Tencent Cloud Audio and Video Re-Growth

Author | Yang Li

Produced by | Leifeng.com Industry Group

Leifeng.com: Although the public switched telephone network has existed since Bell invented the telephone, voice transmission using the Internet has only begun to rise 20 years ago. Today, short videos, live broadcasts, video conferencing and other application fields have become a red ocean, but the new war has not yet ended at the pace imagined. Audio and video communications are increasingly moving towards cloud-based integration, which has also prompted major cloud service providers to secretly mobilize their reserve forces in an attempt to prepare for the arrival of a new round of dividends.

Speaking of privilege, Tencent must be one of the companies in the long history of China's Internet that has accumulated 20 years of audio and video communication technology.

Li Yutao, who has been in charge of Tencent QQ communications and underlying audio and video technologies for many years, has taken over a series of team businesses including TRTC real-time audio and video, IM instant messaging and CDN streaming media distribution. Since last year, he has been fully committed to integrating the three basic networks into the RT-ONE™ network.

These signals have been there for a long time. In 2020, the emergence of popular Internet products such as Tencent Conference and WeChat Video Account has made relevant teams experience multiple collaborations. Although these businesses are relatively independent at present, there is room for integration both in terms of underlying technology scheduling and scenario application.

Recently, Leifeng.com had an in-depth visit with the Tencent Audio and Video team and found that using Tencent Cloud Audio and Video capabilities as a fulcrum, integrating and scheduling resources and providing them to end customers in the form of packaged services (API and SDK) is the inevitable result of Tencent Audio and Video's self-research, self-use, and cloud migration.

Li Yutao, vice president of Tencent Cloud, has been a representative of QQ communication services, from supporting QQ's massive business to opening up the platform. Liao Long, currently the product director of Tencent Cloud CDN (streaming video distribution), has served as the backend R&D of QQ Space, QQ Farm and other businesses. Looking at the other business line leaders who communicated with us this time, they all have more than ten years of business experience in the fields of communications and audio and video. Leifeng.com also learned that after the establishment of the Tencent Cloud audio and video brand, the team involved has already had three or four hundred people.

Recently, Tencent Cloud officially launched the Tencent Audio and Video brand and introduced a number of solutions including interactive cloud games, immersive content creation domains, real-time communications and business collaboration. The reason that enabled Tencent Cloud to change this situation may be traced back to 20 years ago.

Tencent is an Internet company that started out as a communications technology company. Its accumulation of technology and advantages in market share are undoubtedly obvious to the industry.

On February 10, 1999, the first version of QICQ 99a was officially released. From the beginning, QQ had audio and video call functions, but for a long time afterwards, Tencent relied on Tencent Research Institute and third-party technical service providers, and there was no independent team responsible for audio and video technology.

This is a story from the beginning of Tencent Audio and Video.

However, by 2011, with the development of QQ's massive business and Tencent's own business, as well as the black box problem caused by the use of third-party technology, Tencent began to have the idea of ​​self-development. In the same year, Google acquired GIPS and open-sourced it through the WebRTC project, Skype was acquired by Microsoft, and Zoom was founded. Many leading players in the audio and video communication field in the industry began to get agitated.

It is worth mentioning that many products such as QQ, Skype, and Webex used the GIPS engine at the time. In the same year, Tencent also began to establish an audio and video center and gathered people to develop the first generation engine TRAE (Tencent Realtime Audio Engine). In the same year, the TRAE engine officially replaced GIPS and became the core engine of QQ audio and video services.

During this stage, Tencent completed the initial accumulation of QQ in audio and video business and technology.

Since 2015, QQ has started to open up its business, providing IM and real-time audio and video communications to third parties in the form of APIs.

At the same time, Tencent's self-developed audio and video engine began to accelerate. In 2015, Tencent opened its self-developed SPEAR audio and video engine; in 2016, Tencent opened its OpenSDK engine; in 2017, Tencent developed the third-generation engine XCast. For example, today's Tencent Conference is developed based on XCast. The audio and video center of that year also gradually grew and developed into the current Tencent Multimedia Laboratory.

During this period, Tencent's audio and video technology was opened up in the form of PaaS cloud services. The original service scenarios mainly used text and pictures have now been expanded to scenarios with video and streaming as the main usage forms; the original game live broadcasts and short videos that were biased towards consumer scenarios have now been expanded to online education, video conferencing, e-commerce live broadcasts and other multimedia applications that are biased towards industrial Internet. Changes in the demand for audio and video market have also given many manufacturers, including Tencent Cloud, the opportunity to penetrate their capabilities into more scenarios.

The 2020 IDC report shows that Tencent Cloud Video Cloud Solution ranks first in the industry in market share, and 90% of domestic audio and video users use Tencent's audio and video capabilities.

As the only cloud vendor in the industry that has integrated real-time communication network TRTC, instant messaging network IM and streaming media distribution network CDN, Li Yutao does not think that Tencent Cloud is integrating just for the sake of integration.

"From a technical perspective, the evolution of the integration of the three networks has begun to converge many years ago. But it is precisely because the epidemic has further stimulated the demand for online video calls that we have found that more and more customers, especially those from traditional enterprises, have begun to move their businesses online."

