How DTS Audio Integrates into DVD VOB Streams

This article explains how Digital Theater Systems (DTS) audio is integrated into the DVD Video Object (VOB) container format. It covers the underlying MPEG-2 Program Stream architecture, the packet-level identification mechanisms used to distinguish DTS from other audio codecs, how playback synchronization is maintained, and the standard compliance rules governing DTS implementation on commercial DVD-Video discs.

The MPEG-2 Program Stream Container

A DVD VOB file is an implementation of the MPEG-2 Program Stream (PS) standard (ISO/IEC 13818-1). The VOB container acts as a multiplexer, packaging separate elementary streams—MPEG-2 video, audio tracks, sub-picture overlays (subtitles), and navigation data—into a single, interleaved data stream designed for sequential playback.

Within this container, data is divided into fixed-size packs of 2,048 bytes, matching the sector size of a physical DVD. Each pack contains a pack header and one or more Packetized Elementary Stream (PES) packets containing the actual payload data.

Identification via Private Stream 1

Unlike MPEG audio, which possesses dedicated stream IDs in the MPEG-2 standard, DTS was integrated into the DVD specification as an extension. To maintain backward compatibility with standard MPEG-2 demuxers, DTS audio is encapsulated inside MPEG-2 Private Stream 1 packets, which use the stream ID 0xBD.

Because Private Stream 1 is also shared by other audio formats (such as Dolby Digital AC-3 and Linear PCM), the DVD specification uses a sub-stream ID located immediately after the standard PES packet header to differentiate formats:

When a DVD player encounters a PES packet marked 0xBD, the demultiplexer reads the subsequent sub-stream ID byte. If it falls between 0x88 and 0x8F, the data is routed to the DTS audio decoder or output bitstream buffer.

PES Packet Structure and DTS Framing

A DTS audio PES packet inside a VOB stream consists of several layers:

  1. Pack Header: Contains the System Clock Reference (SCR), which synchronizes the player's internal master clock to the stream.
  2. PES Header: Contains the stream ID (0xBD), packet length, and timing metadata, including Presentation Time Stamps (PTS).
  3. DTS Sub-stream Header: Contains the specific DTS sub-stream ID (0x88–0x8F) followed by frame-synchronization metadata.
  4. DTS Payload: Raw DTS Coherent Acoustics audio frames.

DTS frames on a DVD are typically encoded at a sample rate of 48 kHz with bit depths of 16 or 24 bits. They operate at one of two standardized bitrates: half-bitrate (754.5 kbps) or full-bitrate (1509.75 kbps), structured into 5.1 discrete surround channels.

Synchronization and Interleaving

DVD audio and video must remain synchronized across variable bitrates and potential disc-reading latency. The VOB stream organizes data into Video Object Units (VOBUs). A VOBU typically represents 0.4 to 1.0 seconds of content, beginning with an MPEG-2 Group of Pictures (GOP) anchored by an I-frame, followed by the corresponding audio and sub-picture packets for that segment.

DTS audio frames are interleaved alongside video packets within each VOBU. The decoder relies on the Presentation Time Stamps (PTS) embedded in the PES headers. The PTS dictates the exact moment an audio frame must be converted to an analog signal or passed via S/PDIF or HDMI to an external receiver relative to the master System Clock Reference (SCR).

DVD Specification Requirements

Under the DVD-Video standard set by the DVD Forum, DTS is classified as an optional audio format.

Because early hardware players often lacked built-in DTS decoders or digital passthrough capabilities, DVD specifications dictate that a disc containing a DTS track must also include at least one mandatory baseline audio format—either Dolby Digital (AC-3) or Linear PCM (LPCM) in NTSC regions, or AC-3, LPCM, or MPEG-1 Audio Layer II in PAL regions. This ensures universal playback while allowing DTS to integrate as an alternate high-bandwidth multi-channel stream.