TITLE
    DVD-Video: Overview
Article ID:
Created:
Modified:
24485
6/16/98
11/19/99

TOPIC

    This article provides information and a general overview of Digital Versatile Disc (DVD) Video technology, particularly as it relates to Apple. Included in the article are the following sections:

    • Acronyms Explained
    • DVD-Video Compared
    • DVD-Video Quality
    • Video
    • Audio Options
    • Interactive Video
    • Subtitles

    This article deals only with DVD-Video. For an overview of DVD technology, see the following Tech Info Library article:

    Article 24451: " DVD Technology: Overview "


DISCUSSION

    Acronyms Explained

    There are 4 main areas that digital video can be divided into:
    • physical, shiny disc distribution (similar to distribution on VHS video tape)
    • collection of original footage (similar to Hi8/VHS/S-VHS camcorders)
    • over-the-air broadcast television
    • cable/satellite broadcast television

    DVD-Video is the physical, disc based distribution channel. All of the DVD-xxx variations are based on the same shiny disc. However, the format used by the differing versions can create compatibility issues.

    DV (and its variations) is the most common collection of original digital footage.

    DTV is used fairly often to describe the overall digital over-the-air broadcast method. With specific formats noted as:
    • HDTV - High Definition TV, a 16 x 9 ratio image at greater than a 1000 lines of horizontal resolution. There are two resolutions defined by the United States Federal Communications Commission (FCC). Both can be presented in progressive (typical of computer monitors) or interlaced (typical of NTSC televisions).
    • SD TV - Standard Definition TV, a 4 x 3 ratio image at approximately 720 lines of horizontal resolution. Also available in interlaced and progressive. Takes only a small part of the HDTV bandwidth allowing as much as four separate programs to be sent down the same HDTV channel.

    DSS is the generic term for the small dish satellite systems used to deliver multitudes of channels to the home. Cable TV operators deliver a similar digital signal via cable to their subscribers.

    The one thing to remember about all of these different digital video formats is that hardly any of them are directly compatible with any other format. To get from one format to another will require some form of transcoding.

    DVD-Video Compared

    Let's look at DVD-Video and how it relates to the other formats.

    DVD-Video uses a MPEG-2 based compression technique for the video image. MPEG-2 is also used with DSS, digital cable, HDTV and SDTV.

    However, DVD-Video employs a variable bit-rate compression scheme. Variable bit-rate compression allows the compressionist (the person making the decisions about how much compression to use) to apply a higher compression ratio to a low detail, low action scene; while allowing a high detail, high action scene to be compressed at a lower ratio. You want to do this because the higher detail/action scene would show artifacts more as the compression ratio is increased. A lower detail/action scene doesn't change as often and thus can be compressed at a higher ratio without artifacts appearing. A disadvantage of variable bit-rate compression is that it is not a real time operation. At a minimum, two passes are required to achieve any advantage.

    Another element of the MPEG-2 stream used with DVD-Video is the additional information encoded in the video stream. Navigation information, language information, subtitle information, parental control information, Dolby Digital audio, and so forth are all contained in the DVD-Video MPEG-2 video stream.

    DSS and digital cable use a constant bit-rate compression. Since these distribution channels are dealing with real time events, there is no time for the two pass method used in variable bit-rate compression. Therefore, the image is always compressed the same amount whether it is a low detail/action scene or a high detail/action scene.

    DSS and digital cable do not provide navigation, subtitle, parental control and alternate language abilities found in DVD-Video. These delivery channels would not be able to understand these data elements.

    The MPEG-2 stream coming from over-the-air broadcast channels contains a different structure than either of the other MPEG-2 based data streams. Resolution is one of the first things that differs from the other two.

    DV tapes do not use an actual MPEG-2 compression. While the basic theory used for compression is the same for DV and MPEG-2, the DV compression is not MPEG-2.

    DVD-Video Quality

    DVD-Video uses MPEG2 compression. Any time an image is compressed there is the opportunity for image degradation. A good compressionist can create such a good image that 90% of the viewers will not be able to see any compression artifacts in the movie. However, a less than average compressionist can totally ruin a movie by not paying attention to detail.

    There is also the issue of viewing a movie on a computer screen versus a television screen. A computer screen is capable of providing a much higher quality image than does a NTSC television screen. There are titles which may be perfectly acceptable on an NTSC television that will look very poor on a computer screen. This difference will most likely be traced back to a less than ideal job of compressing the original material--but the viewer may likely believe it to be the fault of the computer.

    Video

    General Video Information

    There is DVD-ROM. There is MPEG-2. There is DVD-Video.
    • DVD-ROM is the big picture.
    • DVD-Video is one way to use DVD-ROM.
    • DVD-Video is one way to use MPEG-2

    Thus, DVD-Video is a specific use of MPEG-2 video on DVD-ROM in a specific manner.

    A DVD-Video disc contains a stream (a track) of MPEG-2 (Main Profile@Main Level, also known as MP@ML) video compressed and encoded in either a Constant Bit Rate (CBR) or in a Variable Bit Rate (VBR). The DVD-Video specification also allows the lower resolution MPEG-1 CBR and VBR video.

