Editor/Wang Yuhao
Planning/LiveVideoStack
On the evening of December 26, the wechat mini program opened its live streaming capability to five application scenarios including social networking, education, medical, government affairs and people’s livelihood and finance. Compared with native apps and browser-based H5 apps, what impact will miniprogram live have on the audio and video technology ecosystem? Will the inherent flow advantage of wechat bring opportunities or traps to developers and operators? LiveVideoStack invited a number of representative technical people to share their views and thinking.
LiveVideoStack: For this real-time audio and video function provided by the small program, can it meet our general requirements for live? For example, what kind of latency does it have? Whether can satisfy lian Mai and so on?
Liu Lianxiang: The live broadcast of small program is divided into “ordinary live broadcast” and “low delay support”. Ordinary live broadcast supports 2-5s delay, and low delay is within 400ms. Ordinary live broadcast scheme can support external CDN, and there is no problem with mass distribution. The low-latency scheme supports a maximum of 10 streams, which is not suitable for live streaming.
Jun Hejun: Small program live broadcasting can meet most of the needs, because the main demand for live broadcasting is audio and video, and other peripheral functions are also easy to realize. As for Tencent’s live broadcast products I have experienced before, the delay can be reduced to a relatively low level of 100 milliseconds, which is quite rich in this aspect. It has yet to be tried and verified about the lian-mai live broadcast.
Gao Zehua: From the current limited API information, it basically includes collection, encoding, push stream, pull stream, decoding, playback and other functions, and claims to support real-time interaction. But for general non-interactive live broadcast, the implementation is relatively simple, wechat small program can meet the requirements. As for the function of linmai, it is still necessary to analyze its ability and advantages through some testing methods after it is made.
Zhou Sijin: It can meet the general requirements of live broadcasting. We just finished the preliminary test, and the delay is about 1-2s, which is better than the general CDN. But for even mai function, delay a little bit big, the effect is somewhat unsatisfactory.
Yuan Rongxi: The DEMO test provided by Tencent Cloud, a small program, shows that it can provide live broadcasting and RTC by using the RTMP protocol standard. Live streaming can carry out push streaming and play streaming, which is no different from ordinary live streaming platforms and should be compatible with most CDN manufacturers. Wechat mini program specifically mentioned that if acceleration is needed, Tencent cloud UDP + RTMP can be used to reduce the delay, about the delay size officially announced is 400 ~ 800ms, basically meet the real-time demand, of course, the specific need to go through strict testing. From the current analysis, UDP+RTMP is actually a private protocol, because it involves public and private key encryption, Tencent UDT transmission and other protocol details, in other words, just ordinary live broadcast can use other CDN cloud, but to smaller delay, such as Lianmai may only use Tencent cloud.
Peter: The RTC mode guarantees relatively low latency through the private RTMP over UDP protocol. However, it is not clear whether low latency calls can be guaranteed under weak networks. I suggest you to test it.
Zhang Lu: At present, the push-pull flow component provided by the small program mainly supports RTMP protocol, so it should be good for delay insensitive, one-way broadcast business support, such as show business. For some services with frequent interaction, such as one-to-one teaching and in-show connection, the quality of audio and video may fluctuate greatly due to TCP connections. In view of this problem, small program also provides a solution, with Tencent cloud services, small program can use UDP protocol more suitable for audio and video transmission. In addition, it is not known whether the server side of the small program can open the access ability of audio and video data, if open, business side should be able to develop more rich audio and video products.
Jiang Yuqing: The real-time audio and video provided by small program is a technical solution of RTC. In essence, it is not much different from the existing WebRTC, and the daily live broadcast can be fully satisfied. Its delay level, is mainly based on the RTC network building, and is not open interface itself, according to the current some RTC network point of view, the quality of the service provider network under the condition of stable condition, basic can do around 400 ms delay, such delay is not so easy to detect, can support the demand such as wheat. And at present, even the application of mainstream low delay audio and video networks such as MCPK Claw doll is also based on RTC system in the construction. It can be said that these terminals and the low-latency networking functions that can be done on the web are well supported.
