From the era when communication is king to the twenty-first century when the Internet is booming, audio and video technology has always been an essential skill for many technology companies. A mobile phone connecting the world, a social account to interact with friends from all over the world, is nothing new; even ultra-low latency, ultra-high-definition picture quality, immersive interactive metaverse and full truth interconnection, are rapidly fermenting, it seems to be about to be catalyzed by the underlying Moore's Law expires.
Fresh applications fade away easily, and classic techniques always shine. Today's audio and video technology can be described as blooming everywhere. Real-time audio and video communication network (TRTC), instant messaging network (IM), streaming media distribution network (CDN), these technical terms you may not have heard, but you are exposed to online classes, online meetings, e-commerce live broadcasts, video social networking every day is inseparable from the support of these technologies.
However, the maturity and richness of audio and video technology have also brought trouble to many developers. More and more multi-terminal developers in the development of audio and video communication related applications, in order to achieve different product capabilities, often need to integrate multiple SDKs, not only increase the development of joint commissioning workload, but also increase the difficulty of work.
Closely related to this, the requirements of the application layer such as transport optimization, service adaptation, and content distribution directly affect the user experience. How to meet this kind of demand at a low threshold and low cost is also a major difficulty.
Ten years to sharpen a sword. After 21 years of accumulation and precipitation, Tencent Cloud Audio and Video has created a comprehensive audio and video product matrix, and now standing at the door of the era of full interconnection, in the face of many industry pain points and difficulties, Tencent Cloud Audio and Video will integrate the three networks into one, face the difficulties, and set off a wave of audio and video.
The future has come, the three characteristics of the era of full interconnection
In December 2020, Ma Huateng, chairman of the board of directors of Tencent, talked about the future trends and changes of the Internet in the Tencent annual magazine, and predicted the arrival of the true Internet era.
Ma Huateng mentioned that the Internet has come to this day, and a series of basic technologies such as real-time communication to audio-visual video have matured; the rapid improvement of computing power has promoted richer changes in the mode of information contact and human-computer interaction. This is a process from quantitative change to qualitative change, which means the integration of online and offline, and the full integration of physical and electronic methods. The doors to the virtual world and the real world have opened, whether from virtual to real, or from real to virtual, are committed to helping users achieve more realistic experiences.
The advent of the era of full interconnection actually has a premonition. As Tencent Cloud Audio and Video Expert Engineer Chang Qing mentioned in an interview, looking at the entire Internet industry, we will see three major changes in the application level in recent years.
The first is to pursue lower and lower transmission delay. Customers are becoming more and more demanding on latency. For example, in the online chorus solution, the sound transmission delay of two users has been compressed to less than 70ms, close to the network transmission delay between the two points. This can be considered very demanding compared to the previous RTC transmission delay requirements.
Second, the use of the combination of virtual and real is becoming more and more common. For example, Tencent Conference supports virtual background technology, so that its images can be perfectly integrated with beautiful background pictures. At the same time, with the popularity of web assembly technology, related reasoning libraries and models have also begun to land on browsers, which makes WebRTC web page users also enjoy this function.
Third, VR technology may enter a stable period of development out of the "vase" stage. With the maturity of technologies such as Wi-Fi6 wireless streaming methods, and the increasing power of built-in chips, the comfort and playability of VR devices have improved significantly. At the same time, the relevant ecology is becoming more and more mature, and it is likely to enter a healthy development stage of stable user growth in the future.
From online chorus, to video conferencing, to the maturity of VR technology, it is not difficult to find that behind it there is audio and video technology as a key support. In other words, the advent of the era of full interconnection is inseparable from excellent audio and video technology. Tencent cloud audio and video has long been prepared.
Three networks in one, open up the "second pulse of Ren Dou" of enterprise audio and video construction
Since the QQ era, Tencent has been committed to the research and development and application of audio and video technology. Today, Tencent has long been a technical leader in the field of audio and video. As the primary position of infrastructure aggregation, Tencent Cloud Audio and Video has already built three world-class networks based on the "cloud, edge and end" infrastructure - real-time audio and video (TRTC) network, instant messaging (IM) network and streaming media distribution (CDN) network.
