repository·develop·Indexed 12 days ago
https://github.com/ossrs/srsA high-efficiency, real-time video server supporting multiple protocols including RTMP, WebRTC, HLS, and SRT. Designed for low-latency video streaming across various platforms and architectures, SRS supports integration with tools like FFmpeg and OBS.
SRS uses several different Dockerfiles depending on the intended use case:
Dockerfile: Used for release and general testing.trunk/Dockerfile.pkg: Used for creating package binaries.trunk/Dockerfile.builds: Used for testing build processes.trunk/Dockerfile.test: Used for running tests.trunk/Dockerfile.cov: Used for testing and code coverage.gMock is a C++ framework used for creating mock classes to facilitate better system design and testing. It provides a declarative syntax for defining mocks and allows for the creation of partial (hybrid) mocks that combine real and mock object behavior.
Key capabilities include:
Note: gMock is part of the GoogleTest C++ testing framework and is subject to its requirements.
SRS is an open-source, high-efficiency, real-time video server. It operates on a typical publish (push) and subscribe (play) model, working with clients like FFmpeg, OBS, and VLC to receive and distribute streams.
SRS can convert between various protocols, such as converting RTMP or SRT to HLS, HTTP-FLV, or WebRTC.
Oryx (formerly SRS Stack) is an all-in-one, out-of-the-box, open-source video solution designed for creating live streaming and WebRTC services. It integrates several core technologies into a single-node application to provide a complete media workflow.
Core Integrated Components:
Oryx is an open-source, out-of-the-box video cloud solution (SRS Stack) built on Go, Reactjs, SRS, FFmpeg, and WebRTC. It is designed for various streaming scenarios including:
SRS (Simple Realtime Server) is a high-efficiency, real-time media server designed to receive streams from publishers and deliver them to players.
Workflow:
SRS supports various protocols for both input and output, allowing for protocol conversion (transmuxing) to suit different delivery needs.
graph TD
subgraph Publishers
P["FFmpeg, OBS, Larix, vMix,<br/>hardware encoders, browsers, apps"]
end
S["SRS"]
subgraph Players
L["FFmpeg, VLC, ffplay, ExoPlayer,<br/>IJKPlayer, browsers, hardware, apps"]
end
Publishers --> S --> PlayersSRS (Simple Realtime Server) follows a release cycle identified by specific codenames. For a detailed list of features available in each version, refer to the SRS Features documentation.
/* Release History Summary */
* Release 8.0 (2026~Now): Code: Free
* Release 7.0 (2025~2026): Code: Kai
* Release 6.0 (2023~2025): Code: Hang
* Release 5.0 (2022~2023): Code: Bee
* Release 4.0 (2020~2021): Code: Leo
* Release 3.0 (2018~2019): Code: OuXuli
* Release 2.0 (2015~2017): Code: ZhouGuowen
* Release 1.0 (2013~2014): Code: HuKaiqunSRS 5.0 introduced significant improvements to core streaming protocols and architecture:
libsrt.SRS relies on several third-party libraries to enable specific streaming and protocol features. Depending on your required functionality (e.g., WebRTC, HLS, or Transcoding), you may need to ensure these dependencies are available in your build environment.
http-parser.nginx.srt.openssl (specifically for delivering H.264+AAC streams).openssl (version 1.0.*) and libsrtp.ffmpeg and opus.gperftools.SRS provides native support for several streaming protocols, each with specific use cases and limitations:
libsrt.SRS supports a wide range of client-side technologies for both publishing (encoding/streaming) and playing back live streams. The general workflow is:
Encoder $\rightarrow$ SRS/CDN Network $\rightarrow$ Player.
Supported technologies include:
SRS 6.0 (codename: Hang) provides comprehensive HEVC (H.265) support across multiple delivery protocols and recording formats:
libsrt v1.5.3, configurable default_streamid, and reduced SRT-to-RTC latency (down to 200ms).