Liu Xueci: The upgraded real-time audio and video recording and playing capabilities provided by wechat mini program are explained by two documents: Real-time Audio and video recording Component and Real-time Audio and video Playing Component. So far, it seems to be mainly live, but it also provides rTC-based point-to-point real-time calls, and we haven’t seen multiple calls yet.
Livestream should be based on the integration of Tencent Video cloud, which is no problem for general livestream needs. Liangyan cloud has also conducted a comprehensive evaluation of the delay performance of Tencent Video cloud livestream SDK, which is about 1 second. And integrated into wechat small program will solve the problem of wechat live, which is also the pursuit of many users, with social filter, whitening and other functions. With the current social marketing sharing strategy, I believe that a large number of technology companies that only do ordinary live technology are basically a death blow.
As for Lianmai, currently from the interface did not see the multiplayer interactive video part, so I can not comment on it, it is estimated that it will not work for the time being. However, as the version improves, it should be supported soon. Technically, it wasn’t too much of a problem.
Xu Jianlin: Through the actual test of two-person video call, the delay of two mobile phones under the same wifi is about 700ms, which is consistent with the description of 300-800ms in the document. There is no accurate measurement for one time in the same city, but the subjective perception of delay is not obvious. However, 700ms is relatively high for scenarios with strong real-time interaction. At present, other manufacturers such as Powerinfo, Agora, and Zego provide the target delay of connecting the mic service from 200 to 500ms. This difference was significant in the case of latency sensitivity. In addition to the delay, the demo is also good at echo suppression.
Zhan Xiaokai: I think the real-time audio and video function provided by the small program is very reliable. The general live broadcast and even the requirements of the mic can be met, and the delay of the MIC is acceptable. But should be stable in a short period of time of App will not be migrated to small application platform, after all, for every field platform, maintain good relationship with the high-end user host, and to provide better experience (Native experience will be better than small program) is the most important, and if each big manufacturer to attempt a fast traffic function, or fry a concept, Applets are definitely a good choice.
LiveVideoStack: Wechat small program open real-time audio and video recording and playback functions, does it mean that small and medium-sized enterprises or startups without too much development ability or cost can spend a small cost to get the function of live?
Zhao Jiayu: For enterprises, if they want to gain dividends from audio and video related businesses, they must first pass the technical barrier. After all, not all companies can achieve high threshold and difficult development. Most companies will choose to integrate the AUDIO and video SDK of third-party companies to quickly realize audio and video capabilities in their own applications.
For the introduction of real-time audio and video functions of small programs, I personally think that the opening of the ability of wechat small programs again expands the imaginary boundary of small programs, which can be applied to many scenarios, such as online bank/securities account opening, online insurance loss, multi-person meetings, online education and so on. Enterprises already in wechat can use the latest audio and video functions to realize audio and video functions more quickly, but also rely on the ecology of wechat. For startups with small programs as the main platform, the open audio and video live function can easily realize the live function.
Jun He: I believe this is also one of the original intention of small program, is to enable traditional enterprises and small and medium-sized enterprises with weak research and development ability to get access to the Internet quickly and at low cost. The education sector is favorable for small and medium sized and traditional educational institutions, which can quickly get involved in online education at a lower cost.
Lin zhengxian: Indeed, it has lowered the threshold of live audio and video broadcasting. At the same time, because the small program is embedded in wechat, so the promotion cost of the relevant business will become lower. I think it will have a great future in education, e-commerce and other fields. On the other hand, its release may have some impact on the promotion of WebRTC.
Jiang Yuqing: From the perspective of product development, the two interfaces provided by LivePlayer are very simple and easy to use. Product access is mainly the development of product form, and the development cycle is not very long. It’s kind of out of the box. It doesn’t cost much to develop. However, the main cost of live broadcasting does not lie in development. In the case of high number of users, the cost mainly comes from network transmission. Then the subsequent cost of live broadcasting depends on how network transmission is charged and used.
Peter: It is indeed easier to live broadcast, and the advantage of wechat is that iOS and Android do not need to make two sets of apps, and it is also easy to spread; But the cost of CDN is still to bear their own, this is the big head.
