Why is your encoder so slow? The curse of next-gen standards
summary
Claim: next‑gen codecs slow not because of immature software but because of design choices like motion field merge modes that create non-differentiable enumerations. An optimized VVC encoder from Fraunhofer HHI demonstrates the tradeoffs: enabling implicit tools gives big wins in fast presets, while exhaustive searches over signaling tools deliver marginal bitrate gains — a 10× complexity jump from faster to medium yields ~40% improvement, another 10× to slower only ~12% — and the optimized encoder achieves those Pareto tradeoffs in practice.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:33Video coding history
- 2:06How does an encoder work
- 3:06HVC
- 4:32MERM
- 5:11VBC
- 6:59Coding complexity
- 7:57Optimization
- 8:33Presets
- 10:03Conclusion
related talks
11mVirtues of Recycling in Multi-rate Encoding
16mMAE: More Adaptive Video Encoder for Consistent Low Latency in High-Quality Real-Time...
11mVirtues of Recycling in Multi-rate Encoding
23mDecoder Complexity Aware AV1 Encoding Optimization
29mHow to build a high quality 'live' HD VP9 Encoder
30mA paradigm shift in codec standards: MPEG-5 Part 2 LCEVC, higher quality with any...
20mBoosting Video Codec With RISC-V Vector Extension
15mWhat Codec should I use?