How to build a high quality 'live' HD VP9 Encoder
summary
Claims a working FPGA-based VP9 encoder that delivers live savings close to reported VOD gains (YouTube ~50% offline, Netflix ~36%) by cutting bitrate roughly 30% versus H.264 (example 3 Mbps x264 vs 2 Mbps VP9) and running 1080p60 real-time at 2 Mbps vs 3 Mbps H.264. Uses exhaustive rate-distortion optimized mode decision, look-ahead rate control, segmentation and FPGA DSP/SRAM scaling; hardware prototype and demos show objective metrics (VMAF/PSNR/SSIM) near HEVC and a deployable PCIe/driver stack.
AI-written summary · about
chapters (8)
- 0:00Start
- 15:34ANATOMY OF A HARDWARE ENCODER
- 18:30COMPUTATIONALLY INTENSIVE BLOCKS
- 19:10EXAMPLE: 32 x 32 DCT TRANSFORM
- 21:38NGCODEC ENCODER FPGA STRUCTURE
- 23:10ENCODER RESOURCE UTILIZATION
- 23:59DEVELOPMENT CYCLE
- 27:07RECAP
related talks
10mAdventures with VP9
26mNew video coding techniques under consideration for VP10 - the successor to VP9
26mDASHing Into the Wild with VP9
30mA paradigm shift in codec standards: MPEG-5 Part 2 LCEVC, higher quality with any...
17mContent Aware Encoding for low latency live streaming encoders using deep learning
31mAV1 Update
27mHighly-Efficient SVT-AV1-based Solutions for VOD Applications
17mDelivering better video encodes for legacy devices