Nowadays, in order to solve industry difficulties such as SDK integration and reduce the threshold for enterprises to build audio and video applications, Tencent Cloud Audio and Video has launched a "three-in-one" RT-ONE™ audio and video communication basic network, integrating the above three networks to build a network communication base for the industry's audio and video communication PaaS platform.
RT-ONE™ multiplexes three networks to form a technical superposition advantage, which can meet the requirements of unified scheduling, unified interface mode, nearby access, transmission optimization and service adaptation, and greatly reduce the access threshold of audio and video products. Compared to previously isolated networks, the RT-ONE™ has two core advantages:
First, the integration of technology is more thorough. After the completion of the fusion, the traditional CDN network has absorbed many of the technologies of RTC and launched a special product such as "fast live broadcast" that takes into account the characteristics of RTC low latency and high CDN concurrency. Users can enjoy lower latency and stuttering rates, and can achieve high concurrent viewing of millions of people. In addition, the RTC network reuses many of the highly concurrent components of the CDN network, absorbing the distributed design concept, resulting in higher concurrency capabilities and greater stability.
Second, product interoperability is more natural. In the past, RTMP live broadcasting using Tencent Cloud Audio and Video needed to open a live broadcast service, and RTC Lianmai needed to open a TRTC service, and the two sets of services depended on each other. Customers who have already used the live broadcast service must re-dock the program to use TRTC Lianmai, while customers who use TRTC for online education must turn on the live recording service to complete the video recording. Based on the "RT-ONE™" network, customers can use the V2 interface of mobile live broadcasting to achieve seamless switching between TRTC and live broadcast services, and TRTC's recording services can also be upgraded to achieve a better user experience and more flexible customization capabilities.
Take RT-ONE™ empowering a live stream with goods as an example. In this scenario, the customer needs the ability to use both TRTC, IM, and CDN.
First of all, the anchor introduces the product to the audience, gives the audience a voucher, the audience conducts a second kill through the commodity link, snaps up, sends a bullet screen to interact with the anchor, etc., these behaviors require the use of IM's messaging ability;
Sometimes the anchor also has to interact with the audience, which requires the use of real-time audio and video communication capabilities provided by TRTC.
At the same time, the audience for a live broadcast with goods is very large, and the number of viewers may even exceed one million at the peak, which requires the streaming ability of the CDN.
The above mentioned is only the most basic ability in the e-commerce live broadcast scenario. In addition, there are PK with goods, like gifts, grab red packets, lottery, etc. are also commonly used capabilities, but also need to rely on TRTC, IM, CDN three networks to achieve together.
It can be seen that RT-ONE™ is not only to meet the more and more three-network parallel needs of customers, compared with the traditional independent TRTC + IM + CDN solution, the deep integration of RT-ONE™ network through the reuse of three networks of "cloud, edge, end" infrastructure, in terms of access threshold, network quality, cost of use are more competitive, and in terms of performance, RT-ONE™ > TRTC + IM + CDN.
If the "technology" and "product" of the enterprise are regarded as the "Ren Dou Er Pulse" of the enterprise, then the arrival of RT-ONE™ can be described as a comprehensive martial arts secret book, designed for enterprises to open up and build audio and video applications of the "Ren Dou Er Pulse". However, only practicing internal skills is not enough to resist the enemy - in addition to RT-ONE™, Tencent Cloud Audio and Video released the "Tencent Cloud Vision Cube" at the Global Digital Ecology Conference on November 3, integrating terminal capabilities for enterprises and repairing both inside and outside.
Terminal capability integration reduces the threshold for developers to use
At a time when live video broadcasting is hot, audio and video applications from all walks of life are emerging in an endless stream. However, in order to achieve different product capabilities - such as beauty and beauty, anchor audience even mai, bullet screen comments, etc., developers often need to integrate multiple SDKs, at the same time, different functional modules are cluttered and inconsistent, API call rules are complex, which greatly increases the developer's development and joint debugging workload.
