Audio (and video for that matter) are stored as "streams" of data. Usually the audio streams are combined with the video stream. When they are separated out, they are called "elemental streams".
DVD-Video discs can carry up to 8 streams of audio using a number of non-compressed and compressed audio coding formats. Each audio stream can contain from 2 (stereo) to 6 (surround sound) channels depending on the source material available. Multi-channel audio will be down-mixed to stereo in players where there is no surround sound decoder. The audio coding formats specified differ slightly between PAL and NTSC.
So why would anyone need 8 streams ?? Well, multiple streams allow multiple languages on a single disc. So you could offer a stereo soundtrack with English, Spanish. French, and Chinese.
Audio Coding Formats
Normally, when used with video, a compressed format will be used. Audio coding formats available for DVD Video include MPEG-2, AC3 (Dolby Digital), LPCM and DTS.
Most movies use either MPEG2 or AC3. LPCM takes a lot of bandwidth (up to 6 Mbps !!), and since many older players do not support DTS, the movie houses do not use it. However, DTS is now just beginning to come out with some new audio DVD titles.
5.1 Surround Sound
This is the new rage . . . surround sound !! The added feeling of realism is incredible. You can hear someone sneaking up from behind. You can hear a missile being fired from afar, and listen as it whizzes over your head and explodes a mile behind you and slightly to the right. Actually, the newest audio systems support 6.1 and even 7.1 - but DVD movies max out at 5.1, which is plenty.
MPEG2 and AC3 both support 5.1, which uses 6 speakers (front left, front right, center, rear left, rear right, and subwoofer). A separate channel is created for the subwoofer called the LFE (Low Frequency Effects). The rear speakers are usually referred to as "surround left" and "surround right". The 5.1 standard recommends the following speaker placement:
The sound is not equally distributed. Instead, they have modeled the strengths of the speakers after the way that we hear. Our ears pick up sound in front much easier than sound in the rear. Here is the typical sound distribution of a DVD movie:
MPEG-2 (Motion Pictures Experts Group)
The MPEG video encoding formats include audio encoding, which uses lossy compression.
AC3 (Audio Code 3) = Dolby Digital 5.1
AC3 provides up to 5.1-channel surround sound. The coding format is lossy so some of the original audio quality will be lost. Bit rates from 64kb/s (mono) to 448kb/s are available. Full 5.1 channel surround sound requires at least 384kb/s, but Dolby recommends using the maximum 448kb/s. Stereo audio is normally encoded at 192kb/s.
Recently, the vendors of DVD authoring utilities have begun to add in AC3 encoding support, at a steep additional price. The attraction there, is that it offers much more intensive audio compression than MPEG-2, which leaves more space on the disc for video.
LPCM (Linear Pulse Code Modulation)
Linear PCM (LPCM) offers an alternative uncompressed audio format that is similar to CD audio, but with higher sampling frequencies and quantisations. LPCM offers up to 8 channels of 48kHz or 96kHz sampling frequency and 16, 20 or 24 bits per sample but not all at the same time. These values compare with 44.1kHz and 16 bits as used for CD audio. The maximum bit rate is 6.144 Mb/s, which is much higher than AC3 or MPEG-2 coding. LPCM offers high quality (similar to DVD-Audio) but its high data rate leaves little bandwidth for video.
DTS (Digital Theater Systems)
DTS Digital Surround is an optional 5.1 channel audio format that has become quite popular for audio DVD. DTS uses lossy compression with a sampling frequency of 48 kHz at up to 20 bits per sample. The data rate can range from 64 kbps to 1.536 Mb/s, with typical rates of 768 and 1536.
DTS, for Digital Theater System) 5.1, was introduced by Steven Spielberg with
the release of
Multi-language streams Audio
At least three surround sound streams are possible while maintaining sufficient data for the video if either Dolby Digital or MPEG-2 encoding are used. Three examples for implementing multiple languages are shown in the table with the corresponding bit rates needed.
|1 surround sound channel||
448 + (3 * 128)
|3 mono (centre) speech channels|
|1 surround sound channel||448 + (3 * 256)
= 1,216 kb/s
|3 stereo speech channels|
|3 surround sound channels||3 * 448
= 1,344 kb/s
In each case the data rate for surround sound Dolby Digital is the