- SDK downloads
- Sample codes
- Quick starts
- Common features
- Advanced streaming
- Advanced video processing
- Advanced audio processing
- Best practices
- Error codes
- API Documents
- API Documents
- Function overview
Low-Latency Live Streaming (L3) is a stable and reliable live streaming product that can provide developers with multi-terminal strong synchronization, tens of millions of concurrency, and millisecond-level latency. It can achieve live streaming of large classes, e-commerce Live, show live, watch and other scenes together.
1 Products Comparison
|Category||Real-time Audio and Video||Low-Latency Live Streaming||CDN Live Streaming|
|Typical scenarios||Frequent audio and video interactions between viewers and anchors are required, such as small classes, online meetings and other scenarios.||Strong synchronization is required between viewers, and the anchor needs to respond to the audience’s text, barrage, or reward information immediately, such as live streaming of large classes, watching movies together and other scenes.||The audience does not need to interact with the anchor audio, and there is no strong requirement for the synchronization of the live content between the audience, such as game live streaming and live show show.|
|Delay situation||The content delay between the viewer and the anchor linking the microphone is less than 400 ms.||The content delay range between the viewer and the host is between 600 ms and 1000 ms.||The content delay between the viewer and the host is more than 3000 ms.|
|Synchronization||Strong synchronization between audiences.||The synchronization between the audience is good.||Poor synchronization between viewers.|
2 Key Features
Main Function Description Low-latency streaming High-concurrency, stable and reliable, multi-terminal synchronization low-latency live streaming distribution service, supporting up to tens of millions of concurrent. Live streaming with microphones Supports multiple anchors linked with microphones, multi-person real-time communication and other functions, which can realize up to 50 channels of linked microphones with live streaming. And users can connect to the microphone and pull streaming with low latency at the same time. Basic beauty Supports basic beauty functions, including setting effects such as whitening and dermabrasion.
Main Function Description Screen Sharing The host can simultaneously display the screen content to other users in the channel, support designated sharing a certain screen or window, and support designated sharing areas. Voice Changer/Sound Effects SDK has built-in multiple voice changers and sound effects, which can increase the fun of live streaming and enhance the live streaming experience. Media Player A feature-rich media player that can play local or online media resources during the live streaming and push them to the live streaming room. Echo cancellation The engine has its own echo cancellation function. In the case of dual talk, each party has a clear voice, a good duplex experience, and supports three echo cancellation modes: comfortable, moderate and aggressive. Noise suppression The engine has its own noise suppression function. Combined with the psychoacoustic model, it can increase the signal-to-noise ratio by more than 20 dB without compromising the voice quality. Automatic gain A feature that automatically adjusts the receiver's audio gain (volume). After this feature is enabled, the sound pick-up and volume can be implemented more dynamically and make the volume more stable. Multiple encodings Support multiple hard coding and multiple soft coding to meet the coding requirements of multiple application scenarios and network environments. Multiple decoding Support multiple hard solutions and multiple soft solutions to meet the decoding requirements of multiple application scenarios and network environments. GPU processing All data goes to GPU memory, with good performance, good stability, and low latency.
3 Applicable Scenarios
- Live large classes: ultra-low latency online classroom experience, perfectly adapted to diverse teaching interactions, strong synchronization between teachers and students, and students, teaching immersion is comparable to offline classrooms.
- Show live streaming: The host quickly responds to the real-time barrage, and the audience immediately rewards with a millisecond delay to create a new show experience and detonate the live streaming.
- E-commerce live streaming: Assist new e-commerce gameplay such as spike and issuance, create richer e-commerce interactive scenes, and help improve the platform's UV conversion rate and GMV.
- Watch together: Massive audiences from all over the world gather in the live room to watch movies, concerts, competitions and other content simultaneously. The low-latency experience helps to share the joy with zero distance.
- Online auction: low-latency and high-smooth bidding experience, strong synchronous bidding process, to ensure fairness and openness in every link from quotation to hammer transaction.
4 Feature Indexes
|SDK package size||
Express-Video SDK (standard version) installation package increment size is as follows:
Express-Video SDK (including whiteboard function) installation package incremental size as follows:
|Live streaming delay||600 ms ～ 1000 ms.|
|Video quality||Up to 4K resolution, 1 fps ~ 60 fps.|
|Massive Concurrency||More than 200 BGP nodes worldwide, supporting tens of millions of concurrent.|
Low-latency live streaming supports platforms such as iOS, Android, Windows, macOS, Web, and WeChat applets. See the table below for specific compatibility requirements.
|Platform||Support Version||Support Architecture|
|iOS||7.0 or above||
|Android||4.1 or above||
|Windows||Windows 7 or above||
|macOS||10.10 or above||x86_64|
|WeChat Mini Program||-||-|