DIGITAL-S (Digital S), a digital video recorder for broadcasting;

The format of DIGITAL-S video recorder is 1995. In April, Japanese JVC Company introduced a new professional digital video recorder for broadcasting. It is a video recorder developed on the basis of S-VHS technology, and adopts the efficient coding digital technology of S format. It plays the image signal of S-VHS. The width of the video tape is 1/2 inches, which is designed according to the COIR60 1 standard. It adopts 4: 2: 2 sampling, carries out intra-frame coding (M-JPEG) at a digital sampling speed of 50Mb/s, and the compression ratio is 3.

Specification number -S

Video recorder mode component video recorder mode

Image sampling mode 4:2:2 component mode

Sampling frequency y: 13.5mhz, r-y/b-y: 6.75mhz.

Quantization bit number is 8 bits

1 number of recorded samples in line 720

The number of lines recorded in 1 frame is 578 lines (625/50) and 482 lines (525/60).

DCT mode of image compression, 1/3.3.

Mode/compression ratio

The recording rate is 50Mbps.

Sampling and quantization of uncompressed sound 48KHz 16bit.

The number of channels is 2ch (format can be 4ch).

The total recording rate is 99Mbps.

Error correction Rs coding

Image internal code (85,77), external code (149, 138)

Internal voice code (85,77), external voice code (14,9)

Modulation mode SI-NRZI mode, 24-25 conversion

Tracking mode control pulse mode

Drum diameter/speed 62 mm /75 rpm.

The number of tracks/frames is 12 (625/50) and 10 (525/60).

The tape speed is 57.795 mm/s (625/50),

57.737 mm/s (525/60)

Relative velocity14.5m/s

Track spacing (width) 20μm m.

The shortest recording wavelength is 0.58 μ m.

Metal coated magnetic tape with magnetic tape, with bandwidth of12.65mm.

The cassette used is the same as the W-VHS cassette (188mm× 104mm×25mm).

The recording time is 104min( 124min is under development).

The deck mechanism is based on the mature and reliable S-VHS deck mechanism. The magnetic drum is 62mm in size and divided into three layers. The upper and lower drums are fixed, and the middle drum rotates with the rotating head. This form can reduce the tension of the tape, and separate the recording head and playback head of the clip video recorder, and the playback head is 90 degrees ahead, so this machine has advanced image pre-reading function: that is, before the video recorder records a new signal, the digital signal on the tape can be played in advance at the same point. With this function, the functions of scribing and drawing overlapping of the two-in-one editing system can be completed without a third machine.

1, magnetic tape and magnetic box

The tape used by Digital S is W-VHS tape, which can record and replay HDTV broadcast. It is a metal-coated tape with a magnetic resistance of 1800 oster, which is the highest performance of its kind. High signal-to-noise ratio can be obtained at high frequency, and the dustproof structure of the tape box is strengthened in the aspect of tape running performance. The recording time of W-VHS tape is 104 minutes.

2. Tape recording format

The recording and playback heads are each arranged in pairs with four video heads with an azimuth of 15 degrees, and two tracks are recorded at the same time. In the 625/50 system, the video data of 1 frame is recorded by 12 track.

The audio signal can record four kinds of sound data in one frame of image, and these sound data are recorded in groups of every two tracks. The channel of digital S can have four channels in format, but there are only two channel models on the market at present.

Digital S is equipped with four playheads. Even if one or two heads are occluded, the image of the previous frame can be occluded because the image is changed every two frames.

3. Video and audio processing of digital TV.

Firstly, analog Y, CR and CB should be digitized in 4: 2: 2 format, and the brightness sampling frequency 13.5MHz should be synchronized with the horizontal synchronization signal. According to the standard of 4: 2: 2, the number of effective lines per frame in 625/50 format is 576, that is, 23 to 3 10, 335 to 622. DIGITAL-S adds two lines of inverse process, mainly to save the original data of the inverse process. See the following table for the sampling specifications of image signals.

Fg 525/60 system 620/50 system

Sampling frequency y

CR/CB 13.5MHz is synchronized with the horizontal synchronization signal.

