As the digital multimedia is widely used, video decoding becomes a basic element in embedded system design. There are a variety of video standards, depending on the product to implement one or more standards. Of course, this is not all, the video is only part of the multimedia code stream, and there are also audio or voice that requires parallel processing. Therefore, the synchronization layer of a precise processing store or data stream is required. In addition, video decoding itself is high in performance, requires different system architectures previously based on voice-based and information applications; this is a special challenge to the portable system, while desktop applications are also facing these issues. General Video Standards and Codecs Joint Video (JOINTVIDETEAM, JVT) consists of a video coding expert group of ITU (VCEG) and ISO / IEC moving image expert group (MPEG). VCEG Develops Voluntary Standards, advanced mobile image encoding for session and non-session / video applications. MPEG development international standards for moving image, audio, and combinations of two combinations, encoding, decompression, processing, etc. to meet various applications. In summary, JVT has developed the most popular video standards including iTUH.262 / MPEG2 and H.264 / MPEG-4AVC. MPEG-2VIDEO / H.262: MPEG2 (ISO / IEC13818-2), also known as ITU-TH.262, is the most widely used video coding standard in consumer electronic video equipment. MPE2 video is used for digital TV broadcasts: including ground, cable and direct satellite broadcasting. It can reach 720x576 pixel imaging at a fixed frame rate of 25FPS (PAL) or 30FPS (NTSC). In addition, it is also necessary in DVD video. MPEG-4-SP / ASP: ISO / IEC14496-2 describes the SimpleProfile (SP) / AdvancedSimpleprofile, ASP). Among them, SP is used for next-generation portable terminals and narrowband internet. ASP increases several tools, and the encoding efficiency has increased by 1.5 to 2 times. Both two are increasingly receiving MPEG-4-AVC / ITU-TH.264: Advanced Video Coding (AVC) is a multimedia standard developed by ISO / MPEG and ITU-T Joint Technical Committee. AVC provides higher compression ratios, better video quality, and more than MPEG2, is expected to be used for Internet broadcasting and mobile communications. WindowsMediaVideo (WVM) / SMPTEVC-1: WMV9 is Microsoft's multimedia standard, which includes support flow processing, variable bit rate, and fault tolerance with MPEG-4-AVC / H.264 comparable. In addition to the family computer, WMV9 is currently used in digital projection in the cinema. The encoding used in the movie can be a constant bit rate CBR or a variable bit rate VBR of 7 to 12 Mbps, and reach the DVE resolution (720x480). Design principles Specific target applications determine the system requirements and drag the choice of system design. Personal Media Player (PMP): The personal media player is a portable device with large capacity memory, decoding the stored file, can watch video, listen to music, or browse digital photos. Since it is a portable player, power consumption is critical. Since the coding and stream processing functions are not required, and the screen size is usually relatively small QVGA or CIF format, which can be solved well by DSP-based scheme. The advantage of using DSP is also easy to support a variety of video and audio standards. For example: LSI provides sufficient video processing capabilities for such applications for the authoritative DSP core ZSP500 of high performance embedded systems. In addition, the ZSP500 can also provide outstanding audio decoding capabilities, so that the DSP can run at lower frequencies; thereby making the audio / video subsystem power consumption. Memory, keyboard, display, and file systems can be performed by microcontroller. Figure: Standard video codec scheme. Of course, some PMPs have video output features that connect the player to the ordinary TV. Although the PMP integrated display screen is relatively small, the current video decoding size usually reaches the D1 format. Depending on the function list requirements of the player, the device needs to integrate 1 or more video decoding criteria; for the D1 format video decoding in the TV frame rate (25-30 frames / sec), the hardware video decoder is the best choice. Multiple video decoding criteria may require multiple hardware modules. The implementation of multiple decoding criteria will improve system cost, but will not increase power consumption, because only one module is activated at the same time. Multi-standard audio decoding can be implemented in a small and efficient embedded DSP core without increasing additional hardware costs (except for the space of the storage decoder). The microcontroller will be used to control the DSP and video hardware decoders. Single DSP solutions can also be used when they are only used to play audio - such as a typical MP3 player. The DSP in this system performs audio decoding, file / stream data processing, and also controls keyboard, display, and memory. Since the DSP processing audio decoding is more efficient, the use of ZSP400DSP instead of the microcontroller can significantly reduce power consumption. In addition, DSP-based solutions can be upgraded to support more intensive audio standards (such as AACPLUs), and the microcontroller may not process additional loads. Video mobile phone: The video size of this class is QCIF or CIF, and the frame rate is from 10 to 30 frames per second. Do you need to support a variety of video / audio decoding criteria depends on the functional requirements of the phone. The video will be playback of stream video or short video sequence. With software decoding without additional hardware, it can meet