Programmable DSP may use in realizing each kind of existing codec encoder-decoder and the future arranges the decoding standard. The present tendency will be issues every two years new will arrange the decoding standard, each new standard will meet the need more DSP cycles. Therefore, the choice has the compatible development blueprint DSP platform (for example ZSP) to be important, like this through system upgrade, but the redesign then satisfies future system request.
Multimedia will designate all elements (including writing, audio frequency, frozen picture, video frequency and graph) integrate in the sole media object. Flows the technology (Streaming technology) to be possible while the media object is read, listens to or onlooking live transmission these objects. Before RealAudio Corporation in 1995 provided the first commercial class media product, after the major part Internet media file needed to download completely, could broadcast. But the present, does not need to wait till the class audio frequency to cut the piece (clip) downloading to complete can broadcast, but may playback in the transmission process.
Flows the media data to transmit by the server end, and by client side receive and real time display. When receives the buffer contains avoids data losing sufficiently enough many information, the client side may start to broadcast the audio frequency/video frequency, avoids information losing. The special-purpose multimedia server usually uses in implementing based on the network class transmission. The multimedia servers may transmit the data continuously, but will not present the eruption type transmission and the long time stagnation, therefore before playbacking will start, the client side only needs the very few cushion data.
Already appears on embedded DSP the audio frequency/video compression algorithm provides the class transmission to need the real-time performance the key. Because they have to the numerical data carry on the code and decoding ability, this kind of algorithm is called the codec encoder-decoder.
Although the class transmission usual and the distributed computer network’s relations are closest, but other form’s digital communication needs to flow the transmission similarly. The digital audio frequency broadcast (for example DRM, XMSR, Sirius Satellite Radio), the digit broadcast television (for example Direct TV, South Korea’s T-DMB), the 3GPP handset and the blue tooth handset and so on needs to use the codec encoder-decoder, satisfies the class transmission request. In addition, the codec encoder-decoder in the memory compression and so on the non-class transmission application is also very useful.
The standardized codec encoder-decoder provides the topnotch the interoperability. The class transmission audio frequency standard includes: MPEG1/2, Layer 3(MP3), digital Dubi AC-3, MPEG2 AAC, WMA and Ogg Vorbis. General video compression standard including MPEG2, MPEG4 SP/ASP, MPEG4 AVC/H.264 and WMV.
Standard audio frequency codec encoder-decoder
MP3:MP3 uses in indicating MPEG1 Layer originally 3, but in the daily application already developed to contains Layer 1, Layer 2 as well as Fraunhofer Institute expands MPEG 2.5. MP3 obtains one of highest cognition codec encoder-decoders, has the biggest user group in the Internet codec encoder-decoder. But to achieve near CD quality audio frequency, to the content which certain difficult to deal with needs to be higher than the 192kbps speed.
MPEG1, Part 3 (ISO/IEC 11172-3): Has defined the duplex sound track, the sampling rate is 32, 44.1 or 48KHz, the encoding rate from arranges the decoding method 32 to 384kbps. This standard described three related methods: Layers I, II, and III. Layer III provides the highest compression ratio, but the order of complexity is also highest.
MPEG2, Part 3 (ISO/IEC 13818-3): Has provided two important improvements to the MPEG1 standard. First, the low bit rate demand through enables “low sampling rate (LSF)” the expansion standardization to obtain satisfies. This codec encoder-decoder has provided 16, 22.05, and the 24kHz sampling rate encoding method; Next, the MPEG1 pattern expands for the support to 12 sound track’s voice data. The Fraunhofer low frequency expansion, namely MPEG2.5, has provided the MPEG2 half sampling frequency choice: 8th, 11.025 and 12kHz.
Digital Dubi (AC-3): At present, digital Dubi has the biggest multi-sound tracks to arrange the decoding user group. Through the multi-sound track conformity to the sole code object, digital Dubi has realized the high quality, the low complexity audio frequency compression. Although this algorithm with codes sound track’s quantity to have nothing to do with, present realized the plan already to accept the SMPTE suggestion, namely has used uses in the bass by 5 full bandwidth audio frequency sound track and 1 the minute band width sound track being composed 5.1 sound tracks: Respectively be left, middle and right, the left encirclement, the right encirclement and low frequency expands (LFE).