Jian-lin xu: Indeed, the official document has a DEMO key deployment tutorial, I with a basic no pain points (but small program category can’t choose the wrong, and if you use development domain name, need to open the debug client can access, while the Server is not so stable, 502 for a while, a moment not return authentication information, but this is all the business layer). Streaming media development is very demanding for technical reserve, especially for weak network scenarios and low latency. Of course, compared with existing vendors, I think it just provides another choice, but this choice comes with the advantage of wechat traffic.
Zhan Xiaokai: It is certain that startups or individual developers can get the functions of recording and broadcasting at a relatively small cost, but at present the server can only choose Tencent cloud, which is actually a limitation on the one hand.
Gao Zehua: It is difficult to do systematic analysis before getting the final test comparison data. According to internal sources, this engine is not implemented by the built-in WebRTC interface of wechat browser. But the SDK of Tencent cloud encapsulation. If so, there is an opportunity to make some real time audio and video effects.
Liu Xueji: From the perspective of development ability, it is theoretically so, but the premise is that enterprises still need basic small program development ability, generally speaking, the technical requirements are reduced. As the growth of large user quantity on cost, monomer cost will certainly drop. From this point of view, a lot of technical strength can not keep up with THE CDN, IDC will be relatively difficult in the future. It is worth noting that Ali Cloud, Jinshan cloud such enterprises how to deal with. On the other hand, wechat small program is mainly S2B, for small and medium-sized enterprises, and even some small studios, is really good.
Sijin Zhou: Indeed, for small and medium-sized companies, there are more opportunities to choose.
Yuan Rongxi: It is very good news for start-up companies that wechat small programs support audio and video functions, which means that a few lines of code can fully have professional audio and video system functions, allowing developers to focus on business development and greatly saving development costs. For example, some time ago, we helped a friend to solve the online doll grabbing function in the small program, and it took a lot of effort to achieve it through JS + MPEG1. Now Tencent provides the ability of small program, and only a few lines of code to complete a low-delay online doll grabbing, and the stability is much better than MPEG1.
Liu Lianxiang: Up to now, all kinds of CDN and SDK have been very complete, and the access cost is actually very low.
Zhang Nu: I think this function is mainly applied to short video products.
LiveVideoStack: that for large enterprises or their own r & D ability of the enterprise, small program this function open will attract layout, after all, wechat itself has become an indispensable tool in life.
Lin zhengxian: I’m optimistic about it. Because it is still lightweight, many of the heavier features of the broadcast side are not integrated. Second, it also broadens the audience of our business. In the past, there was a big problem with sharing to mobile phones, especially for iOS phones, HLS is basically the only choice. But HLS has problems with high latency, and now small programs have undoubtedly taken a big step forward.
Yuan Rongxi: Audio and video has become a just need and standard function of Internet applications, any cloud companies and large enterprises that provide relevant services will not miss this opportunity, and then there should be a group of cloud manufacturers to do audio and video adaptation and support for micro channel small program, because micro channel small program is a huge traffic entrance.
Zhan Xiaokai: Yes, it is very possible, after all, wechat is such a large flow entrance, but if it is no problem for foreign markets, wechat flow in India and Indonesia is not much, and this year, companies in India and Indonesia can still have the potential for development.
Zhang: The community operation based on wechat has been very mature. Small programs will also be an important flow entrance, so for enterprises that rely on flow, small programs are certain to win the battlefield. Small programs will undoubtedly increase user engagement if they can provide richer native applications.
Xu Jianlin: Definitely, especially the “self-built service” mentioned in the document, which provides more possibilities for large enterprises and even other manufacturers.
Peter: Wechat users have not developed the habit of watching live broadcasts. We need to observe the data.
Jiang Yuqing: It will indeed affect the layout of some enterprises’ livestreaming technologies. The field of live streaming itself is moving towards low latency and high interactivity. It is also one of the core points of future live broadcasting profit. The influence of wechat itself will make live streaming products closer to life. That is, the audience will increase. In this case, it should be beneficial for its own live broadcasting platform.
Zhou Sijin: Large enterprises should carry out the layout, as a good supplement to the APP end.
Liu Xueci: This is mainly from the business point of view, wechat small program support video, such as these large enterprises rely on video business, small program can play a drainage role, will naturally attract small program layout, which is also the small program hope to see. At present, wechat has already added social super entrance APP and real-time audio and video functions, which will further promote the development of small programs and attract further upgrading of similar e-commerce videos, which is the video ecosystem Tencent hopes to see.
Gao Zehua: This is part of the enterprise layout of wechat, which can further increase the platform stickiness of wechat and make wechat gradually closer to the system. In addition, it will also become an enterprise entrance to Tencent cloud, which is a win-win situation for both departments for Tencent.
Liu Lianxiang: The live streaming of small programs is just an extension of the original live streaming, and the live streaming of small programs has very strict industry restrictions, which will not cause drastic changes, but will divert some of the traffic of the original App.
He Jun: Big enterprises should also layout, after all, wechat is a very important traffic entrance, it is also very important to take the initiative. As far as online education is concerned, the live streaming of wechat mini program has applicable scenarios for some “situational teaching”, which can make up for the traditional teaching mode of teachers using desktop client, students using APP or desktop. And it can be combined with the wechat ecosystem, and the enterprise’s public account, service account seamless combination, combined with the existing wechat ecosystem (such as: wechat pay, wechat we media recommendation, etc.).
LiveVideoStack: The first batch of open categories of small program this update has five categories: social live broadcasting, online education, medical care, government affairs and people’s livelihood and finance. Will the opening of this ability also attract other industries? Like camera surveillance and so on.
Jiang Yuqing: Yes, the application of low-delay audio and video transmission is only at the initial stage. In addition to the currently known applications, many other fields, in fact, will also have application scenarios and forms, including the mentioned camera monitoring.
He Jun: I don’t think so for the time being. The original intention of wechat mini program is to be a popular, more universal and easier to spread product. In terms of “camera monitoring”, it is more like a functional demand that does not have the above characteristics. In order to give full play to the capabilities of small programs, existing products need to be packaged and operated in an “Internet” way.
Liu Xuezi: There is no doubt about it. Of course, Tencent is promoting basic video functions as a platform, aiming to provide interface functions of basic PaaS layer and video cloud services. Generally speaking, it will not take all of them, but will leave the innovation of application business layer to the social third party companies to support. Therefore, the next trend should be the gradual transformation of traditional video business companies such as video surveillance and video conference to promote the connection between traditional video business and wechat applets by using wechat real-time video function interface. Traditional video surveillance based on private network private cloud, video conference companies must be transformed, of course, part of the industry must be private network, such as military industry. The traditional Red Sea market of video surveillance will have new development hotspots. After all, video surveillance + wechat, both of these two volumes are too large, and will produce the effect of 1+1>2.
Liu: : There will be attractive, but the current viewing restrictions are too large, camera monitoring is not allowed within the scope of wechat. Other industries can wait and see.
Yuan Rongxi: PERSONALLY, I think the application of small program audio and video in the industry is still relatively inadequate, especially in education and medical care. Take education as an example, online education can not be done simply by setting up an audio and video service. The ultimate goal of online education is the teaching effect, which requires a variety of interaction methods to skillfully cooperate to achieve, for example: Barrier-free writing synchronization, teaching and research system, material presentation and interaction, teaching quality monitoring and so on, some applications may start to make small programs to try online education, but there will be no breakthrough in the effect. This function of the small program will be spread out from the entertainment broadcast and other aspects, after all, the degree of entertainment acceptance is easier than industry applications. For the future, miniprograms will be a key part of the Internet of everything, not just camera surveillance.
Mr. Gao: I think it’s totally possible.
Peter: There are privacy issues with cameras. I guess everyone is sensitive.
Zhou Sijin: There are similar needs in other industries. Tencent has chosen these five categories because it sees huge market opportunities in these five categories. In fact, there are many opportunities in other industries, and camera monitoring is also a “rigid need”.
LiveVideoStack: In this open category also includes the field of online education, does the live streaming function provided by the small program mean an opportunity for traditional educational institutions? For large class, small class or one-to-one class, which is better suited to the platform? How can the “whiteboard” function, which is crucial for education and different from other live streams, be implemented?
