When it comes to audio and video interactive call systems, there are currently two more on the market, one is the WebRTC-based video interactive system, and the other is the one-to-one interaction of RTMP/RTSP. Since the research and development team of TSINGSEE Qingxi Video is doing research on this recently, we will also discuss and compare these two interactive systems with you.
WebRTC
Many interactive systems are based on webrtc, and its biggest advantage is convenience. Now that WebRTC technology is built into the browser, users can realize real-time communication through the browser without using any plug-ins or software. For developers, before Google open-sourced WebRTC, the technology to achieve communication between browsers was in the hands of large companies. The development of this technology is a very difficult task. Now developers use simple HTML tags and JavaScript. API can realize the function of Web audio/video communication.
But in terms of video streaming transmission, webrtc also has its inevitable flaws. Its server deployment is very complicated, which is not conducive to private deployment. It cannot be used in some high-privacy scenarios, such as public security and municipal systems; the transmission is based on UDP, which is very It is difficult to guarantee the transmission quality. Because UDP is an unreliable transmission protocol, in a complex public network environment, various bursts of traffic, occasional transmission errors, network jitter, timeouts, etc. will cause abnormal packet loss, which will be to a certain extent Affect the quality of audio and video communication; and the entire framework system is not flexible enough.
RTMP/RTSP
Based on the existing RTMP, RTSP push, or built-in RTSP service, playing mature development kits, the product has high stability, and the industry's first ultra-low latency feature; adding noise suppression, echo cancellation, automatic gain control and other features to ensure the call effect; The use of common RTMP and RTSP servers, such as nginx, SRS or Darwin Stream Server (native version), is more conducive to private deployment.
to sum up
Regardless of the above implementation logic, it specifically involves the following parts: the acquisition terminal obtains the original yuv/rgb data, and the h.264/h.265 encoding is packaged into the required format, such as rtmp or rtsp and the data is packaged and sent; At the streaming end, the data is obtained through the rtsp or rtmp protocol, and the audio and video data parser is performed, and then decoded, and the decoded original data can be drawn.
EasyRTC is the web real-time interactive call system in the cloud side-end architecture of TSINSEEE, which combines the advantages of webrtc and rtmp/rtsp solutions, supports one-to-one, one-to-many and other video conference interactive calls to meet voice social and online education And training, video conferencing and telemedicine scenes.
The video transmission of EasyRTC can be divided into three parts as a whole:
1. The user side pushes the data stream to EasyRTC by using terminal devices such as mobile phones and computers;
2. The EasyRTC video conference cloud service processes the reception by means of noise reduction, increase, and echo cancellation. The data stream is pushed to other terminal users;
3. The client (Windows, Android, iOS, H5) receives the data sent by EasyRTC to complete a whole video call.
If you have such needs, please contact us to understand, EasyRTC provides a trial version for everyone to test, welcome to consult.