Digital Dubi supports the nimble broadcast way: 1 sound track to 5.1 sound tracks, 32, 44.1 or 48KHz sampling rate, bit rate from 32 to 640kbps. After the decoding audio frequency may match the broadcast system to provide the best quality sound effect which automatically has nothing to do with the acoustic disposition.
aacPlus series codec encoder-decoder: Coding Technologies Corporation already developed a series of obtained the codec encoder-decoder which the international standard organization widely uses. MPEG2 has used AAC, provides close CD by 128kbps the quality, even if regarding specially complex content also so. aacPlus v1 by the DVD forum, DVB, organizations and so on Digital Radio Mondiale, 3GPP2 and ISMA decides as the standard. aacPlus v2 started at the end of 2004 commercial, has been assigned is in the 3GPP high quality audio frequency codec encoder-decoder, the aacPlus v 2 all modules is the MPEG-4 audio frequency standard constituent.
The AAC:aacPlus series codec encoder-decoder is revolves the AAC core which MPEG2, Part 7 (ISO/IEC 13818-7) describe to establish. AAC provides 8, 11, 12, 16, 22, 24, 32, 44, 48, 63, 88 or the 96kHz sampling rate, as well as reaches as high as 48 sound track’s audio frequencies, each sound track bit rate may reach as high as 288kbps. It has defined three close related plans: Low complexity (Low Complexity), Main and expandable sampling rate (SSR). Low complexity AAC-LC needs the very few processor resources, therefore usually uses in the embedded application.
MPEG4, Part 3 (ISO/IEC 14496-3): Increased consciousness noise for MPEG2 AAC to substitute the (PNS) tool, therefore the definition was MPEG4 AAC. PNS through to kind of noise signal parametrization code, thus simplifies these signals the expression method. Cannot in PNS and MPEG2, MPEG4 time domain noise reshaping (TNS) confuses.
aacPlus V1: Sometimes this codec encoder-decoder is called “highly effective AAC” (HE-AAC). Its conformity basic AAC codec encoder-decoder and the frequency band duplicated the (SBR) technology. SBR is one kind of band spread technology, may cause nearly any audio frequency codec encoder-decoder when the bit rate drops 30% still could guarantee the acoustic fidelity. SBR adds on some coding parameter through the use frequency band low half partial informations to express the frequency band the high half partial informations. The SBR technology may also use in other codec encoder-decoders, for example the union brought MP3 SBR to constitute the MP3Pro codec encoder-decoder.
aacPlus V2: Increases the parametrization stereo sound (PS) technology in aacPlus V1, has formed aacPlus the V2 codec encoder-decoder. The PS technology uses the left sound track and some extra coding parameter, produces the right sound track, further cut the bit rate. aacPlus V2 may achieve the DVD5.1 sound track quality under 160 Kbps, may achieve near CD stereo sound quality in 48Kbps, may achieve the extremely good stereo sound effect under 32 Kbps, may achieve the entertainment quality stereo sound effect under 24Kbps, when is lower than 16Kbps may achieve the high quality single track effect. the aacPlus V2 efficiency causes the motion digit broadcast possibly to apply newly into.
WMA:WMA is in Windows Media Series which Microsoft is authorized a series of widespread use audio frequency codec encoder-decoder. This series newest edition is WMA9, WMA9 Professional, WMA9 Lossless, WMA9 Voice and WMA9 Variable Bit Rate(VBR). In the embedded application, WMA9 is this series most common codec encoder-decoder; Provides 16 /320kbps double channels, the sampling rate reaches as high as 48KHz. “Professional” supports 24, the 96KHz sampling rates and reaches as high as 128 arrives at 768kbps 7.1 sound tracks. Is the same with digital Dubi, after the decoding audio frequency may match the broadcast system automatically, provides with the acoustic disposition has nothing to do with the best quality sound effect. “Lossless” uses in the CD archive, the compression ratio between 2:1 and 3:1. “Voice” uses in reducing the pronunciation to 20kbps. Although VBR is not ideal regarding the majority of class application, but WMA9 and “Professional” can by the invariable bit rate code. “Lossless” always uses the VBR function.
Ogg Vorbis: In order to not need the patent expense the open resources, has is approximate in the MP3 acoustic fidelity. “ogg” is the vessel form, but “Vorbis” is the audio frequency codec encoder-decoder. Because it has avoided related with the MP3 game music according to each game collection permission expense, therefore Ogg Vorbis the utilization ratio rises day by day in the computer games manufacturer.