Liu Lianxiang: I believe that the open live streaming of small programs will have a great impact on the education industry, especially the one-to-one and small-class classes. Small programs are naturally suitable. Based on the small program of the whiteboard we have been in the development of this demand can look forward to.
Xian Niu: From the perspective of the characteristics and technology of the online education industry, large classes are likely to be the first to benefit from the real-time voice and video capabilities of wechat mini programs. For small classes, since rtMP-UDP allows no more than 10 users to pull low-latency streams, the number of small classes can not exceed 10, but in fact, a slightly larger class will be more profitable. In addition, educational industry-specific capabilities such as powerpoint sharing and whiteboard doodling need to be developed by developers themselves, and the ability to open applets does not necessarily allow developers to develop them themselves.
Zhang Nu: at present, wechat has been applied well in the big class scene. For small programs, there must be a very big improvement in the user experience in the two aspects of delay and students raising their hands. For important functions such as whiteboard and annotation in online education, it should not be particularly complicated to rely on the development language of applets.
Sijin zhou: As far as interaction is concerned, I think it is more suitable for small class and one-on-one. For large classes, it can reduce the delay and improve the experience of interactive live broadcasting. There is no plan for whiteboard.
Yuan Rongxi: For online education, some applications may try the function of small programs, most likely large classes. Why is that? The big class doesn’t care much about teaching results, it focuses on absorbing traffic and converting it into high-value 1-on-1 or small-class classes. Small class and 1V1 short time can not see the small program application, the reason is the teaching effect problem. Because all apps that offer 1-on-1 or small-class classes have their own unique ways of interacting, and even their own hardware, it’s not just a small program that can handle it.
: and the king of the traditional education institutions a rare opportunity to incoming Internet, but most need to change the traditional education institutions or thinking mode, the Internet technology is only a small portion of the complex scenarios (such as a programming class teaching, the teaching scene itself is more complex, not perfect show on the mobile end), may be need traditional client mode.
Wechat applets embed their own encapsulated canvas, which can do simple whiteboard functions based on it, but some advanced whiteboard applications (such as WebGL/AR/VR) applets still lack the ability, which is expected to be enhanced in the subsequent version of applets.
Peter: Wechat authentication payment is relatively easy, whiteboard technology is not difficult to implement, but it is not easy to operate on mobile phones, the experience may be problematic.
Mr. Gao: We can think of it as an entrance to Tencent cloud. If not now, maybe in the future. However, education has its own special attributes, and experience is very important for students and teachers who are paid to attend lectures. From the perspective of education, there are many details and functions to be provided. In my opinion, for the future “wechat classroom”, more applicable mode, or to guide to other more professional online education platform.
Jiang Yuqing: In addition to the real-time audio and video interface, the previous version of the small program has actually opened a drawing interface similar to canvas. In other words, the function of “whiteboard” can be realized in wechat. Transmission through websocket interface of small program and drawing interface can fully meet the needs of whiteboard function. For ordinary users, audiences can be quickly expanded without downloading a separate APP. For traditional educational institutions, it is also a brick to add to the high interaction of education. I think it’s a great opportunity.
However, the core problem of education is not APP or small program, but the quality of teaching provided. Among the excellent educational resources of traditional schools, the threshold of live broadcasting limits the possibility of these excellent teachers to expand education live broadcasting to a large extent. The launch of this small program, for the traditional education industry, should actually lower the threshold of live education. It should be a great opportunity for both live education and traditional education.
Liu Xue time: I think that live and one to one is more suitable for public class, but it is only in view of the small 2 B or c, B for large enterprise, education institutions interactive video technology ability is just one small aspect, based on considerations of cost and business will be more, so you can’t say 1 to 1 is very suitable for this kind of platform, depending on the concrete analysis of concrete object.
