festifun

monster/festifun

Fork 0

Commit Graph

Author	SHA1	Message	Date
m3ultra	157c783e25	Lane A (M1): ingest + GCC-PHAT audio sync Fills the foundation stubs in ingest.py and audio_sync.py. - ingest: ffprobe probe + mono 16 kHz WAV extraction per video (ffmpeg); idempotent run_ingest reuses existing rows so synthetic -> ingest -> sync works without unique-constraint collisions. - sync: PHAT-whitened GCC-PHAT pairwise offsets with sub-sample parabolic peak refinement, scored by peak-to-second-peak ratio; confidence-weighted global least-squares solve with cycle-consistency rejection (>50 ms); windowed drift estimation (ppm) with a 5 ppm deadband; disconnected sync-graph components -> offset_ms=None (spec pitfall #5). - Persists offset_ms/drift_ppm/sync_confidence via db.update_video_sync and exports data/work/sync.json. Verified on the synthetic fixture: synthetic -> ingest -> sync recovers the ground-truth offsets (0 / +1370 / -842 ms) to <0.001 ms and drift 0 ppm, well inside the spec M1 tolerances (+/-10 ms, +/-3 ppm). pytest: 43 passed (24 foundation + 19 new in test_ingest.py and test_audio_sync.py). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	2026-07-16 01:13:56 +10:00
m3ultra	cf6a6fb2c8	Foundation (M0 + M3 + frozen contracts): scaffold, synthetic fixtures, full API What now works: - M0 scaffold: pyproject (all spec deps), uv/py3.12 env, `python -m festival4d` CLI registering synthetic\|ingest\|sync\|reconstruct\|events\|serve. Vite hello page. - Synthetic fixture (synthetic.py): 3 shifted-audio videos (offsets 0/+1370/-842 ms), camera-arc poses, stage point cloud -> points.ply, seeded events + anchors, ground_truth.json. `python -m festival4d synthetic` populates data/ + DB. - DB schema exactly per spec §2 (db.py) + CRUD helpers all lanes use. - M3 API (api.py) full against synthetic data: manifest/poses/pointcloud/anchors/ events/detect/annotations; Range-capable video serving (206 verified); CORS for any localhost origin. - Frozen geometry contract: geometry.colmap_to_threejs (M5 math) + unit test (3 known vectors, random round-trip, scipy oracle); mirrored frontend/src/lib/pose.js with identical POSE_TEST_VECTORS. Lane-B stubs: slerp_pose, ray_from_pixel, triangulate_rays, nearest_point_on_ray. - Classifier contract (events_ai.py): MomentClassification model + MomentClassifier protocol + Gemini/Claude/Local provider stubs. - Lane-owned modules stubbed with final signatures (ingest, audio_sync, frames, sfm, events_ai); cli/api catch NotImplementedError and degrade gracefully. - plan/CHANGE_REQUESTS.md created; plan/status/foundation.md updated. Acceptance: pytest 24 passed; serve endpoints verified via curl + browser (video seek, manifest fetch cross-origin, pose.js self-test, 0 console errors). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	2026-07-16 00:45:55 +10:00

Author

SHA1

Message

Date

m3ultra

157c783e25

Lane A (M1): ingest + GCC-PHAT audio sync

Fills the foundation stubs in ingest.py and audio_sync.py.

- ingest: ffprobe probe + mono 16 kHz WAV extraction per video (ffmpeg);
  idempotent run_ingest reuses existing rows so synthetic -> ingest -> sync
  works without unique-constraint collisions.
- sync: PHAT-whitened GCC-PHAT pairwise offsets with sub-sample parabolic
  peak refinement, scored by peak-to-second-peak ratio; confidence-weighted
  global least-squares solve with cycle-consistency rejection (>50 ms);
  windowed drift estimation (ppm) with a 5 ppm deadband; disconnected
  sync-graph components -> offset_ms=None (spec pitfall #5).
- Persists offset_ms/drift_ppm/sync_confidence via db.update_video_sync and
  exports data/work/sync.json.

Verified on the synthetic fixture: synthetic -> ingest -> sync recovers the
ground-truth offsets (0 / +1370 / -842 ms) to <0.001 ms and drift 0 ppm,
well inside the spec M1 tolerances (+/-10 ms, +/-3 ppm). pytest: 43 passed
(24 foundation + 19 new in test_ingest.py and test_audio_sync.py).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

2026-07-16 01:13:56 +10:00

m3ultra

cf6a6fb2c8

Foundation (M0 + M3 + frozen contracts): scaffold, synthetic fixtures, full API

What now works:
- M0 scaffold: pyproject (all spec deps), uv/py3.12 env, `python -m festival4d`
  CLI registering synthetic|ingest|sync|reconstruct|events|serve. Vite hello page.
- Synthetic fixture (synthetic.py): 3 shifted-audio videos (offsets 0/+1370/-842 ms),
  camera-arc poses, stage point cloud -> points.ply, seeded events + anchors,
  ground_truth.json. `python -m festival4d synthetic` populates data/ + DB.
- DB schema exactly per spec §2 (db.py) + CRUD helpers all lanes use.
- M3 API (api.py) full against synthetic data: manifest/poses/pointcloud/anchors/
  events/detect/annotations; Range-capable video serving (206 verified); CORS for any
  localhost origin.
- Frozen geometry contract: geometry.colmap_to_threejs (M5 math) + unit test (3 known
  vectors, random round-trip, scipy oracle); mirrored frontend/src/lib/pose.js with
  identical POSE_TEST_VECTORS. Lane-B stubs: slerp_pose, ray_from_pixel, triangulate_rays,
  nearest_point_on_ray.
- Classifier contract (events_ai.py): MomentClassification model + MomentClassifier
  protocol + Gemini/Claude/Local provider stubs.
- Lane-owned modules stubbed with final signatures (ingest, audio_sync, frames, sfm,
  events_ai); cli/api catch NotImplementedError and degrade gracefully.
- plan/CHANGE_REQUESTS.md created; plan/status/foundation.md updated.

Acceptance: pytest 24 passed; serve endpoints verified via curl + browser (video seek,
manifest fetch cross-origin, pose.js self-test, 0 console errors).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

2026-07-16 00:45:55 +10:00

2 Commits