Standard video frequency codec encoder-decoder
Union video frequency group (JVT) is composed of ITU video frequency code expert group (VCEG) and ISO/IEC movement image expert group (MPEG). The VCEG development voluntary standard, uses in talking and the non-conversation class sound/video frequency application advanced migration image coding. MPEG develops the international standard, uses in moving the image, the audio frequency and both combination content compression, the code, the solution compression, processing and the code expression and so on, satisfies each kind of application. In brief, JVT had already developed including ITU H.262/MPEG2 and the H.264/MPEG4 AVC most popular video frequency standard.
PEG2 Video/H.262: MPEG2 (ISO/IEC 13818-2), is also called ITU-T H.262, is in the present expense class electron video equipment uses the most widespread video frequency to code the standard. The MPE2 video frequency uses in the digital television broadcast: Including land, submarine cable and direct satellite broadcast. It can in 25fps(PAL) or 30fps(NTSC) fixed frame rate issuing 720×576 picture element image formation. In addition, it is also in DVD-V the essential codec encoder-decoder.
MPEG4-SP/ASP:ISO/IEC 14496-2 described MPEG4 Simple Profile(SP)/Advanced Simple Profile(ASP). And, SP uses in the next generation portable terminal and the narrow band Internet. But ASP increases certain tools, the coding efficiency enhanced 1.5 to 2 times. Their two obtain more and more acceptances in the market.
MPEG4-AVC/ITU-T H.264:AVC is the multimedia standards which develops by ISO/MPEG and the ITU-T union Technical committee. AVC provides a higher compression ratio, the better video quality and compared to a MPEG2 higher fault tolerance, uses in the Internet broadcast and the two-way charges for cellular phones hopefully.
WVM/SMPTE VC-1:WMV9 is Microsoft’s multimedia standards, its characteristic including support class processing, changes the bit rate as well as the fault-tolerant tool which compares favorably with MPEG4-AVC/H.264. Except uses in the family computer, WMV9 also uses in the digital projection at present in the movie theater. In the movie uses the code may be 7-12Mbps constant bit rate CBR or changes bit rate VBR, and achieves the DVE resolution (720×480).
Embedded DSP class media solution
Because a series of reasons, DSP becomes the class media codec encoder-decoder’s ideal choice. First, each kind of different codec encoder-decoder and transforms unceasingly the standard request solution must be programmable; Next, the majority of codec encoder-decoders operate the intensity, but DSP is the design uses in the highly effective mathematics operation; Moreover, the power loss and the cost are in the translatory flow transmission important considerations, but the DSP essence has provided the low power loss, the low cost optimal fit.
The typical audio frequency/video frequency class media system usually simultaneously uses the internal memory and the exterior memory. The internal memory is by the DSP essence clock speed movement active memory; The exterior memory is quite slow, the price is also cheap. Arranges the decoding instruction to save in exterior memory, but downloads to the internal memory in carries out. Because the video frequency class data quantity is huge, only if must, it usually place outside the piece, but the audio frequency class data may internal provincial appointment Italy lay aside, but may also according to need to hang some IP module on the system SoC main line.
Around compatible DSP platform
The present tendency will be issues every two years new will arrange the decoding standard, each new standard will meet the need more DSP cycles. Therefore, the choice may the DSP platform which evolves according to the compatible development blueprint be important, like this through system upgrade, but the redesign then satisfies future system request. ZSP provided adapted the flexibility which and the performance the multimedia standards transformed unceasingly must.
LSI the logical company’s ZSP product department provides the software compatible DSP essence which the entire series may synthesize, and provides the widespread sound/video frequency standard code, in the product development blueprint essence is the code is compatible. The widespread third party partner network guaranteed that the new standard will be rapid available. May with ease adapt based on the ZSP sound/video frequency system with the emerging sound/video frequency standard.
Each generation of (G1/G2/G3) ZSP based on the construction which easy to program. The ZSP essence optimizes specially for the low power loss application, is motion application and so on individual sound/video frequency player ideal choices. ZSP has 16/32 bit data channel, supports control performance which high grade audio frequency processing and the video processing need.
The G2 essence has the formidable association processor connection, supports inserts to the essence carries out the assembly line hardware accelerator. The hardware accelerator may loosely or the coupling carries out the assembly line closely in ZSP. The close coupling accelerator may regard as ZSP the set of instructions expansion, this causes them easy to program and the use, even if based on the C code.
The ZSP essence has the remarkable compiler, not only support assembly code highly effective development, but also supports the C code the highly effective translation; In addition the ready-made available standard code, had guaranteed the fastest system design and realizes.