For the educational whiteboard, using wechat small program, H5 whiteboard should be a necessary way, but in interactive education, taking The bright Eye Class of Bright Eye Cloud as an example, whiteboard signaling and video control signaling are closely combined and difficult to separate. Applets currently do not provide whiteboards, so it is worth watching in the future. If applets do not provide whiteboards, this can be a real headache for developing educational products. This requires enterprises to provide their own combined solutions. The most extreme, is as bright as the classroom, audio and video bottom and white board are their own implementation, their own integration, so there is no problem.
LiveVideoStack: Applets provide audio and video capabilities, will there be an impact on professional audio and video service providers?
Gao Zehua: The audio and video capabilities provided by Xiaoprogram are more like an entrance to Tencent cloud. Large-scale high-quality paid audio and video services are complicated, and many services need to be cross-regional, cross-industry, cross-age, cross-class and so on. Applets provide audio and video capabilities that can impact small, scattered, domestic developers and small businesses. The purchasers of audio and video service providers are usually large and medium-sized manufacturers, which have a lot of customization requirements. Small procedures can not be met temporarily, so the impact is limited.
Exhibition XiaoKai: I don’t think in a short period of time is formed on the impact of professional audio and video service providers, small program specific to the production environment, after all, at least need a period of time, and the server must go tencent cloud, if tencent cloud of nodes deployed (such as foreign) has a bottleneck or service has a bottleneck, so many platform would be embarrassed, After all, with multiple service providers and then eager flow for the platform is easier to control costs, product stability, if tied to a manufacturer, it may be a lot of passive.
In addition, there are some advanced short video processing, audio processing is not mature (for the national K song sound processing, video processing, Tencent should not open source); In addition, for the short video community, karaoke community, live show and other fields in a short time will not be transferred to the small program to develop new functions; However, it is a good thing for developers, which can make the development cost of multimedia lower and the overall development easier to use. It can let some people or companies with ideas quickly achieve something, fast trial and error, fast iteration.
Liu Lianxiang: Small program ordinary live broadcasting is not bound to service providers, but the current plan to reduce latency requires Tencent cloud services, which will have a certain impact on audio and video service providers, but it will not have a great impact on the demand for audio and video in the whole industry. The two audio and video service providers said they could do a lot of services that Tencent could not provide.
Xu Jianlin: It varies from enterprise to enterprise. If the underlying technology principle is basically the same, the impact will be greater.
Liu Xueci: This is beyond doubt, the impact object is mainly to professional audio and video technology service providers, how much impact depends on the pace of Tencent. As I said earlier, for the pure live streaming technology service providers, the impact is just around the corner, which is a very cruel thing, and they have to think strategically to face this impact.
Yuan Rongxi: For professional video service providers, small programs will not bring impact, but a wave of huge opportunities. Video services will be looking for ways to accommodate this access method by channeling miniapp traffic to their services, much as Apple’s Safari supports WebRTC.
Jun: Yes, especially some small audio and video suppliers without their own characteristics will receive a big impact. Tencent has great advantages in infrastructure, research and development capabilities and so on. Professional audio and video suppliers may need to make their own characteristics in the segmentation of vertical fields, with irreplaceable. You can also try a live streaming platform for small programs, or a third-party outsourcing service.
Zhou Sijin: It will indeed form some impact, but the stability of small program audio and video capabilities, scale and actual effects have yet to be tested. In addition, the value of professional audio and video service providers lies in differentiation and specialization, which has its essential difference.
Jiang Yuqing: It shouldn’t be a big problem at this point. As a result of the long-term work in the pan-entertainment website, it can be seen that the anchor is from the shallow to the deep. It starts with a simple push stream tool provided by an APP or platform that is easy to use. In the future, I will gradually use professional open source push stream tools and even paid push stream tools. The effect of more small programs is to bring in new anchors and lower the barriers for them to live. For real experienced anchors alone, due to some limitations of small programs, when they are inconvenient to push the stream of other games, beauty and other problems, they will naturally be transformed into users of professional audio and video service providers.
Not only will it not cause too much impact, but the expansion of the number of users can be expected under the environment of lowering the threshold and increasing the number of anchors. For professional audio and video services, the benefits should be more than the impact. It is also a good offer for professional audio and video service providers. To give a simple analogy, when the Beijing-Shanghai high-speed railway was completed, most people thought it would be a big impact on the aviation industry. However, a few years later, they found that due to increased liquidity, there were more people making planes, and the market also doubled and doubled. Small procedures open for audio and video services in the field, should have the same effect. The lowering of the threshold of live broadcasting will also have more market prospects.
