# The Ship — stem player

Full 6-stem mixer for **The Ship** (10 tracks). Same Web-Audio engine as the other stem players: every lane decoded to a buffer on one AudioContext clock (sample-locked, no drift), per-song select, transport, scrub, solo / mute / volume per stem.

## Tracks (10 of 11) — all separated
Every song below is separated into **6 stems** (drums, bass, guitar, keys, vocals, other) via demucs `htdemucs_6s`:

1. Seedy Shade — demo 5 · violin · 72bpm
2. Flake of the Year — demo 5 · w/ intro
3. Gene
4. Six Feet Under
5. Sleep Vs Death
7. MOP — Mix 02-06-26
8. JGBFTL — demo 4
9. MHL — demo 5 · w/ piano
10. Trapped in Wonderland — demo 10
11. Ambulance (encore)

**Missing:** #6 *Good Morning Captain* — the dropped file was a `.textClipping` placeholder with no audio. Add the mix and re-run the builder and it slots in as a 6-stem song too.

## Hosting + build
- Audio host: `drwu-stems.vercel.app/theship-stems/stems/<song>/{drums,bass,guitar,piano,vocals,other}.mp3` (60 lanes, all 200)
- Separation: `demucs -n htdemucs_6s --mp3 -d cpu` (output staged on /Volumes/Black, ~8 min/song)
- Build: `python3 ~/Projects/drwu-htmls/scripts/build_theship_stems.py` (lane-aware; auto-detects which songs have stems on disk)
