · Supporting

IKUMPlay

SupportingWorking PrototypeAI Experiment
IKUMPlay terminal karaoke pipeline searching tracks through Spotipy, acquiring media with yt-dlp, separating vocals with Docker Spleeter fallback, parsing synced LRC lyrics, and playing back through pygame with Rich terminal visualization and keyboard controls.SIG-IKP · MEDIA PIPELINESearchSpotipyAcquireyt-dlpSeparateSpleeterLyricsLRC syncPlaybackpygameRich UIterminalRUNTIME SURFACEspectrogram · keyboard controls · cleanup
System diagram

Verified summary

Terminal-based karaoke application combining song search, media preparation, source separation, synchronized lyrics, and playback.

Problem or experiment question

Can a terminal workflow acquire songs, separate vocals, and play back synced lyrics with visual feedback?

What was actually built

  • Spotify search through Spotipy
  • yt-dlp acquisition path with Docker/Spleeter fallback
  • Synced lyrics/LRC parsing and pygame playback
  • Rich terminal interface with spectrogram/visualizer and keyboard controls

Important technical decisions

  • Terminal-first UI keeps focus on media pipeline transparency
  • Spleeter fallback supports vocal separation when primary path fails
  • External services chosen for search and acquisition speed during prototyping

Architecture or workflow

  • Search layer: Spotify metadata lookup
  • Acquisition layer: yt-dlp and separation fallback
  • Playback layer: pygame with LRC synchronization
  • UI layer: Rich terminal visualization

Known limitations

  • Hardcoded Spotify credentials were observed and must not be published
  • Committed virtual-environment artifacts inflate repository size
  • External media and service availability affect operation
  • Copyright and platform terms require careful public wording

Repositories

Related projects