Before this, although the demand for IM+RTC+CDN scenario integration existed, it was not prominent. A key opportunity point was due to the epidemic.

"In the past, there were two technology stacks for audio and video. One was live streaming, which provided massive distribution and low-cost services. The other was real-time audio and video communication for a small number of people, such as the audio and video communication services of overseas manufacturers and domestic QQ and WeChat."

Five years ago, more people believed that the two technology stacks were targeting different scenarios, but in fact, by last year a more obvious signal was that the two technology stacks had begun to converge at the business demand level.

"For example, when RTC vendors provide online education services, they may first meet the needs of small classes, and then large classes of 10 or 20 people. As they move up, educational institutions will certainly raise demands for low latency, high concurrency, and low-cost services. At this time, the challenge for RTC vendors is that they lack certain technical scalability and do not have infrastructure capabilities like CDN."

Timing is an online education platform. The online study room scenario solution it launched relies on the capabilities of Tencent Cloud.

Timing founder Guo Xiangyun told Leifeng.com that there are three reasons for choosing Tencent Cloud: "First, the cost-effectiveness, because we are still a startup company; second, the stability of the service; and third, the speed of response to problems."

Timing is one of the few companies on the market that attempts to enter the market from the perspective of education + social interaction. In Guo Xiangyun's view, there are many study rooms in offline scenarios, and some users will feel a strong sense of loneliness when studying. If social methods are integrated, it may alleviate the pressure of studying. Through the introduction of online study rooms, Timing has also transformed from a simple learning record tool for personal use to an online study room product with social attributes.

This approach exactly responds to the integration trend of IM+RTC.

Li Yutao told Leifeng.com: "In the post-epidemic era, more and more customers are calling services on Tencent Cloud with three coexisting requirements: high-concurrency, high-bandwidth CDN traffic distribution, high-quality real-time audio and video interaction, and IM communication services."

For example, before this, many customers have chosen to move to the cloud. When calling IM and RTC capabilities, it is more complicated than CDN, and requires higher access costs and the establishment of a technical team.

To this end, Tencent hopes to provide customers with one-stop service, working backwards from customer needs to the underlying technical architecture. First of all, the infrastructure IDC and network construction are naturally together; in terms of technical capabilities, the three networks need to have the ability of local access, massive scheduling, and transmission optimization. Based on these points, Tencent Cloud integrates the three networks together, reuses the underlying infrastructure, and jointly builds technical capabilities. When it comes to upper-level application solutions and product packaging, it can provide customers with one-stop SDKs and APIs.

This further lowers the threshold for customers to directly call audio and video components.

It can be seen that technology is still the lifeline of Tencent Audio and Video, but it is worth noting that there are many audio and video manufacturers who have been in business for many years and serve a large number of customers, especially in the RTC market. And this team not only serves external customers, but also handles most of the audio and video services within the Tencent system, such as real-time audio and video communication, instant messaging, and streaming media distribution. It is necessary to serve the needs of the group and to "receive customers" from the outside. With two identities, the pressure is imaginable.

When it comes to empowering external scenarios, Li Yutao does not hide the challenges he encounters.

"Our biggest difficulty lies in the diversity of our business. The business needs and performance requirements of customers in different industries and enterprises are completely different. This is the first problem we have to face in the process of transforming from ToC to ToB. At the same time, competing in the external market environment means that we will definitely encounter competition from multiple products, especially in the RTC and IM scenarios. Some customers choose the SDK deployment method, and it is difficult to replace them."

According to Li Yutao, the team is also constantly looking for new entry points for demand. For example, if some customers have a demand for a multi-cloud environment, this means that the team can provide customers with the ability to complete multi-cloud scheduling with one click in a minimal form; for example, in some pan-Internet voice chat and social scenarios, the audio and video cost budget of such enterprises is not high. Can they rely on Tencent's own ecological resources, such as integrating the Tencent Music copyright library, and provide it to customers in the form of mini programs or SDKs?

Through various attempts like this, relying on its absolute advantages in the CDN field, Tencent Audio and Video will follow up with solutions in new scenarios and gradually build trust with customers.

At present, in addition to continuing to consolidate its social video applications and overseas business lines, the Tencent Audio and Video team has actually also clearly defined the directions that it will focus on in the future, including online education, online finance, online medical consultation, industrial manufacturing and other industries.

The epidemic has become a catalyst for the expansion of the audio and video industry, and has actually allowed more industry customers to transfer their businesses online through audio and video services. For example, in the education industry, a large number of offline education institutions will build their own online education SaaS tools, and will also consider adding real-time interactive or interactive functions to improve user activity and satisfaction.

In the future, the way users interact will increasingly move towards scenario applications represented by ultra-high-definition 4K/8K, cloud games, VR/AR, video AI, etc. Tencent's "Full Authentic Internet", which is a way to achieve full or near-real-person communication and interaction, is also becoming more realistic due to the solution of network transmission, multi-terminal integration and other technologies, as well as the exploration of more user scenarios.

This article is originally written by Leifeng.com, and the author is Yang Li. Please reply "reprint" to apply for authorization. Reprinting without authorization is prohibited.