What are the factors that affect the audio-visual effect of video conference system?

In the application of video conference system, the factors affecting video and audio effects are mainly concentrated in three aspects:

1) QoS of the network

2) 2) Performance of MCU and terminal;

3) Meeting room design.

First, the network quality of service (QoS)

At present, E 1 private line and IP are the main networks commonly used in video conference system. E 1 private line can provide end-to-end exclusive bandwidth based on circuit switching and time division multiplexing, so the network itself has a perfect transmission quality guarantee mechanism. In most cases, the main factor affecting the transmission effect of E 1 dedicated line is the quality of transmission equipment and transmission lines. For such a factor, we can often improve it by replacing transmission equipment and reducing the line error rate.

However, IP network is based on statistical multiplexing and packet switching technology. When voice, data, video and other services need to be transmitted at the same time, its traditional "best effort delivery" mechanism exposes many problems, the most important of which is that it can not provide end-to-end bandwidth guarantee for each service, which will lead to large transmission delay and jitter. Therefore, we must optimize the IP network through technical means to reduce the influence of the network itself on the video conference system. At present, these technical means have developed into an important branch of IP system, namely Quality of Service (QoS).

The so-called QoS refers to the ability of the network to provide better services for specific network traffic through various technologies. Its main purpose is to achieve priority control, including bandwidth, delay, jitter and packet loss. Almost all networks can use QoS to get the best efficiency.

QoS technology can be divided into three categories, including best effort service, comprehensive service and differentiated service, among which differentiated service is the most widely used. In differential service, the network classifies, queues and manages each packet according to its QoS label. These tags can be IP addresses, TCP port numbers, or specific fields in IP packets.

In actual network planning, network devices (such as routers) are required to provide QoS guarantee mechanism through various technologies with the help of complex traffic management systems, and divide different priorities according to service types, such as voice is the best, video is the second, and data is the last, and then allocate network resources according to these priorities.

For video conference, in order to ensure the bandwidth of video service, the router must be able to identify and classify the video service packets in the passing IP data stream, and then provide bandwidth guarantee and priority delivery service through congestion management mechanism. In this way, when the network is congested, the transmission effect of voice and video services can be guaranteed. At present, mainstream router manufacturers can provide QoS support based on classification, marking and congestion management.

Second, the performance of single chip microcomputer and terminal

In addition to the network should provide a good QoS guarantee mechanism, the video conference system equipment itself should also have good performance, in order to truly ensure the effect of the conference. These performance factors include the video and audio coding and decoding technology adopted by the system, the design structure of the equipment, and the adaptability of the equipment itself to the harsh network environment.

1, video and audio coding and decoding technology

Video and audio coding technology is the key technical index of video conference system and an important factor affecting the conference effect. At present, the video coding technologies used in video conference system mainly include H.26 1, H.263, H.264, MPEG-2, MPEG-4, etc. The audio coding technologies mainly include G.7 1 1, G.722, G.728, G.729, etc.

Among them, two video coding technologies, H.264 and MPEG-4, can achieve high-definition dynamic image effect in low bandwidth with small coding delay. As a new generation of video coding and decoding standards, their advantages are very obvious.

In audio coding, MP3 is an efficient sound compression algorithm, its frequency response range is between 20Hz and 20KHz, and its sampling frequency reaches 44. 1KHz, which supports dual-channel coding, so it is widely used.

2. Design structure of equipment

In the early days, many MCU and terminals in the video conference system used PC as the hardware structure, and the operating system was based on Windows. This kind of equipment has great limitations in coding and decoding performance, packet forwarding efficiency, stability and security, which leads to low video and audio quality and long delay.

As a professional conference room application, most of the current video conference systems choose MCU and terminal equipment based on embedded design architecture. This is mainly because the embedded system has simple instructions and high real-time performance. Combined with special codec DSP, high quality and low delay video and audio signal processing can be realized, and it is also stable and safe.

3. Adaptability of equipment to harsh network environment

Network QoS can ensure the transmission effect of video conference to a certain extent, but its function is very limited, especially in some harsh network environments. The adaptability of the video conference system equipment itself to the harsh network environment will also have a great impact on the conference effect. These adaptive capabilities include IP priority setting, IP packet sequencing, IP packet repetition control, IP packet jitter control, packet loss retransmission and rate automatic adjustment.

1)IP priority (IP priority)

When the network plans the QoS technology of differentiated service mode, it can classify the service packets entering the data network by various matching means, including IP address, IP priority and so on.

Among them, audio, video and RTCP (multicast) data streams can be prioritized by using the IP priority part in the IP packet. When the network uses IP priority for traffic matching, the video and audio packets sent by the video equipment with the IP priority field information modified can be put into the queue for processing, so as to ensure the priority transmission of the video conference code stream.

2)IP packet sorting

Usually, the best-effort transmission mechanism of the network cannot guarantee the correct order of the packets it forwards. For H.323 video conference system, if video devices receive IP packets in sequence, it will bring out-of-order problems, and the loss or delay of data packets will lead to the freezing of video images or the interruption or jitter of sound.