Peter: It is a new target platform for livestreaming manufacturers. This platform still needs livestreaming CDN. Existing Android iOS native platform also needs live push and pull stream SDK; No real RTC capability and little impact on RTC vendors. However, if wechat does offer real-time audio and video calling capabilities through small programs, it will still have a certain impact on RTC manufacturers.
LiveVideoStack: What advice do you have for developers developing multimedia related services based on applets?
Peter: Try some innovative ways of playing, beyond the existing product form. There may be opportunities.
Liu Xueci: After all, it is a multimedia application, and the relevant multimedia foundation is still needed, such as some basic concepts, such as sampling rate, frame rate, bit rate, delay, as well as live broadcast and vod related technologies, such as RTMP and HLS. If possible, it is also best to learn more about some popular streaming media frameworks such as FFmpeg and WebRTC from the Internet, and you can also do some simple cases.
Zhou Sijin: STILL suggest to experience and test the related function.
Liu Lianxiang: Cautiously optimistic, although the small program of audio and video capabilities are gradually open, but the restrictions are still very large, in addition to the category of restrictions, need to think about their own business is very wechat requirements of the scene fit.
Jun He: The characteristics of small programs are “small and light”, suitable for “short frequency fast” Internet development mode, fast iteration, fast trial and error. Can combine Tencent cloud and other powerful Tencent infrastructure, reduce the cost of research and development, after all, small program must be the most perfect combination with their own products. Some audio and video service providers can consider making a live broadcast platform for small programs. In addition, applets allow commissioning of third party development and maintenance, which may be an opening for traditional service providers to try software outsourcing.
Gao Zehua: No special advice, just like normal API, quick development and quick use. Take advantage of business opportunities. If anything, if you encounter difficulties in the development of real-time audio and video services, you can discuss and share more, rather than build a car.
Afterword.
Since “Mini Program + Live” has just been launched, its functions, performance and stability have not been widely verified, so we will continue to pay attention to it. If you have practices and thoughts that you would like to share, you can leave a comment or contact [email protected].
About this issue (the following list is not in particular order)
Peter, head of the video technology team of a first-tier Internet company
And Jun, tutorABC (formerly VIpABC) big front End principal
He has more than 10 years of experience in front-end and back-end r&d and architecture, and has worked in Internet companies such as Hujiang.com and Tuniu.com. Good at large-scale front-end project architecture, front-end engineering, front-end and Nodejs server performance optimization, etc. Now I am in charge of the front-end department of TutorABC, as well as the browser side and client side of the audio and video teaching platform “Tutormeet+”, and I am committed to creating WebRTC high-performance and rich interactive front-end solution in the field of Internet education.
Gao Zehua is chief audio and video architect of Agora. IO
11 years of music voice codec learning experience, understanding dozens of audio codec standards. Successively led audio projects in Zhonglei Electronics, Shilan Microelectronics, MOTOROLA and Hong Soft Technology. Successively responsible for chip development, embedded system, PC software, mobile APP audio and video subsystem design. Have unique views on the development and application of audio and video communication technology.
Jiang Yuqing, panda live front-end technology expert
She received her bachelor’s degree from Sichuan University and her master’s degree from Dalhousie University in Canada. FedoraRepository Team Member. Now I am engaged in H5 player kernel development, bullet screen engine development and WebGL development at Panda Live.
Lin Zhengxian, head of YY Live Broadcast Department
From 1999 to 2011, he successively served as chief engineer and senior architect in three well-known communication equipment companies. In 2011, he joined YY as T4 engineer and R&D director. Now I am in charge of the technical team of audio and video encoding and decoding, computer vision, audio and video transmission and distribution, etc. Rich experience in wireless transmission, IP core network, Internet access, audio and video live broadcasting and other fields; Led the team to win the company’s technical awards for many times, and applied for a number of patents related to network and audio and video.