the performance requirements of these devices. Moreover, most calls integrate microcontrollers and DSPs, so video decoding can be implemented by DSP, which also makes low-power consumption skills possible. Other phones may contain some of the functions of the PMP or digital video camera, and a simple DSP design cannot meet the requirements, which requires additional hardware acceleration. Video Desktop Phone: The device provides video conferencing features that need to be completed under strong real-time restrictions. If the frame size and the frame rate are limited, the single DSP solution is still feasible. However, most desktop phones have QVGA or larger display; in addition, users typically expect more voice / video quality than wireless mobile devices. This means that the frame rate should be higher, and the frame rate / card frame rate should be as low as possible, and it may be necessary to use a hardware solution to complete video codec. Voice codec can be implemented by low cost DSP (such as ZSPNEO) or microcontroller in the system. Mobile Digital Camera (DVR): For this class device, multi-standard codes are not necessary, most of them use only single video codec standards. Pure hardware solutions are usually preferred due to large size, high frame rate and low power consumption. Without DSP, the system microcontroller is used to drive the video and audio hardware modules. Set top box: Set-top boxes need to complete the processing of high quality video streams and decoding of recorded video files. Like DVR, the requirements of the video quality (frame size, frame rate, bit rate, and fault tolerance) make it almost only by hardware decoding. Of course, because unlike power consumption requirements, a multi-DSP core can be used to use a multi-DSP core with some basic hardware acceleration scenarios to design a system that can handle various video codecs. Depending on the functional requirements, the set-top box may require a certain flexibility: especially for media players, digital VCRs, or network streaming media. The audio can be processed with software in the DSP. In order to achieve this flexibility, select different codecs to complete playback, encoding, decoding, etc., at which point the DSP for audio processing will be a high-performance processor such as ZSP500 to achieve the latest home theater. Surround sound, sound processing and other functional requirements. Basic design configuration for this purpose, we can consider the following four basic design configurations: The first design contains 1 microcontroller and 1 DSP (MC + DSP); the second design includes 1 microcontroller and 1 DSP, but DSP also controls 1 video encoding / decoding hardware module (MC + [DSP + VHW]); the third design uses 1 microcontroller, DSP and video encoding / decoding hardware module (MC + DSP + VHW) In this design, the microcontroller controls the DSP and video hardware modules; the last design contains 1 microcontroller, 1 video encoding / decoding hardware module, 1 audio encoding / decoding hardware module (MC + VHW + AHW) . In the above designs, the microcontroller is responsible for typical embedded control tasks: including user control connections (such as joystick control), USB / UART / Ethernet drive, protocol layer (such as TCP / IP, HTTP), and the like. Table: Typical Bandwidth MC + DSP-Microcontrollers and DSPs in Various Video Formats for low video resolution (CIF), software upgrades, support for multiple tones, and video standards. DSP is used for audio decoding, video decoding, and voice / video synchronization. Although the performance is limited, the system is very flexible, and this platform can easily implement multiple audio and video decoding format support. MC + [DSP + VHW] - This video hardware module is used for high resolution video / decoding. DSP management audio / decoding, also responsible for audio / video synchronization, and can also be used in painting or other video superposition function. One advantage of the system is that the audio / video subsystem can be designed as a standard multimedia / decoder, easy to implant the system without adding too much complexity. The DSP is the controller of the system multimedia part. Since the multimedia / decoding system and the microcontroller connection are loose, it can be easily integrated into many existing microcontroller systems, making this program with considerable attraction. The edit / decoding system can be treated as an ASSP product having a standard local bus interface. MC + DSP + VHW - In this configuration, the DSP is used for audio / decoding, while the microcontroller is used to implement sound / video synchronization. This requires more complex microcontroller design, but can be used in power consumption compared to the MC + [DSP + VHW] system. Since the microcontroller must coordinate DSP and VHW, other control tasks and all coordination operations are also implemented, so that the program has a lot of difficulties. One variant of this configuration is performed by DSP, audio / decoding, and video coding is still executed by hardware, which requires a powerful DSP, but makes system flexibility and supports multiple video decoding standards. MC + VHW + AHW - In this configuration, the microcontroller performs all tasks outside the sound, video / decoding. Tone / video synchronization is also performed by the microcontroller. This solution is inflexible in the speech subsystem (only the audio codec in the original design can not be upgraded), and the MC + DSP + VHW is very similar. The advantage is that it can work with a particular application, with the best power consumption compared to various previous programs. Editor in charge: GT, read full text
Our other product:
