Meet the Packets: How audio travels into your browser
summary
Traces a real pipeline: audio gets encoded and segmented into HLS or MPEG DASH, converted using FFmpeg or the uploader's Amazon elastic trans coder, placed in S3 and served as playlists plus fragmented .ts or .mp4 segments. Congestion control matters: BBR increases bandwidth 5–15% for some groups versus Reno which treats loss as congestion. A single-file HTML audio player wired to dash.JS or HLS playback handles manifests and it works in browsers (HLS is required on iOS Safari).
AI-written summary · about
chapters (12)
- 0:00Start
- 5:39Adaptive Bitrate Streaming Protocols
- 6:16HLS & MPEG DASH
- 7:22You've got options
- 7:55Amazon Elastic Transcoder
- 9:04HLS Audio Stream
- 10:59MPEG-DASH Audio Stream
- 14:01Traffic Congestion
- 14:24TCP: A Round Trip
- 16:37Congestion Window (sending rate)
- 17:25TCP Reno
- 20:45Client Players
related talks
58mModern & secure adaptive streaming on the Web
45mModern & secure adaptive streaming on the Web
46mModern & secure adaptive streaming on the Web
43mMedia Source Extensions
27mPerformance of Low-Latency DASH/CMAF and Low-Latency HLS Systems
37mThree Roads to Jerusalem
43mModern Streaming Video for the Jamstack
24mVideo streaming on the web: Live and recorded video streaming