6.75 MHz

Total number of samples /y line

CR/CB 858 pixel 429 pixel 864 pixel 432 pixel

Number of valid samples /y line

CR/CB 720 pixel 360 pixel

Effective lines/frame 480 lines 576 lines

V-AUX uncompressed line 2 line 2 line

Effective lines: odd fields

Even field 23...262

285……524

23……262

285……524

The quantization form y, CR/CB is 8-bit linear quantization.

Quantization stage 1 ~ 254

Quantization level YCR/CB white level 235/ black level 16

Gray scale 128

Digital Channel Interface (DCI) can include various identification data of digital video recorder, channel coding and modulation processing required for effectively recording and playing back digital data on magnetic tape, etc.

In order to ensure the interchangeability of magnetic tape, Digital S uses three ITI signals (see ITI0, ITI 1 and ITI2 in Figure 4) for tracking servo. The track width recorded at 15 degree azimuth is 20μm, and the width of the playback head is 1.4 times of the track width. However, due to azimuth recording, it is not easily affected by the noise of adjacent tracks, so that a good playback image can be obtained.

When inserting and editing, the noise of the same azimuth signal is more serious than that of the adjacent track. In order to solve this problem, the digital S is edited by erasing signals with a width greater than two tracks with a rotating degaussing head, and then recorded with a recording head with the same track width. Some guard bands can be formed at the entry point and the exit point, thereby preventing the image quality at the editing switching point from deteriorating.

4. Track structure

The numbered digital signals are divided into blocks and sorted. The method of blocking and sorting is shown in Figure 3. CR and CB form DCT blocks at an interval of 8×8 pixels (DCT pixels convert each frame of an image into a plurality of 8×8 pixel blocks), and then a macro block is formed by four blocks (two Y blocks, a CR block and a CB block) of a luminance signal and a chrominance signal at the same position on the image screen. Like synchronous blocks, macroblocks are error concealment units in error correction. Fifty-four adjacent macroblocks form a super block, and adjacent tracks are recorded at different azimuth angles to increase the signal-to-noise ratio of useful signals during reading. So the super block is divided into two groups, each group has 27 macroblocks, which are recorded by positive and negative azimuth angles respectively. Every frame * * * has 720× 576× 2 = 829,440 pixels (625/50), and * * * has 829,440 ÷ 64 ÷ 4 ÷ 54 = 60 super blocks. Similarly, for 525/60, there are 50 super blocks per frame. These 60 super blocks are divided into 12 lines, with 5 super blocks in each line, recorded on the 12 track, divided into 6 pairs, and then recorded on the tape in turn. The partition on each track is shown in Figure 4.

Each track starts from the lower end of the tape, and * * * is divided into interpolation and track information segment 0 (ITIO: insertion and track information -O), video segment I (video -i), interpolation and track information segment I(ITI- 1), subcode segment (SVBCOE), audio segment I (audio -i) and audio segment I (audio Each region has a front code and a back code. In addition, there are two prompt tracks and control tracks in the transport direction, which can obtain good editing operability.

J (Video -j) and Interpolation and Track Information Segment 2 (ITI-2). There is a coding gap between the start area and the end area, namely G0 to G8. Each region has a front code and a back code. In addition, there are two prompt tracks and control tracks in the transport direction, which can obtain good editing operability.

Betacam of SX:

B etacam-SX is a Sony digital video recorder, which adopts the extended 4: 2: 2p @ ml standard of MPEG-2 MP@ML. Therefore, the video recorder has a high compression ratio (10: 1) while ensuring high image quality. A video recorder equipped with SDDI (Serial Digital Data Interface) interface can transmit recorded data with coaxial cable at a speed of 4 times the playback speed.

The connotation of the 4: 2: 2p @ ml compression scheme is based on the image quality and transmission rate of the main class MP @ ml(main level summary) in MPEG-2 standard. At the same time, considering the requirements of transmission materials and production, the 4: 2: 0 sampling in the main class and the main class is changed to 4: 2: 2 standard sampling, which relatively improves the quality of transmission materials. Then MPEG-2 compression is applied, in which, considering the editing requirements, a GOP(GROUP OF PICTVRES) sequence composed of I-B frames is adopted, so that the editing system of data stream is synchronized with the GOP boundaries of two frames and four fields, and the switching point is always selected on the I-B frame, thus realizing high-precision editing and ensuring the production quality of the in-point picture.

