Release / build (push) Successful in 5m13s
The sounds now come from the recordings in audio_assets/, cut and cleaned up by tools/make_sounds.py into one WAV per sound (shared composeResources/files/sounds), loaded by both apps instead of being synthesised at runtime: - your turn: the nudge ring; to call trump: the nudge twice - 10 seconds left: the reminder ring; someone joins: its first strikes - the last seconds: the clock's tick - a card played, the deal, a trick gathered: the card recording - won, lost: the happy and the sad melody - trump called, the hand over: still synthesised (no recording yet) Each is trimmed so it starts at once, faded out instead of cut, the card recording's rumble filtered out, and every sound set to the same loudness, its loudest 100 ms at -20 LUFS (K-weighted, BS.1770), about 6 dB below the old chimes, with peaks under -4 dBFS. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>