DVCPRO is based on the original DV consumer-based video format. This appendix gives an overview of both the DV and DVCPRO digital video formats. It also compares various available video formats in the industry. This appendix consists of these sections:
DV is a ubiquitous video format developed by the HD Digital VCR Consortium, which is composed of more than 50 companies, including Sony, Matsushita, Philips, Thomson, Toshiba, Hitachi, JVC, Sanyo, Sharp, and Mitsubishi. See the “Specifications of Consumer-Use Digital VCRs Using 6.3 mm Magnetic tape” (distributed to Consortium members) for complete details of DV.
The DV initial specification was introduced on July 1, 1993, and now covers both Standard Definition (SD) and High Definition (HD). Table F-1 shows some SD specifications; Table F-2 shows audio recording method specifications.
Y: 13.5 MHz
Standard DV: 270 minutes
Digital compression method
Discrete Cosine Transform (DCT)
DV supports 4:1:1 for 525/60 and 4:2:0 for 625/50. The bit rate is kept constant and is dependent on luminance sampling structure, quantization depth, ratio of luminance to color difference samples, number of frames per second, and the compression ratio (5:1 for DV).
|Note: For sampling pattern illustrations and descriptions, see “Sampling Patterns” in Appendix C in Appendix C, “Pixel Packings and Color Spaces ”.|
Figure F-1 summarizes the compression process used in DV.
The steps performed by DV are as follows:
Pixels are assembled into appropriate blocks.
DCT is applied to each block.
Macroblocks of six DCT blocks are formed. These macroblocks are then selected at various places of the picture (hence the shuffling process) and categorized into classes of importance.
This method better preserves the fine details of the picture throughout the entire frame.
DV applies the popular compression method Discrete Cosine Transform (DCT) to each macroblock.
The assembling process combines four luminance blocks with two chrominance blocks (4:1:1 or 4:2:0) to form a macroblock. Five macroblocks are selected from five different regions in a picture to form a video segment. DCT is applied to each macroblock.
This mathematical formula transforms each block's pixel signal levels into frequency domain coefficients. The coefficients are quantized at different levels to achieve the best compression without affecting quality. The rate control process makes sure a constant bit rate is achieved by varying the quantization scale.
Variable-length coding (VLC, or Huffman encoding) is applied to shorten the data length. Deshuffling puts VLC-coded data into 77-byte packets. Highly detailed macroblocks requiring more data space might use other less-detailed macroblocks to store information.
Overall, this compression process differs from MPEG or M-JPEG compression in these ways:
DV compresses a picture using only intraframe, and each frame is compressed to a fixed size. This method makes editing easier.
Intraframe avoids all motion artifacts associated with interframe coding, a trade-off for its higher bandwidth requirement and a lower compression ratio.
On one hand, considerable effort was expended to make MPEG codec nonsymmetrical: the encoder is significantly more complex than the decoder, making distribution appropriate. On the other hand, DV is designed to be symmetrical, compact, and cost-effective. The same chipset can perform both compression and decompression.
For tape requirements, DV uses 1/4-inch wide tapes at a track pitch of 10 μm. Each track has four primary data areas:
Insert and track information (ITI): this section includes location information, track pitch, servo information, and the application ID of a track (APT). The APT indicates the data structure of that particular track.
Audio: this section contains audio information, audio auxiliary data, and an APT.
Video: compressed video data, video auxiliary data, and an APT are stored in this section.
SubCode: SMPTE/EBU time code, absolute track number, and APT are stored here.
Additional space (editing guard bands) is also included between each of the areas discussed above.
A 525/60 system has 10 tracks per frame; a 625/50 system has 12 tracks per frame. Each track corresponds to a horizontal band of the picture.
For audio, DV allows several modes:
The 48-KHz mode captures two channels of AES/EBU professional digital audio using 16-bit linear samples.
Two channels are also possible for 44.1 KHz and 32 KHz, also using 16-bit linear samples.
A four-channel mode is available for 32 KHz, using 12-bit nonlinear samples. The audio is divided into frames, which are grouped onto video tracks. For example, in 525/60, channel 1 audio is grouped on tracks 0-4, and channel 2 audio is grouped on tracks 5-9.
Developed by Panasonic, DVCPRO is a nonproprietary video format standard for the production, broadcast, and distribution of digital television video. It is also known as SMPTE D-7; see that documentation for full details of the DVCPRO video format.
DVCPRO is built on the DV format (6.35 mm), but is simpler and thus more suitable for the professional and broadcast markets. Instead of metal evaporated tape, metal particle tape is used. Track pitch is increased from 10 microns (DV) to 18 microns. Tape speed is increased from ~18.8 mm per second to ~33.8 mm per second.
As a subset of DV, DVCPRO uses the same video compression method as DV for 525/60. The video signal is sampled at 13.5 MHz for luminance (Y) and 6.75 MHz for color differences (Cb and Cr). Then the sampled video data is compressed via DCT, 8-bit quantization, and VLC.
Audio is recorded via two 16-bit audio channels at 48 KHz, synchronous with video. In addition, another cue audio track is included to allow intelligible audio during variable play speeds. A control track is also added to help minimize tape preroll and more accurate frame editing.
For 625/50, DVCPRO uses a 4:1:1 sampling structure instead of 4:2:0 as used in DV. The DIVO-DVC option board supports 4:1:1 sampling for both 525/60 and 625/50, and supports 4:2:0 625/50 for DV video format. VITC carries additional information and is recorded in the subcode and video auxiliary regions, like DV.
Despite its tape-oriented technology, DVCPRO, like DV, is an open format designed for use with nonlinear editors and server-based systems. As with DV, DVCPRO has a 4X transfer speed record-and-play capability.
DVCPRO for 525/60 is backward-compatible with DVC; you can play 525/60 DVC tape on a DVCPRO tape deck.
At 25 Mbps and using a 4:1:1 sampling structure, which yields a 5:1 compression ratio, the DIVO-DVC board can accomplish a wide range of work flexibly and cost-effectively.
For information on DVCPRO, see
Proposed SMPTE Standard 306M for Television Digital Recording — 6.35-mm Type D-7 Component Format — Video Compression at 25 Mb/s — 525/60 and 625/50
SMPTE's web site is http://www.smpte.org/.
Table F-3 compares DVT options.
DV (Sony, JVC, Panasonic)
Bit rate (Mbps)
525/60 frame size
720 x 480
720 x 480
720 x 480
720 x 480
720 x 487.5
625/50 frame size
720 x 576
720 x 576
720 x 576
720 x 576
720 x 583.5
Not on all models
Extended play (12.5 Mpbs)
Audio frequency (KHz)
32, 44.1, 48
Locked and unlocked
2 and 4
Main, timecode, take,
Track pitch (μm)
18 (plays 10+15)
18 (plays 10+15)
Tape speed (min/sec)