Home

DVD Intro

DVD Video

Guide

Features

Specs

Technical

Players

Movies

Glossary

DVD Format Definition

Video Specifications for DVD-Video

DVD-Video can accommodate both MPEG-1 and MPEG-2 compressed video content.

• MPEG-1 is equivalent to a VHS quality tape.
• MPEG-2 is higher quality and can be displayed on a 4x3 screen, 4x3 letterboxed, or 16:9 screen.
• The MPEG (Motion Picture Expert Group) standard is used for the compression of digital video and audio data.
• The MPEG standard is based on the H.261 standard.
• The MPEG standard uses non-square pixels, unlike most computer displays. This poses a special set of workflow adjustments when created graphics and menus for a disc.

MPEG-1 attributes:

* CBR means that simple and low-action scenes receive the same data rate as complex and high-action scenes. Only about 60 minutes of content can be recorded to a DVD-Video disc, single-sided, single-layer.

MPEG-2 attributes:

* VBR means that simple and low-action scenes receive less data rate and complex and high-action scenes receive more in order to display those scenes without artifacts. Simple scenes use less disc space. Complex scenes use more. Ultimately, this means more content can be recorded to the disc than with CBR encoding.

**Even though the DVD-Video disc with MPEG-2 content can display the 16:9 aspect ratio, the same pixel resolution is used as 4:3. Picture quality does not increase. Letterboxing discards even more pixel picture information for the black bars, rather than for movie display.

***Conditionally High means that in the hands of a bad compressionist, even the MPEG-2 content can contain artifacts and look bad, especially if insufficient bandwidth/data rate is allowed.

GOP...Group of Pictures

• The method for compressing video content in the MPEG standard is through a Group of Pictures.
• There are two common methods of compressing digital video for distribution.
  1. INTRA-FRAME: Often called Spatial Compression, this method removes redundancies within a single frame of video. Each frame of video is still discrete and can be accessed as a unique point in time through a timecode address or frame advance function. M-JPEG is a prime example of an Intra-Frame Codec and flavors of M-JPEG are used in many non-linear editing systems.
  2. INTER-FRAME: Often called Temporal Compression, this method removes redundancies across a group of pictures over time. MPEG, Sorenson, RealSystem and WindowMedia are Inter-Frame Codecs.
• MPEG uses both INTRA- and INTER-FRAME compression across a Group of Pictures in a motion sequence. Frames are designated as I-Frames, P-Frames, or B-Frames.
  1. I-FRAMES (Intra Frames): Often called a keyframe, an I-FRAME is a single frame in which all of its content is examined independently by the compressor, regardless of the frames preceeding or following it. The complete frame is divided into blocks, which are then transformed using discrete cosine transformation (DCT). That information is then quantified and compressed.
     • When scanning forward and backward on a DVD-Video, the I-Frames are the contents that are displayed.
     • Less sophisticated MPEG-2 compressors or real-time encoders may only use the I-Frame technique, resulting in larger file sizes (and less material on the disc).
     • A DVD-Video that was composed of nothing but single frame images would have to use the I-Frame method of compression exclusively (see the Slice of Life Digital DVD Experiment).
  2. P-FRAMES (Predicted Frames): Often referred to as delta frames, P-Frames are frames following an I-Frame whose content or color has changed from the preceeding I-Frame. Because a P-Frame only contains data that has changed, it can't be accessed and displayed on its own; it requires an I-Frame to fill in the sequence's original content.
     • Many MPEG-2 compressors use just the I-Frame and P-Frame method of compressing video.
  3. B-FRAMES (Bidirectional Predicted Frames): Also referred to as delta frames. Additional content redundancies can be removed during compression by looking at changes in frame content or color based on both previous and subsequent frames. A B-Frame cannot be accessed and displayed on its own; it requires an I-Frame and surrounding P-Frames to fill in the sequence's original content.
     • The most sophisticated MPEG-2 compressors use I, P and B Frame compression.
  4. I-P-B Frame Order: A typical arrangement of I, P and B Frames in one second of MPEG video would look like this:

     I BB P BB P BB P BB P BB followed by
     I BB P BB P BB P BB P BB

      

CHROMA & LUMINANCE SAMPLING FOR DIGITAL VIDEO: Video is a representation of the true colors of the real world. Think of video as a "sample" of those colors. Hopefully, the "sample" is good enough that we humans perceive there is little difference between the real world and the video representation of it.

The sampling is split between color attributes and brightness or luminance attributes. In actuality, the human eye (and brain) places more importance on luminance values rather than chroma values. Because of this perceptual preference, video samples of the real world should also be conducted more frequently for luminance.

The samplings in the video signal are divided into three elements; one for luminance and two for color.

Different video formats have different sampling values. The sampling rate for a video format is expressed as three digits separated by colons, e.g., 4:1:1

The more samples, the better the quality of the video and the color depth. Much of the differences are perceptual. However, a higher sampling rate does mean that a video stream containing more picture information can be manipulated more cleanly, such as when using "green screen" mattes or computer generated special effects.