docs: Add Audio Lifecycle Refactor plan (Task #56)

Created comprehensive design document for refactoring the tracker-synth
relationship to eliminate initialization order dependencies.

Proposed solution: AudioEngine class that manages both synth and tracker
as private members, providing order-independent initialization and clear
ownership semantics.

Three options analyzed:
- Option A: Unified AudioEngine (recommended)
- Option B: Registration Handle System
- Option C: Reference-Counted Resources

Estimated effort: 2-3 weeks with incremental migration path.
Binary size impact: ~500 bytes (acceptable).

See doc/AUDIO_LIFECYCLE_REFACTOR.md for complete design rationale,
implementation plan, and decision matrix.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

1 files changed, 12 insertions, 0 deletions

diff --git a/TODO.md b/TODO.md
index 3322304..fbe4060 100644
--- a/TODO.md
+++ b/TODO.md
@@ -113,6 +113,18 @@ This file tracks prioritized tasks with detailed attack plans.
     - [x] **Enhanced Procedural Noise:** Implemented a multi-octave Value Noise generator for higher-quality skybox textures.
     - [x] **Scene Integrity:** Restored proper object indexing and removed redundant geometry, ensuring the floor grid and objects render correctly.
 
+## Priority 1.5: Audio System Architecture (Task #56)
+**Goal:** Refactor tracker-synth relationship to eliminate fragile initialization order dependency.
+
+- [ ] **Task #56: Audio Lifecycle Refactor**
+    - **Problem:** Current system requires synth_init() before tracker_init() or spectrograms get cleared
+    - **Impact:** Fragile, hard to test, prevents composition, causes subtle bugs
+    - **Proposed Solution:** Unified AudioEngine class that manages synth and tracker as members
+    - **See:** `doc/AUDIO_LIFECYCLE_REFACTOR.md` for detailed design options and implementation plan
+    - **Effort:** 2-3 weeks (incremental migration path available)
+    - **Size Impact:** ~500 bytes (acceptable for robustness gain)
+    - **Priority:** Medium (current workaround functional but fragile)
+
 ## Priority 2: 3D System Enhancements (Task #18)
 **Goal:** Establish a pipeline for importing complex 3D scenes to replace hardcoded geometry.
 - [ ] **Task #18.0: Basic OBJ Asset Pipeline** (New)