Video equipment supporting IP packet sorting function can solve this problem. When IP packets arrive, the video equipment will verify their sequence, and the unordered packets will be returned to maintain the continuity of audio and video streams sent to the end users.

3)IP packet repeat control

When an IP packet passes through the bearer network, it may produce multiple copies, or it may produce multiple copies when the system adopts retransmission mechanism to adapt to the harsh network environment, which will cause the freezing of video images or the interruption of sound. Video equipment supporting IP packet repetition control can correct this error through this function to maintain the continuity of audio and video streams sent to end users.

4) jitter control

When audio and video IP packets leave the sender, they are evenly arranged at regular intervals. After passing through the network, this uniform interval is destroyed by different delay sizes, resulting in jitter. Jitter will lead to inconsistency between audio and video streams on the target terminal. Video equipment supporting jitter control can eliminate jitter through jitter buffering to maintain the consistency of audio and video streams received by end users.

5) packet loss retransmission

When the network congestion is serious, network devices (such as routers) will lose some video packets according to the buffer size and related processing mechanisms. In the video conference system, video packets are transmitted by UDP protocol, and UDP itself has no retransmission mechanism, which will lead to the phenomenon of image frame loss or mosaic at the receiving end. Video equipment that supports packet loss retransmission can ensure the consistency of conference images by adding packet loss detection and retransmission mechanisms.

6) Automatic rate adjustment technology

In some harsh network environments, reducing the conference rate is helpful to improve the consistency and actual effect of video and audio. If the video equipment supports dynamic rate adjustment technology, the terminal and MCU can automatically adapt to the capacity and performance of the network by detecting the favorable and unfavorable factors on the network, and provide the best video quality for the end users by dynamically adjusting the video conference rate.

The adaptive bandwidth adjustment function of video equipment is mainly realized by detecting the packet loss rate. If the terminal detects that the packet loss rate exceeds the specified threshold, it will automatically reduce the video conference rate and notify other participating terminals to do the same, thus providing the conference rate with the best video and audio effects.

7) lip sound synchronization technology

In video conference system, video signal and audio signal are coded and transmitted separately. Due to the influence of IP priority and the size of video and audio packets, the synchronization packets of video and audio will arrive in different orders, resulting in unsynchronized lip sounds.

There are two main factors that affect the asynchronization of lip sounds: network transmission delay and video and audio processing delay are different.

When audio and video packets leave the sender, the audio packets are synchronized with the corresponding video packets. However, when passing through the bearer network, various queuing algorithms will treat audio packets and video packets differently. This will destroy the synchronization relationship between audio packets and corresponding video packets. The end result is that the voice and mouth shape are out of sync. Video devices supporting lip synchronization can correct this problem by using RTP timestamp information in ip packets. With RTP timestamps, the device can determine which audio packet corresponds to which video packet. Further readjust the corresponding video and audio packages to ensure the synchronization of voice and mouth shape.

At the transmitting end, the time spent processing audio is different from the time spent processing video. The factors that affect this problem include the difference between the speed of sound and the speed of light, the size and shape of the room, and the complexity of audio and video coding algorithms. In order to avoid the time difference, the equipment supporting lip sound synchronization can increase a certain delay at the beginning of the audio stream to obtain the synchronization of sound and mouth shape; Audio delay can also be increased or decreased at the receiving end to correct inappropriate delay settings at the sending end. In this way, when the remote meeting place receives the video conference sound and image, lip sound synchronization can be realized. Meet the needs of some users, and the highest cost is 654.38+10,000 yuan, which can realize stereoscopic projection. This product to be introduced is a four-dimensional cosmic stereo image generator VCT-2005.

The AP converter converts three input stereo 3D signals into two passive 3D output signals. This single left-eye and right-eye information is input to two projectors, such as a low-cost LCD projector, and then viewed through a polarized 3D eyepiece, so that a high-quality 3D image effect can be obtained.

Usually, when we watch a stereoscopic image with a monitor, we will feel the picture flashing when the refresh rate is lower than 60HZ. This is because the images in the display are alternately output by two images, which can actually only reach half of the output refresh rate of the display. Stereo glasses are synchronized with the computer, even if the projector refresh rate is high, it is useless, and the high-refreshed image can't reach the projector at all. Such frequent flashing images will cause the eyes of the viewer to be tired. At present, "Stereo Image Generator" uses polarized light technology to transmit video signals from VCT or PC to DLP/LCD projector and project 3D images on the big screen. The left and right eyes can see two different images at the same time through polarized light angle. Image output is not limited by refresh rate. The left and right eyes see two uninterrupted images output at the same time, which, like our daily habit of watching images, will not flicker, completely eliminating the fatigue caused by flicker. It can make the audience integrate into the digital environment and have the real feeling of being there. With data gloves, space tracking locator, joystick, etc. , can achieve the effect of human-computer interaction.