1, tape recording format

The format of Betacam-SX orbit is shown in figure 5. Use 1/2 inch metal coated tape. Unlike other video recorders, the track is not organized by frames, but by GOP (equivalent to two frames, namely I frame and B frame). For 525/60 system, each GOP is organized into 10 ramp tracks after adding error correction code; For the 625/50 system, use 12 to tilt the track. In the arrangement of tracks, video tracks are placed on both sides, and 8 audio data and 2 system data are placed in the middle. There are also three longitudinal tracks, an auxiliary track, a time code track and a control track.

The main parameters of SX are as follows:

Video processing: 525/60 625/50,

4: 2: 2 processing, 4: 2: 2 processing,

8 bits/sample 8 bits/sample,

507 lines/frame 608 lines/frame,

Video auxiliary data: 1 line/field,

Number of tracks: 10 /GOP 12 /GOP,

Tape speed: 59.515mm/sec, 59.575mm/sec,

Track spacing: 32μm,

Drum diameter: 81.4mm.

Drum speed: 75Hz

Recording rate: about 40Mbps

Video compression: 10: 1 compression, and the output bit rate is about 18Mbps.

Video quality: y: 25hz ~ 5.5mhz, metal band.

R-Y, b-y: 25hz ~ 2.0mhz, metal belt.

Serial number & gt48dB

Audio: 4-channel linear PCM,

16 bits/sample

48kHz sampling

Hard disk capacity: 90 minutes DNW-A 100P

Hard disk editing: the minimum editing segment is 0.5 seconds.

EDL table editing and output

Tape to hard disk: 4x speed loading

Video input and output interfaces: SD 1, SDD 1, analog component,

Simulated recombination

Audio input and output interface: AES/EBU digital, analog.

The track width of Sx is 32mm, which is wider than the other two models. Generally speaking, the wider the signal track, the higher the signal-to-noise ratio and the smaller the code rate of digital signals.

2. Digital video processing

When editing the hard disk, first copy any audio and video data of the tape to the hard disk, and then edit it in any order according to the file index. This process will not really change the order of the contents of the disk, but it will change the EDL. You can see the interface of SDDI codec in Figure 6, so that the data recorded on the tape can be copied to the hard disk at four times the playback speed. Betacam-SX has strong error correction ability. Even if two tracks are lost in a GOP, error correction can still be carried out normally.

Another feature of Betacam-SX is that the playback head is doubled. Generally speaking, the magnetic head can read data normally, which reduces the requirement for track accuracy. Using this technology, the image without clutter echo can be played back in the range of 1 double speed.

DVCPRO:

DVC (Digital Video Recorder) is a domestic DV digital component video recorder jointly developed by more than 50 companies in the world such as Panasonic. DVCPRO is a digital video recorder format that fully excavates and expands domestic DV, and its biggest feature is small cassette and long playing time. Because of the use of 1/4-inch magnetic tape, the tape conveying mechanism is also reduced accordingly. Its compression scheme comprehensively considers all possible parameters such as compression damage, motion code rate, subjective picture quality, durability and multi-generation reproducibility, thus producing an integrated system from ordinary digital TV to HDTV, including all levels. Using 4: 2: 2, 4: 1: 1 or 4: 2: 0 digital components for ordinary TV signals, the resolution is very high (the same as D 1 and D5), and the compression ratio of 4: 2: 2 is 3.3:/kloc-0. 4: 1: and 4: 2: 0 use the compression ratio of 5: 1 and the code rate of 25mb/s. With the decrease of the absolute code rate, DVCPRO can make the compression system work in the frame, so it can control and edit the frame accuracy. Although the compression schemes are inconsistent, the mechanical structure, tape cassette, recording format and data structure have been accurately described and defined. Moreover, most circuits can be packaged into two integrated circuits (that is, the chipset is the same), which is compatible.