Tencent Cloud Cube Audio and Video Terminal Engine (RT-Cube™) aims to integrate terminal capabilities, lower the threshold for developers, and realize the development experience of calling all terminal capabilities with a single SDK.
From the user's point of view, RT-Cube™ is an integrated product of audio and video terminals. Its functional modules are rich and flexible, including the six most popular and practical functions at present, namely: live broadcast, anchor audience even mai / cross-room PK, video recording editing / release upload, live viewing, on-demand viewing, and audio and video calls.
Based on the six major functions, Tencent Cloud Vision Cube (RT-Cube™) provides users with multiple preset SDK versions and many value-added capabilities. Among them, the preset SDK version includes mobile live streaming SDK, short video SDK, audio and video call SDK, player SDK, and full-featured SDK; at the same time, customers can also make advanced configurations according to their own needs, customize the combination of function modules, and create their own SDK. Truly one access, call everywhere, without losing flexibility.
Value-added capabilities are designed to create high-quality audiovisual experiences for our customers. Includes beauty effect SDK, genuine music library SDK and data quality monitoring and many other optimization functions. Truly professional, stable, high-quality terminal services.
Take the pan-entertainment field as an example. Tencent Cloud Vision Cube combines its own capabilities to create low-code solutions for many pan-entertainment customers. Through the full-code open source of the client and the server, the rich classic gameplay "out of the box for skin" can be launched. In the real-time chorus scene, Tencent Cloud Vision Cube brought an ultra-low latency experience upgrade, and in response to the copyright issues that enterprises are generally concerned about, it opened the AME live interactive song library through train, providing more than 200,000 genuine songs.
In the field of enterprise collaboration, Tencent Cloud Vision Cube has launched audio and video call components, multi-person audio and video collaboration components and 1v1 online customer service components, which can be integrated into enterprise applications and easily have the same quality audio and video communication capabilities as WeChat/QQ calls, Tencent meetings, and enterprise customer service.
It's not just pan-entertainment and enterprise collaboration. Combined with the "cloud + terminal" basic capabilities of Tencent Cloud RT-ONE™ audio and video fusion network and Tencent Cloud Cube audio and™ video terminal engine, Tencent Cloud Audio and Video has been applied in social entertainment, live streaming with goods, construction real estate, cultural tourism and other industries, riding the dust in the audio and video market.
Dynamic encoding technology has been upgraded, and bright eyes bring a new high-definition field of vision
User experience is everything. As the leading product company in China, Tencent is well versed in the way of product success, and it also adheres to this law in the field of audio and video.
Based on intelligent dynamic coding technology, Tencent Cloud Audio and Video has created a high-speed HD solution for bright eyes. Through the combination of intelligent scene recognition, dynamic encoding matching, and image quality repair enhancement, it realizes the ability to provide higher definition services at a lower bit rate, aiming to bring users a new high-definition horizon.
Bright Eye Ultra HD solution mainly has four advantages: HD low code, ultra-clear field of view, picture quality repair and high customization:
- HD low code: Based on intelligent scene recognition, dynamic coding technology, CTU/line/frame three-level bitrate precision control model, Mingyan Extreme Speed HD solution can provide higher definition streaming media services with lower bit rate (average saving of 50% +) for live broadcast, on-demand and other industries.
- Ultra-clear vision: Based on ultra-high-performance encoding algorithms, Mingyan Extreme Speed HD solution supports high-resolution video real-time encoding up to 8K, integrating ultra-resolution, HDR, widening color gamut and other technologies, which can generate ultra-high-quality video for users and provide the ultimate clarity of view.
- Picture quality repair: Through 3D drying, color enhancement, super resolution, interpolation and other processing technologies, the Bright Eye Extreme Hd Solution can effectively eliminate noise, mosaic, jitter, frustration and other problems in old films, improve the clarity of the picture, and regenerate the picture quality.
