Skip to content
This repository has been archived by the owner on Jan 23, 2023. It is now read-only.

Latest commit

 

History

History
49 lines (39 loc) · 2.34 KB

README.md

File metadata and controls

49 lines (39 loc) · 2.34 KB

Adaptive streaming is a popular and rather elegant technology to adapt the audio-video streams provided over the internet dynamically to the users' needs. During research at VRT-medialab, we developed some tools that were useful in creating, testing and debugging these Adaptive Bitrate streaming files. Since these tools might be useful for other people to use, we decided to open-source them.

HTTP adaptive streaming intro

Most HTTP adaptive streaming protocols have the same basic construction. The audio-visual material is encoded multiple times, each with a different bitrate or quality. The resulting streams are chopped up timewise into pieces of equal length (typically somewhere between 2 and 10 seconds each). Each piece of each quality has its own URL and hence is individually accessible.

With this information, the client can navigate its way across the different available qualities. Usually starting out on the lowest bitrate (to assure a fast start), and switching up to higher bitrates if the network, cpu, ... allows.

Contents

This project contains:

  • A segmenting tool. This tool splits the input stream into multiple output files, according to Apple's version of the adaptive streaming protocol (IETF draft http://tools.ietf.org/html/draft-pantos-http-live-streaming-06). The segmenter code compiles into 4 binaries, each using a different splitting algorithm: ByteCount, MP3, ADTS (AAC) and MpegtsH264. Apart from ByteCount, they all try to split every N seconds, but keeping the file structure in mind: i.e. ADTS will cut on frame boundaries, MpegtsH264 will cut on GOP boundaries. It also provides support for Live-stream-mode and supports encryption

  • A few parser scripts to dump binary formats into a "human" readable format. It's by no means an easy read, but has saved us many hours of watching hexdumps.

    • adts-parser.pl will spit out the ADTS-frames and tell you what type of AAC block is inside.

    • h264-parser.pl reads in raw (AnnexB) h.264 streams and figures out the NAL type (I-frame, non-I-frame, ...) of the packet

    • mp4-parser.pl reads in an MP4 container (MPEG 4 Part 14) and prints out the box structure

    • ts-tools.pl parses MPEG 2 Transport Streams. It can spit out the decoded text, but can also extract (demux) a single video or audio-stream out of the TS-file.