Overview
MPEG-1 is a video and audio compression format customized for CD CD media. The resolution of the VCD is only about 352 × 240, and the fixed bit rate (1.15Mbps) is used, so when playing the video of the fast action, due to insufficient data volume, the macro block block cannot be fully adjusted, and the result is caused by the video screen. Blurred square. Therefore, the output quality of MPEG-1 is comparable to the traditional video recorder VCR, which may be the reason why VideoCD is not successful in developed countries. MPEG-1 audio divided into three generations, the most famous third-generation protocol is called MPEG-1LAYER3, referred to as MP3, has become a widely circulated audio compression technology. MPEG-1 audio technology is between each generation, and the compression ratio is higher than the previous generation. The first generation protocol MP1 is applied to the DGC recorded digital audio and Philips company; the second generation protocol MP2 is later applied to one of the European version of the DVD audio layer.
MPEG-1 has the following characteristics:
Random access
flexible frame rate
variable image size
Defined I-frame, P-frame, and B-frame
motion compensation can span multiple frame
half pixel precision motion vector
Quantating matrix
GOF structure
SLICE structure
audio layering
MPEG-1 audio divided three layers, MPEG-1LAYER1, respectively MPEG-LAYER2 and MPEG-LAYER3, and high-level compatible low-level. The third layer protocol is referred to as MPEG-1LAYER3, referred to as MP3. MP3 has become a widely circulated audio compression technology.
MPEG-1Layer1 adopts 192kbit / s per channel, 384 samples per frame, 32 equippex subbands, fixed divided data blocks. Subband codes calculate the number of subband signal quantization BIT number with DCT (discrete cosine transform) and (fast Fourier transform). The quantization noise is lower than the masking value by a psychoacomy model based on the frequency domain masking effect. Quantization uses a linear quantizer with dead zones, mainly for digital cassette (DCC).
MPEG-1LAYER2 uses 128kbit / s per channel, 1152 samples per frame, 32 subbands, belonging to different division. The psychoacoustic model of the common frequency domain and the time domain masking effect is used, and the bit allocation of high, medium and low frequency sections is restricted, and the bits are allocated, proportional factors, and sampled encoding. Layer2 is widely used in digital TV, CD-ROM, CD-I, and VCDs.
MPEG-1LAYER3 adopts 64kbit / s per channel, with a hybrid filter group to increase the frequency resolution, divided into 6x32 or 18x32 subband in signal resolution, overcome the average 32 subband Layer1, Layer2 A shortcoming of low frequency segment resolution is low. The psychoacoustic model 2 is used to add a non-uniform quantizer, and the quantization value is entropy encoded. Mainly used in ISDN (Integrated Business Digital Network) audio encoding.
MPEG-1 is set in 1992, designed for industrial grade standards, which can be compressed for the SIF standard resolution (for NTSC 352x240; for PAL system 352x288), the transmission rate is 1.5MBITS / SC, play 30 frames per second, have a CD (referring to the laser disc) sound quality, the mass level is basically equivalent to VHS. MPEG's coding rate is up to 4-5MBITS / SEC, but as the rate is increased, the quality of the image after decoding has decreased.
MPEG-1 is also used on video transmission on digital telephone networks, such as non-symmetrical digital user lines (ADSLs, video on demand (VOD), and educational networks. At the same time, MPEG-1 can also be used to transmit audio on the Internet or on the Internet.
Using
MPEG-1 once the main compression standard of VCD is the mainstream of real-time video compression, which can be applied to different bandwidth equipment, such as CD-ROM, VIDEO-CD, CD-I. Compared with M-JPEG technology, there is significant improvement in real-time compression, data volume, and processing speed. MPEG1 can meet up to 16 or more 25 frames / second compression speed, at a definition of 500kbit / s compressed stream and 352 pixels × 288 rows, only 2K per frame size. If you look at the VCD to Super VCD to DVD, MPEG1's 352 × 288 format, MPEG2 can have 576 × 352, 704 × 576, etc., which are used to store synchronous and color sports video signals on CDROM, which is designed to achieve VCR. (Analog Tape Recorder Video Videocassetterecorder; VCR) Quality, its video compression ratio is 26: 1. MPEG1 allows images to compress one/38 on the spatial shaft, and compressed up to 1/5 of the data smaller on the time axis. After the MPEG1 compressed data transmission rate is 1.5Mbps, the compressed source input format SIF (SourceInputFormat), the resolution is 352 pixels × 288 rows (PAL system), the resolution of the brightness signal is 360 × 240, the zoning of the chromaticity signal The rate is 180 × 120, 30 frames per second. MPEG1 has a quantity of 4: 1: 1 for the color difference. MPEG1, MPEG2 is a partial screen that passes a different action. In implementation, MPEG1 can be done by means of an existing decoding chip, rather than excessive relying on the host's CPU as M-JPEG. Hardware compression can save computer resources and reduce system costs than software compression.