    An interesting aspect of DVD-Video is the ability to have the 24 frames per second (fps) film encoded at 24 frames a second. With other forms of home video, the 24 fps film standard has to be converted to the NTSC spec of 29.97 fps via a process called 3:2 pulldown. (NTSC consists of 30 frames with 2 fields each.) What 3:2 pulldown does is double one field creating 3 fields for one of 24 film frames - and does this often enough to turn the 24 film frames into 30 video frames (well, actually 29.97 video frames - but that's another math story.) In order to display these 24 frames on an average television, the DVD-Video playback device does an 'on-the-fly' 3:2 pulldown.

    So, why have 24 frames encoded and do on the fly conversion? With current computer displays and with future digital TVs it will be possible to display the actual 24 fps in which the film was shot. As a result, a more correct feel will be imparted to the playback of movies shot on film and transferred to video.

    Video Image Information

    The number of pixels in a DVD-Video frame are typically 720 x 480 of NTSC source material. The traditional method of measuring of NTSC devices has been "lines of horizontal resolution". Using this method:
    • VHS source material is rated at approximately 230 lines for standard 4:3 screens (approximately 172 for 16:9 screens)
    • laserdisc is approximately 425 for 4:3 (approximately 318 for 16:9)
    • DVD-Video is approximately 540 for 4:3 (approximately 405 for 16:9).

    In practical terms, DVD-Video will most likely be rated at approximately 500 lines of horizontal resolution.

    A 720 x 480 pixel count should not be confused with a "lines of horizontal resolution" measurement. The numbers may be different. Here's why: "lines of horizontal resolution" (LoHR) comes from the ability of a device to resolve an industry-standard reference image. Part of the image has groups of lines printed at various distances from one another. LoHR is determined at the point where these printed lines are resolvable into individual lines. So while it is possible to have 720 horizontal pixels, if you cannot visually distinguish each line, you don't have 720 LoHR.

    MPEG uses temporal compression. Temporal compression uses a key frame to record everything in the frame. This complete frame is an I-frame. During temporal compression, subsequent frames consist of only the differences from the I-frame. So a still image can be an I-frame. The still frame can be displayed for a specified amount of time; or it can be displayed indefinitely--generally with a "continue" button on the screen. Still frames can play sound. Menus are generally still frames.

    Subpictures are not still frames. Subpictures are much more restrictive. Subpictures do not have the color range or detail of MPEG-2 video. They are used as overlays on the MPEG-2 video stream for subtitles, closed-captioning; karaoke, menus, and very simple animation. A portion of the DVD-Video interactive command set allows a title developer to perform effects such as scroll, move, color, highlight and fade subpictures. There are 32 subpicture streams available within the DVD-Video spec. Subpictures can contain 4 colors from a palette of 16 colors and four contrast values out of 16 levels from transparent to opaque.

    Audio Options

    A DVD-Video disc can have up to 8 audio tracks (tracks are sometimes referenced as streams). Each of these tracks can be one of three formats:
    • Linear PCM -- from 1 to 8 channels
    • Dolby Digital (also known as AC-3) -- from 1 to 5.1 channels
    • MPEG-2 audio -- from 1 to 5.1 or 7.1 channels

    Note: tracks (or streams) are composed of channels

    Additionally, the Digital Theater Sound (DTS) format is an optional format. DTS is another 5.1 channel format which was developed to improve the audio experience in movie theaters. There is reserved within the DVD standard an audio stream format for DTS. DTS format DVD-Video discs first appeared in late spring of 1998.

    Linear PCM
    Linear PCM is generally the default audio track. It will be sampled at a rate of 48kHz or 96kHz. Audio CDs are sampled at 44.1kHz. A sample size of 16, 20 or 24 bits is allowed. Audio CDs use a sample size of 16 bits.

    According to the DVD-Video specification it may contain up to 8 channels. There is a maximum bit rate of 6.144 Mbps. Linear PCM is uncompressed audio which means the choice for the number of channels may limit the sample rate or sample size used. Typically up to 5 channels should be able to use the highest data rates and sizes.

    While DVD-Video playback devices are required to support any of the above variations, the device could subsample from 96kHz down to 48kHz while not using the full 20 or 24 bits. Still the lowest output from the PCM track would be 48kHz @ 16 bits--better than Audio CDs using 44.1kHz @ 16bits.

    Dolby Digital
    If PCM is not the default audio track, then the default will most likely be Dolby Digital. The Dolby Digital format is compressed data using the AC-3 technique. The source material is normally taken from PCM recordings at 48kHz at up to 24 bits.

    Typically you will find Dolby Digital to be a 5.1 channel format. The "5" breaks down to left, center, right, surround left and surround right. The ".1" channel is the low frequency effects (LFE) channel--the channel that would be directed to a subwoofer. However, there are a number of valid channel formats in DVD-Video Dolby Digital:
    • single mono
    • dual mono channels
    • standard stereo
    • left/center/right
    • stereo w/single surround
    • left/center/right/single surround
    • stereo w/dual surrounds
    • left/center/right/dual surrounds

    With each of the above the ".1" (LFE) channel is optional. Again, the most common channel format you will find for Dolby Digital is the 5.1 arrangement (the last one in the list plus the LFE). You will also find that most regional 1 DVD-Video discs contain a Dolby Digital track. Although as of the late spring 1998 a number of DTS DVD-Video discs are starting to appear.