1, tape recording format

The magnetic tape is made of highly reliable metal particle tape (MP) with a total thickness of 9 microns ... Because the metal particle tape is suitable for longitudinal recording (unlike the metal coated tape), it can provide analog audio prompt track (at the upper edge) and control track (at the lower edge). The use of control track can make the fast servo system lock quickly during the transition from stop to playback, which is the key to shorten the editing time of pre-roll and frame fine editing. Hint track not only provides users with a third audio channel, but also provides another way to obtain audio when digital audio is unavailable during audio search.

The track format of DVCPRO is 2 1.7mm straight drum. Compared with the usual 1/2-inch video cassette recorder, the diameter of the drum of DVCPRO video recorder is 3.5 times smaller, which not only reduces the total mechanical volume and weight of the video recorder, but also reduces the power loss and rotating noise of the drum. The rotating speed of the video recorder drum is 150 rpm, the width of the spiral scanning recording track is 18μm, and the track index (the ratio of the width to the length of the recording track) is three times that of the D3 format, which is similar to the analog format. Six magnetic heads are installed on the drum, namely, two recording heads, two instant playback heads and two rotating degaussing heads. The rotating degaussing head is used to assist the insertion and editing functions. The instant playback head can be used to play back images while recording, so as to check the recording quality immediately. You can also play back the forward and backward directions in slow motion with the recording and playback head. The channels are sampled at 48kHz, quantized at 16bit, and uncompressed digital audio signals of two channels are recorded at the same time.

The ITI (Insertion and Tracking Information) code at the beginning of the track is mainly used to provide insertion and tracking information when inserting code. New data will be recorded after ITI code, and ITI code will not be rewritten except for re-recording. Subsequent audio data segments and video data segments have editing gaps left and right ***3 segments: G 1, G2, G3. After G3, there is a subcode that records longitudinal time code or reverse time code data. There is still a certain margin behind. Every frame 12 track (625/50 system), azimuth is recorded at 20 degrees. The tape speed is 33.8 13mm/s, the audio frequency is dual-channel PCM, 1 channel longitudinal cue track, and the sampling frequency is 48kHz. There are also two independent LTC/UITC time codes. Two time codes can be read at any tape speed.

The track inclination θ is 9. 1784 degrees, and the track width is controlled to 0.4mm, indicating that the track width is 0.35mm, the total length of spiral scanning track is 32.842mm, and the video recording length is 27.548 mm ..

2. Video processing

DVCPRO includes DVCPRO50 and DVCPRO25, which can be switched between four sampling formats: 4: 2: 2, 4: 1: 1 and 4: 2: 0. The compression ratio is also different and can be switched. Their signal structures are compatible, and they can also be switched between 525-line and 625-line TV systems, and they can record and play back materials in 25Mb/s and 50Mb/s formats. DVCPRO realizes 4: 2: 2 recording by using two sets of DVCPRO chipset in parallel. The compression ratio of DVCPRO50 is reduced to 3.3: 1, while that of DVCPRO25 is 5: 1. Here, 4: 1: 1 is taken as an example to illustrate the video signal processing process. The sampling frequency of video brightness signal is 13.5MHz, and the color difference signals are 3.375MHz and 3.375mhz. The brightness resolution is the same as that of D 1 and D5, and the chroma bandwidth is 1.5MHz, which is the same as that of Betacam-SP. Video data compression mainly adopts DCT and variable word length coding.

Firstly, the 4: 2: 2 digital signal is transformed from 21:1to 162Mbps after 4:1. After transformation, it is input to the image compression coding unit, then compressed from 162Mbps to 24.948Mbps, then sent to the recording symbolization processing unit through error correction coding and scrambling, and recorded on the magnetic tape by rotating the magnetic head.

