Why is gapless so hard?
summary
Explains that padding and codec lag force encoders to add zeros or LPC tails, that padding must be trimmed and timestamps adjusted, and DRM/AC3 pass‑through often prevents trimming; reports Opus’s 48 kHz constraint vs 44.1 kHz libraries, libopus LPC trick reduces leakage, and FFmpeg behaves inconsistently; describes a remote clock‑drift measurer built to quantify sync errors (1080p about 10× worse than ideal, 4K error doubles), notes ~5 ms per short frame drift and that YouTube testing found artifacts traced to resampling/FFmpeg interactions.
AI-written summary · about
chapters (10)
- 0:00Gapless audio challenges
- 2:29Audio compression basics
- 6:04Transform codecs limits
- 10:16Why gapless is hard
- 13:22Padding and spectral leakage
- 15:27Opus and resampling issues
- 17:33Community improvements
- 18:58Gapless for video
- 20:43Audio-video sync
- 22:07Interstitials and stitching
related talks
31mIn a Perfect World: Why Media Sources Are a Hassle
12mWhy is your encoder so slow? The curse of next-gen standards
33mThe Challenges of Testing Video Playback at Scale
26mBandwidth Ballet: Optimizing Video Performance for Web Maestros
27mThe Importance of Audio Live Panel
04mSub-frame Latency
22mProviding better video experiences for the next billion users.
30mA paradigm shift in codec standards: MPEG-5 Part 2 LCEVC, higher quality with any...