Changelog 1.1.0

🔊 3D Spatial Audio & Speaker/Projection Enhancements

• Native 3D Positional Audio for In-Game Speakers:
  • Migrated speaker audio playback from flat StaticAudioChannel with manual gain curves to LocationalAudioChannel.
  • Speakers now feature full 3D spatialization (OpenAL / HRTF, stereo panning, pitch, and distance roll-off) centered directly on the speaker's in-game block location.
• Native 3D Positional Audio for Voice Projections (Body Anchors):
  • Upgraded voice projections (routeOwnerVoiceToAnchorListeners) to LocationalAudioChannel positioned at the projection's anchor location.
  • Players standing around an active projection/anchor (e.g. security cameras, drones, intercoms) now hear the projected voice with accurate 3D spatialization and attenuation.
• Elimination of Multi-Channel Audio Crackling & Popping:
  • Resolved Opus filter corruption occurring when multiple speakers or projections transmit concurrently.
  • Replaced shared/global OpusEncoder instances with discrete, stateful stream encoders keyed per (Source : Sender : Listener) pipeline.
  • Deterministic channel UUIDs (UUID.nameUUIDFromBytes) ensure each unique audio stream retains its own independent Opus prediction state without inter-channel bleed.
  • Added automatic encoder lifecycle cleanup on player disconnect and inactivity timeouts (30 seconds).

⚙️ Configuration & Live Reload Synchronization

• Fixed Live Config Reload (/dreamvoice reload config):
  • Fixed an issue where reloading the configuration reloaded Codex in memory but failed to notify or update the active VoiceWallService.
  • /dreamvoice reload config now dynamically applies changes to VoiceWall active state (enabled), modes (STRICT_BLOCK, REALISTIC, etc.) and airDamping live without requiring a server reboot.
  • Updated Codex.VoiceWall.getEffectiveMode() so disabling VoiceWall ("enabled": false) properly forces VoiceWallMode.OFF even if an explicit mode string is present in the configuration.
• Modernized Default config.json Template:
  • Aligned default config.json with the full modern Codex object graph (including diffraction, categories, overrides, reverb, and airDamping).
  • Strict JSON handling in CodexServiceImpl: eliminates stale fallback behaviors and ensures consistency between file configuration and in-game diagnostic commands (/voicewall info).

🧱 VoiceWall & Acoustic Physics

• Fixed Wall Attenuation Gain Inversion & Clipping Elimination:
  • Fixed a critical DSP bug where wall attenuation formula Math.pow(10.0, totalDbLoss / 20.0) amplified audio signals (e.g. +15 dB to +40 dB boost, up to 100x magnification) instead of attenuating them (-totalDbLoss / 20.0).
  • Added sample clamping Math.clamp(..., Short.MIN_VALUE, Short.MAX_VALUE) and soft-knee anti-clipping limiting (AudioLimiter) to prevent digital saturation crackling/popping when voice passes through walls.
  • Full 3D Spatialization Intact: Retained native Simple Voice Chat EntitySoundPacket transmission, preserving client-side OpenAL HRTF, angle, distance, and 3D spatial panning.
• Dedicated Stream Codecs for VoiceWall Audio:
  • Replaced shared receiver codecs with isolated OpusDecoder and OpusEncoder pipelines keyed per (senderUuid : receiverUUID) stream.
  • Prevents Opus Linear Predictive Coding (LPC) state desynchronization when multiple speakers talk simultaneously or within wall attenuation fields.
  • Added automatic cleanup of inactive stream codecs upon player quit and inactivity timeouts (30s).
• Acoustic Raycast Partial Attenuation:
  • Updated speaker and projection wall occlusion checks from binary block checks (ray.isBlocked()) to line-of-sight analysis (!ray.lineOfSight()).
  • Allows material-based sound dampening and wall absorption through blocks instead of completely cutting off audio prematurely.
• Codex Data Mutability:
  • Removed final from Codex.distance and added @Setter / @AllArgsConstructor annotations to support runtime updates and proper deserialization.

🎙️ Global Audio Codec Isolation & Crackling Elimination

• Wiretaps & Hidden Bugs (VoiceWiretapService):
  • Isolated Opus encoders per (wiretapUuid : senderUuid) stream to eliminate multi-speaker corruption around listening bugs.
  • Deterministic channel UUIDs (UUID.nameUUIDFromBytes) ensure channel stability across player reconnections.
  • Automatic stream encoder disposal on player quit and inactivity.
• Walkie-Talkie Radio (VoiceRadioService):
  • Isolated radio Opus encoders per sender to prevent frame predictor corruption between radio broadcasts and other audio features.
  • Fixed Roger Beep audio injection: temporary tone generator encoder is cleanly closed without colliding with ongoing player voice streams.
  • Deterministic channel UUIDs for static radio channels.
• Point-to-Point Transmitters (VoiceTransmitterService):
  • Dedicated transmitter encoders per speaker and deterministic static channel IDs.
  • Complete lifecycle tracking with automatic inactivity cleanup.
• Voice Cassette Recording (VoiceRecordingService):
  • Dedicated recording encoders per speaker protected by AudioLimiter.
  • Guaranteed safe native Opus decoder closure in finally blocks during cassette playback frame stitching.