but there are also many shortcomings. First, the compression ratio is not large enough, in the case of multiple monitoring, the disk space required by the recording is too large. Especially when the DVR host exceeds 8 roads, in order to save a month of storage, 10 80G hard drives are usually required, or more, the hard disk investment is large, and the hard disk failure and maintenance caused by the hard disk. Second, the image clarity is not high enough. Since MPEG1 is only 352x288, other factors such as capacity, simulation of digital quantization loss are considered, and the playback sharpness is not high, which is the main problem of market reactions. Third, there must be certain requirements for the bandwidth of the transmission image, not suitable for network transmission, especially remote multi-channel video transmission, which is unable to achieve remote multi-channel video transmission. Fourth, the number of record frames of MPEG1 is fixed to 25 frames per second, and the frame cannot be lost, and the flexibility is used. From a widely used compressed chip, there is also a lack of effective regulatory means, such as keyframe settings, sampling area settings, etc., resulting in unsuitable applications in security monitoring, cost.
Overall, M-JPEG and MPEG1 are the mainstream technology of the DVR market due to the technical maturation, but the fatal weakness of both is that the hard disk costs, and cannot meet the needs of security and real-time recording.
Features
Random access, flexible frame rate, variable image size, define I-frame, P-frame, and B-frame, motion compensation can span multiple frames , Half pixel precision motion vector, quantization matrix, GOF structure, Slice structure, technical details, input video format.
parameter
maximum pixel number / line: 720
maximum number of lines / shadow: 576
maximum picture / second: 30 < / p>
maximum macro block / shadow: 396
maximum macroblock / second: 9900
maximum element rate: 1.86Mbps
maximum decoding Buffer Size: 376832bit.
Compression mode
MPEG-1 can be understood according to the hierarchical concept, a MPEG-1 video sequence containing multiple GOP, each GOP contains multiple frames, each The frame contains multiple SLICEs.
Shadow is an important basic element of MPEG-1, a shadow is a complete display image. The species of the shadow have four:
i - image / shadow (node encoded image, intracodededPicture) Reference image, It is equivalent to a fixed image and is independent of other image types. Each image group begins with this type of image. Independent encoding when encoding, only the intra coding technology is applied, thus not referring to other frames, similar to the JPEG encoding.
p - image / shadow (predicting code image, PredictiveCodedPicture) contains previous i or P-paint Difference information. When encoding, motion compensation and transportation \ dynamic estimates are used, and the P is predicted by the forward estimate, the previous I-frame or P-frame before reference is used.
B - image / shadow (predicted before and after the encoded image, BidirectionallyPredictiveCodedPictures) contains i from previous and / or after I Or P-Graphics Difference Information. The encoding also uses motion compensation and transportation \ dynamic estimates, predicts forward estimates, backward estimates or two-way estimates, primarily references the previous or back Id or P.
D - image / shadow (indicating the encoded image, DcDirectCodedPicture) is used to quickly take it. The image constructed only by the DC DC component can be browsed when the low bit rate is time. There are rare to use in actual coding.
Decoding standard
MPEG-1AUDIOLAYER1
Type: AUDIO
Agency: MPEG
required Width: 384kbps (compressed 4 times)
Features: Simple encoding, used in digital cassette tape, 2 channels, audio compression scheme used in VCD is MPEG-1 layer I.
Advantages: Compression mode is much more complicated relative time domain compression technology, and simultaneously encoding efficiency, sound quality is greatly improved, and the encoding delay is correspondingly increased. Can achieve "complete transparent" sound quality (EBU sound quality standard)
Disadvantages: frequency width requirements Higher
App field: VoIP
Royalty Free: Free
Remarks: MPEG-1 sound compression coding is the international standard of the first high-fidelity sound data compression, it is divided into three levels:
- layer 1 (Layer1 : Easy to encode, used for digital cassette tape
- layer 2 (layer2): algorithm complexity medium, used for digital audio broadcast (dab) and VCD, etc.
- Layer 3 (Layer3): The encoding is complex, the transmission of high quality sounds on the Internet, such as MP3 music compression 10 times