The specific flow of rate compression is shown in Figure 9. There are 526 effective lines in 625 lines in each frame, with odd fields ranging from 23 lines to 3 10 lines and even fields ranging from 335 lines to 622 lines, with 288 lines per field. The quantization level of luminance signal is from 1 to 254, the white level is 235/ the black level is 16, and the quantization level of color difference zero level (gray level) is 128 * * 255. Like S, the brightness and color difference samples are divided into image blocks according to 8×8, and the image blocks are located according to (x, y), and the level is. Y, CR and CB are divided into blocks according to the frame, so the brightness has 72×90=6480 DCT blocks, and CR and CB each have (18×2)×45= 1620 blocks, and then the four brightness blocks of CR and CB and 1 block form a macro block. For 625/50, 1620 macroblocks are formed, and then 27 macroblocks form a super block, and 60 super blocks can be obtained. 60 super blocks are divided into 12 lines, and 5 super blocks in each line are recorded on 12 tracks, which is the same as S, that is, 12 tracks are recorded in each frame.

In a word, DVCPRO50 is not only compatible with DVCPRO25 in playback and recording, but also fully compatible in data stream, realizing seamless switching. This enables DVCPRO to switch and create a series of programs in servers that store 25Mb/s and 50Mb/s files.

The "format war" has a long history, and there are joys and sorrows for users: the good news is that the performance and price of various machines have been chosen because of numerous format competitions; I am worried that improper choice will bring a lot of inconvenience to the work, and even cause backlog and waste. Therefore, when choosing a digital video recorder, we should choose it according to the specific situation and the use occasion, pay attention to the rationality of resource allocation, the systematicness, openness, compatibility and operating cost of the equipment, and properly consider the foresight to make the best use of it and get the maximum benefit with the least investment.

In terms of image quality, digital Betacam format is 4:2:2 sampling, 10 bit quantization, and intra-frame 2: 1 compression is the highest level in compressed digital equipment. Betacam-SX is the inter-frame compression of 10: 1, and the bit rate of 18Mbps is the lowest among several modern video recorders, thus reducing the running cost. Its image quality is slightly lower than that of digital Betacam, but it can reach B-level. The DVCAM format developed by Sony on the home DV format is 4:2:2 sampling, 8bit quantization and 5: 1 compression in the frame, and the picture quality is close to PVW level. Panasonic's DVCPRO50 is an upgraded product of DVCPRO. The intra-frame compression is 3.3: 1, and the code rate is 50Mbps. The image quality is close to DVW, the bit rate of DVCPRO25 is 25Mbps, and the image quality can reach the level of PVW. DIGITAL-S of JVC company, with 3.3: 1 compression in the frame, the picture quality at 50Mbps is close to DVW level.

In short, these digital video recorders are still in the development stage, and some technical problems need to be further improved. I believe that in the near future, this kind of digital video recorder equipment will be more and more perfect, the price will be cheaper and cheaper, and the picture quality will be higher and higher.

Specification \ format number -S Betacam-SX VCPRO

Image coding standard 4:2:2 component mode 4:2:2 component mode 4: 1: 1(4:2:2) component mode.

Image compression mode The DCT mode of intra-frame compression is based on the P@ML mode of MPEG-2.

The compression ratio is 3.3:110:15:1(3.3:1).

Code rate (about) 50mbps18mbps 25mbps (50mbps).

Audio standard 16bit/48KHz uncompressed 16bit/48KHz uncompressed 16bit/48KHz uncompressed.

4 channels 4 channels 2 channels

Magnetic tape 1/2 inch metal coated tape 1/2 inch metal tape 1/4 inch metal particle tape.

Maximum recording time 104 minutes 180 minutes (large band) 60 minutes (small band) 123 minutes.

Drum diameter/speed 62mm/75rps81.4mm/75rps21.7mm/150rps.

Track number/frame12 (625/50)10 (525/60)12/GOP (625)

10 Articles/Republican Party (525)

Article 10 (525)

The tape speed is 57.8 mm/s (625) 57.7 mm/s (525) 59.575 mm/s (625) 33.8 13 mm/s (625).

The track width is 20μ m 32μ m18μ m.

Record bearing 15 degrees 20 degrees.

Orbit inclination (degree) is 5.57 4.4438+0 9.6438+0.4438.40088888986

Compatibility of replay has S-VHS replay function, which can replay the existing Betacam.

The material DVCPRO-25.50 on the oxide belt and the metal belt is compatible.

The image quality is close to DVW level, which is equivalent to PVW at BVW level (close to DVW).

The life of drum ≥800h is three times that of PVW, about 2400 h >: 1