Liu Lianxiang is the founder and CTO of dotEngine Audio and video calling cloud
Full stack engineer & products, used to work in Vision China, Guokhu. 6 years of product research and development experience, 4 years of multimedia research and development experience, developed montage short video social app, founder of Volute real-time video collaboration platform, founder of play live
Liu Xueci is the founder of Beijing Bright Eye Cloud Vision Technology Co., LTD
Top video technology business experts and lecturers in the industry. I have worked in technology and management of multinational companies and started many businesses independently. At the same time, I served as strategic technical consultant for several mobile Internet companies. Beijing Liangyan Cloud Vision Technology Co., Ltd. focuses on traditional unified communication solutions and video conferencing products, and in 2015 transformed to focus on mobile Internet interactive audio and video cloud business.
Jian-lin xu
Xu Jianlin (Piasy), focus on Android architecture, essential open source library source guide (unrolling the wheel series), Advanced RxJava series blog translation, currently focused on the client real-time multimedia field, WeBRTC-Android source guide, Work on audio and video SDK development at Powerinfo.
Simon Niu is a senior technical expert and architect of THE Company
Master of Computer science from Beijing University of Posts and Telecommunications, master of Business Administration from The University of Hong Kong, engaged in real-time voice and video cloud service technology research for many years, focusing on interactive live broadcast and voice and video social industry.
Yuan Rongxi, senior architect/audio and video technology director, outstanding student
Core system engineer, 16 years of C programmer, good at solving problems, good at building high-performance service system and system performance tuning, like to solve problems and debug technology of the system. In his early years, he was obsessed with P2P communication network, TCP/IP communication protocol stack and authentication and encryption technology. He once realized real-time video transmission system based on P2P Super node technology. In 2015, I joined Student Jun with excellent academic performance and was responsible for building intelligent routing real-time audio and video transmission system and network for student Jun with excellent academic performance to solve the real-time problem of audio and video communication. In recent years, I have focused on storage systems and concurrent programming. I am interested in PaxOS and RAFT distributed protocol. I especially like database kernel and storage engine, and I have been exploring the implementation and transaction processing model of MySQL/innoDB and WiredTiger. Keen on open source, I have proposed some patches for the open source community. In spare time, I like writing long technical articles and reading Tang poetry.
Zhao Jiayu, NetEase Cloud Communication and video CTO
I have been deeply involved in the architecture and research and development of Cisco Jabber, Webex Meeting, Cisco Spark and other distributed real-time communication products. I have many years of experience in overseas work and management of large R&D teams, as well as the architecture design and research and development of complex real-time communication software.
Zhang Nu, technical director of Baiyun
Graduated from Xidian University, joined Beijing VASu Technology Co., Ltd. in 2006 and began to enter the audio and video industry, engaged in enterprise software video conference system development, design, research and development. Since 2014, I have been working in Beijing Baijiashilian Technology Co., LTD., mainly responsible for the design and development of live broadcast products. It involves distributed server architecture, big data transmission, audio and video codec technology, multi-platform client support and other work.
Zhou Sijin is CEO of Beijing Trimaran Technology Co., LTD
Ploycom was a gun-fighting systems architect at the head of video conferencing; Vice President of Technology at V2, a software video conferencing industry leader; Trained in 2 listed companies (Jiaxun Feihong, Jiucheng), r&d director; Currently, HE is leading a group of brothers to start their own business (Beijing Trisolaran Cloud Technology Co., LTD.) as THE CEO, aiming to become the best interactive video live cloud platform in China.
Zhan Xiaokai, Senior Director of National Happiness RESEARCH and Development
Worked at taobao, development ticket search business, to join sing, 12 years after sing the whole process from online to have 400 million users, is responsible for the development of sing audio and video, which involves multiple product lines, including sing, sing studio, Mars, audio and video collection, the mobile platform hardware codec, cross-platform audio and video processing, has a wealth of experience, Live now broadly happy work, responsible for product business, over the next two months there will be a book about the audio and video development, mobile platform was introduced in detail in the book audio and the whole process of development, is also the years engaged in the development of audio and video mobile platform a detailed summary, hope you can help more people to the audio and video field.