In the era of rapid development of Internet , Video applications are everywhere , The demand for video technology from all walks of life began to explode . With the improvement of video resolution , The rapid update of audio and video playing methods and the increasing demand of users for viewing experience , More and more application scenarios put forward higher requirements for audio and video technology .
8 month 24 Japan ,LiveVideoStack union Tencent cloud Produce “ One stop experience Tencent cloud Audio and video and integrated communication technology ” project , The invitation 4 position Tencent cloud Technical experts shared Tencent cloud In high quality 、 Technology evolution of high availability real time audio and video , reveal Tencent cloud Fast iteration behind the whole chain of video products 、 Massive concurrent background architecture and user experience Optimization of self-developed encoder . One stop experience Tencent video cloud is based on long-term massive business service experience and more than ten years of audio and video technology accumulation , Provide streaming from live 、 Video capture 、 Transcoding from editing to storage 、 content distribution 、 Decoding and playing full stack technology solution .
From line quality 、 Packet loss recovery , To experience score —— Tencent cloud The evolution of audio and video quality optimization
In the past year , Tencent cloud To include education 、 Finance 、 Many industries such as online entertainment provide high quality 、 Highly available video call and live link function , How to ensure ultra-low latency video interaction ？ How to ensure the viewing experience of users under the condition of weak network ？ Tencent video cloud terminal R & D general manager Chang Qing passed Tencent cloud TRTC The evolution of service technology gives us a detailed analysis ： From the self-developed player engine to optimize the multi terminal viewing experience , To introduce QUIC The protocol reduces the viewing delay of traditional live link 、 Greatly reduce push stream stuck , Then to the protocol stack optimization strategy from single point optimization to collaborative optimization to ensure the viewing experience under weak network conditions , And through the audio and video encoder fault tolerance to improve 、 Cloud decision system gradually improves audio and video quality optimization .
Evergreen General manager of Tencent video cloud terminal research and development
Tencent cloud PCDN： Dedicated to redefining the Internet of things service framework for fog Computing
review XP2P Development history of , From the rise of live broadcast services in recent years to save bandwidth costs as a starting point , With the development and perfection of ,XP2P Already at the bottom with the help of STUN、 Port prediction 、 Birthday attack 、UPnP The foundation of direct connection has been improved , And directly connected to UDP Connection based on efficient 、 Reliable transport protocol XNTP, So that we can be more strategic , On the implementation of the widely used HTTP agreement , So far, the service framework of the Internet of things has been basically completed , The network load is small 、 High transmission performance 、 Stable features . and XP2P/PCDN Naturally, it became the first application under this service framework , And excellent performance .
Tencent cloud Zhang Peng, a technical expert, thinks that , Once we continue to implement decentralized storage services , Add the name of the center to the discovery service ？ The prototype of a truly decentralized Internet service framework for all things will be perfected , It is a step closer to the optimal architecture that the early Internet workers dream of . Zhang Peng thinks that , Fog calculation is not only as simple as traffic sharing , It also needs a wind to really redefine it , Just like in those days AWS Redefining cloud computing is so great . And the service framework of the Internet of things , Will follow the simple idea that everything is connected and services will continue to develop , Maybe it's the key to redefining fog Computing , That's more exciting ！
Zhang Peng Tencent cloud Senior engineer
Open source collaborative video cloud , Polish video products with an open and cooperative attitude
With the rapid growth of video popularity and viewing demand , Tencent cloud technician 、FFmpeg Zhao Jun, a member of the decision-making committee, believes that , There is one in the video industry “ technology 、 Demand and reality ” The triangle game of , The scene is like a dance in chains , That is, it needs to be in ultra-high definition 、 The balance between computing power and network bandwidth . It is based on such a triangular game , Tencent cloud With “ Open source 、 synergy ” As a weapon , Gradually polish out a complete and efficient video product chain ： One side , Build audio and video codec based on 、 High speed HD solutions with enhanced image quality —— Bright eyes , as well as “AI+ video ” Intelligent Media Retrieval of 、 analysis 、 Audit program —— Wisdom eyes, etc ; On the other hand , Also actively embrace change and open source , The technology that will be used to build these products , Give back to the open source community with an open mind .
Zhao Jun Tencent cloud technician 、FFmpeg Members of the decision-making committee
V265 Encoder ： Give users complete 265 Experience
In the game between the demand of ultra high definition video image quality and the shackles of network bandwidth , Video coding is undoubtedly the focus of all companies , Tencent self-developed video encoder V265 You get it 2018 year MSU Video encoder competition subjective quality of the best results . As V265 One of the technical leaders , Tencent cloud Expert engineer Zhang Xianguo from the code rate control optimization strategy and technical practice landing optimization point of view to do a detailed analysis .V265 Through adaptive rate allocation in different scenes, the video compression rate is significantly improved 、 adopt CTU/ That's ok / The three-level rate control model of the frame realizes the accurate rate control . In practical application scenarios ,V265 Combined with high-speed HD business , Video content analysis is used to predict transcoding rate or quality parameters to meet users' subjective quality requirements , At the same time, it can save bit rate adaptively .
Zhang Xianguo Tencent cloud Expert engineer
Tencent video cloud
Tencent video cloud , Based on Tencent's long-term experience in massive business services , And the technology accumulation in the field of audio and video communication , Continue to build China's leading video cloud service platform , Provide one-stop video cloud solutions and powerful video communication product matrix , Including cloud live 、 Cloud on demand 、 Cloud transcoding 、 live broadcast SDK、 Short video SDK、 Beauty movement effect SDK、 Real time audio and video SDK Products such as , And super HD 、AI Video analysis 、 Digital rights management 、VR Video and other solutions , Widely used in online video 、 Pan Entertainment 、 Live telecast 、 E-commerce new retail 、 Online education 、 Financial and other scenarios .80% Live broadcast the choice of leading enterprises' trust .
Click on 【 Read the original 】 Get the instructor PPT（PPT The content will be updated after confirmation with the instructor , password ：ubfp）
![[QT] subclass qthread to realize multithreading](/img/95/085ef9ecf3478a8c6bad4f75716dd8.jpg)