- Highly customizable: Mingyan Ultra-Fast HD solution has a simple, mature and stable, efficient and flexible API interface, which is equivalent to a customized video intelligent service privatization middle platform, which can build a safe, stable, efficient and reliable video cloud service ecosystem for users.
Live broadcasting, broadcast media, online video and short video are the main scenarios of Tencent Mingyi's ultra-fast HD solution. Taking Tencent's internal business as an example, after using Mingyan Extreme Speed HD, the overall business saves about 70% of storage and bandwidth costs, and due to the reduction of files, the time consuming of the first frame of the video is also reduced by 20%, and the overall playback fluency is greatly improved. For the extreme compression of on-demand scenes, 1080P HD movie video, ultra-fast HD H.264 can maintain the overall subjective clarity at a bitrate of 1.5M, or keep vmAF above 95 points. H.265 can achieve the same effect at 900kbps, and AV1 even achieves 650kbps.
For classroom education scenes, the compression effect is more obvious, and there are more classroom scenes than still pictures, and more coding tools can be used. For PPT classroom scenarios, ultra-fast HD H.264 can be under 67kbps, while maintaining subjective clarity, H.265 can do 35kbps, AV1 can do 28kps, at this time, most of the bitrate of the video has been lower than the bitrate of audio, greatly reducing the storage and bandwidth of video.
At present, Tencent Mingyan's ultra-fast HD solution has reached cooperation with CCTV, Kuaishou, Douyu and other industry leaders to bring new HD video services to these enterprises.
Audio and video opened the curtain of the times, and Tencent Cloud Audio and Video took the lead
The advent of the era of full-truth interconnection is indispensable to the blessing of audio and video technology, and the audio and video field is leading the way with Tencent cloud audio and video.
On January 7 this year, IDC (International Data Corporation) released the "China Video Cloud Market Tracking (First Half of 2021)" report. The report mentioned that in the first half of 2021, the size of China's video cloud market reached 4.37 billion US dollars, an increase of 38.7% year-on-year. Among them, the growth rate of the audio and video solution market reached 47.6%, and Tencent Cloud ranked first in the industry.
And this industry first, in fact, has a long history.
Since 2018, IDC has continued to market in China's video cloud market, and Tencent's cloud solution market share has ranked first, and has achieved "four consecutive championships" so far. Behind Tencent Cloud Audio and Video's continued leadership, IDC mentioned that "Tencent Cloud Audio and Video released the Stereo Cube (RT-Cube™) audio and video terminal engine during the reporting period, integrating all terminal capabilities such as live broadcasting, TRTC, IM, etc., and supporting the RT-ONE™ audio and video communication network, reusing the "cloud, edge, and end" infrastructure of the three networks, with the intention of providing out-of-the-box audio and video application development tools for the entire industry."
It is worth mentioning that IDC data also shows that in the key direction of RTC (real-time communication), the growth rate of Tencent Cloud Audio and Video ranks first among the head manufacturers.
TRTC (Tencent Cloud Real-time Audio and Video) has served more than 5,000 customers in all walks of life in addition to the excellent practice of national applications such as QQ, Tencent Conference, National K Song, and Glory of the King. Its background architecture technology has won the "China Patent Gold Award", the highest award of China's intellectual property rights, and its audio and video codec, audio and video processing and other technical fields maintain a global leading position in technology, and can link Tencent Cloud's advantages of direct on-demand capabilities to provide real-time interaction, direct on-demand, video processing and other audio and video full-link technology solutions, one-stop to meet the needs of customers' audio and video.
With the continuous evolution of domestic audio and video technology and the maturity of related industries, Tencent Cloud Audio and Video, as the industry leader of audio and video, will be deeply rooted in the field of technology, and will continue to explore in the future, and strive to escort countless cutting-edge developers and various cutting-edge products that are rising. Before the advent of the ultra-high-definition video era of 5G+8K, Tencent Cloud Audio and Video is also looking forward to keeping pace with the development of the entire industry and ushering in a new pattern.