MPEG Standards Explained

You frequently talk about video formats, can you explain the difference between MPEG-1, MPEG-2, and MPEG-4 and what their possible uses are?

There are so many different formats floating around, I frequently find myself double-checking to make sure I'm referencing the right thing. Without realizing it, many of us spend a great deal of time watching MPEG-2 format video either on DVD or delivered over cable and satellite to digital subscribers around the world because it remains the accepted compression standard for the television industry. Assuming you have the right software installed on your system, most of the subtle differences between the various video formats should be transparent. Things get tricky when you're missing a codec, which is software designed to interpret the audio or video information contained in a file.

MPEG stands for Motion Picture Experts Group, the standards body made up of many large companies involved in technology and content creation in the video industry.

MPEG-1 compression is the oldest MPEG compression standard. At its highest quality level, MPEG-1 formatted video is compressed to approximately 1.5Mbps. MPEG-1 video compression finds commonalities between frames in a movie to generate subsequent frames, reducing the amount of video information required to store an entire movie. For instance, if several frames in a row all contain the same still image of a living room, the compression tells the file to keep displaying the same image for each of those frames until it hits a frame where the image changes, eliminating the actual duplicate frames from the file to reduce the overall size (I'm oversimplifying a little bit, but that's the basic idea). MPEG-1 compression is made up of several layers of information, with MPEG-1 Layer 3 (or MP3) representing the audio portion of the file compresion. MP3 reduces the file size of audio data by eliminating high and low frequency sounds undetectable by human hearing, reducing the overall size the audio portion of a video file. MPEG-1 video is used for VCD video disks and is typically no better than VHS quality video.

MPEG-2 improves upon the MPEG-1 standard by increasing the data throughput a video file is capable of. Where MPEG-1 maxes out around 1.5Mbps, MPEG-2 is typically compressed to between 3.5Mbps and 6Mbps. MPEG-2 is the standard used by DVD and SVCD formats for encoding video, as well as the digital cable and satellite industries. Like MPEG-1, MPEG-2 is a lossy compression standard, meaning some of the originating video source is removed to make the file smaller. Because MPEG-2 video removes some of the video information from the file, it's occasionally possible to see blocky artifacts in the video file, especially in high motion scenes. This is less true for MPEG-2 video compressed in the 6Mbps range, because the quality level approaches a range where defects are imperceptible to the human eye, but as the quality level is improved the file size gets larger.

MPEG-4 treats compression differently than either MPEG-1 or MPEG-2. Because of this compression difference, MPEG-4 video output offers higher image quality at much smaller file sizes than either MPEG-1 or MPEG-2 are capable of. MPEG-4 compresses files in a range from 5Kbps to 10Mbps making it adaptable for delivering video to everything from cell phones to HD quality output. Instead of interpreting each individual frame of a video, MPEG-4 compresses images by dealing with objects in the video, meaning it efficiently reuses image information without throwing away as much image data. Two of the more commonly available uses for MPEG-4 are DivX movies, widely found on p2p networks and the audio and video data being created through Apple's QuickTime format as MP4 video and AAC audio.