    MPEG-2 Audio
    The definition of MPEG-2 audio allows it to be a multi-channel format, although some playback devices provide only basic stereo left/right delivery. Somewhat like Dolby Digital there are a number of channel configurations available with MPEG-2 audio, including:
    • single mono
    • stereo
    • stereo w/single surround
    • stereo w/dual surround
    • left/center/right
    • left/center/right/single surround
    • left/center/right/dual surround
    • left/left-center/center/right-center/right/dual surrounds

    As with Dolby Digital, the LFE channel is optional on all of the above channel configurations. When the LFE is add to the last one in the above list, it is often notated as "7.1".

    Both MPEG-2 and MPEG-1 formats are supported, although MPEG-1 Level III (also known as MP3) is not supported.

    DTS (Digital Theater Sound)
    DTS is an optional DVD-Video multi-channel audio format (typically 5.1). The DTS format can provide a data rate nearly 4 times as much as Dolby Digital - meaning less possibility for audio artifacts from compression. It also means that other options in the overall data stream may have to be dropped (like less language tracks or no alternate angle tracks, and so forth).

    DTS has several possible channel configurations, including:
    • single mono
    • stereo
    • left/center/right
    • stereo w/single surround
    • stereo w/dual surrounds
    • left/center/right/dual surrounds

    All have the option to include the LFE channel.

    Nearly all playback devices prior to the spring of 1998 do not support DTS. After that time only specific playback devices support DTS. The devices which do support DTS are normally marked with an official "DTS Digital Out" logo.

    SDDS (Sony Dynamic Digital Sound)
    Another option for multi-channel audio is SDDS in 5.1 or 7.1 configurations. However, Sony has not announced any plans to provide support for SDDS in the DVD-Video arena.

    General Audio Info
    The DVD-Video specification calls for all NTSC discs to contain either PCM audio or Dolby Digital audio on at least one track. Additional audio tracks may be any of the other formats.

    DVD-Video playback devices are expected to have a resident Dolby Digital 2-channel decoder (chip, software, etc.) which mixes a 5.1 Dolby Digital down to Dolby Surround stereo so that a standard stereo playback can be achieved.

    All of the standard or optional audio formats provide support for karaoke mode. This consists of 2 channel stereo (L/R) with an optional melody channel and two optional vocal channels.

    Interactive Video

    DVD-Video allows a limited amount of interaction with the disc contents. This interaction is far less than people would be accustomed to with CD-ROM on computers, but it is more than people would expect from a video tape or laser disc.

    The primary method of interaction is through the use of menus. Most DVD-Video discs provide a menu for selecting a variety of elements from the disc, such as playing the movie, playing the background video, hidden easter eggs, and even simple games. Some discs provide alternate story lines which can be selected via the menu system.

    The menus consist of a backdrop image with up to 36 buttons which are selected by using a remote control with left-right-up-down arrows (or a mouse/trackpad on a Macintosh with DVD-Video). Once a button is selected, a select button on the remote control will activate the selected activity.

    Supporting the menus is a very basic command instruction set (not available to the person playing the movie, but available to the disc developer). There are 24 system registers which contain information such as language code, audio and subpicture settings, parental level, and so forth. There are also 16 general purpose registers for command instructions.

    Using the commands provide a method to branch to other commands. Commands are used to control the player's settings, move to different sections of the disc and control such elements as the selection of which audio streams, video streams, camera angles, subpicture, and so forth.

    The content of a DVD-Video disc is organized by 'title' (movie) and 'sections of titles' (chapters). The chapters can be organized through the use of a program chain or 'PGC'. For example, a PGC can be created which will be able to play back only the chapters that would cause an 'R' rated movie to become a 'G' rated movie if the parental control level was set in such a way. Or the PGC could create an alternate story line. PGCs can be used to construct other sequences of chapters as well.

    Following a PGC is seamless. There are no breaks in the flow of the movie - like there is when two sided laser discs are flipped. Different PGCs are different paths through the same material.

    Subtitles

    Subtitles are not generated by the Macintosh computer--they are placed on the disc by the DVD pressing operation. The subtitles are therefore a part of the MPEG2 data stream that can either be "turned on" or "turned off". If they are "turned on" they are composited over the top of the video image. If "off" they are ignored by the process and simply are not displayed. On most DVD-video discs there will be more than one language available for subtitles. Each of these different language subtitles are a part of the MPEG2 data stream. If a specific language is selected to be displayed, that part of the data stream is composited over the video image.

    This means that the Mac OS does not have access to the subtitles such that different fonts can be used. The font is selected and recorded in a subpicture stream prior to the pressing of the DVD-video disc.

Document Information
Product Area: Computers
Category: General Topics
Sub Category: General Topics

Copyright © 2000 Apple Computer, Inc. All rights reserved.