MPEG H transport formats

Overview

MPEG-H audio bitstreams can be formatted as either raw Access Units or as a self-contained MPEG-H Audio Stream (MHAS). These transport formats are usually referred to by their ISOBMFF sample type 4CCs (FourCC) of mha1 and mhm1 respectively ¹.

Raw Access Units

The use of Access Units (AU) without additional framing information is the traditional way of storing encoded audio data in ISOBMFF (mp4) files. They are not self-contained and require meta-information from the mp4-file header to be accessible. They come with almost no overhead and can't be used outside mp4 files, e.g. MPEG-2 TS or HDMI. This format is still supported by the decoder, but due to its limitations, the de-facto standard for MPEG-H Audio is MHAS.

MPEG-H Audio Stream (MHAS)

MHAS is a packetized transport format for MPEG-H Audio, i.e. individual packets can be addressed inside a bitstream without the use of additional meta-data. This provides the framing mechanism, which is required for MPEG-2 TS transport, where LATM/LOAS was used in the past. MHAS can also be used to carry meta-information in addition to the encoded audio data, which is necessary to use MPEG-H Audio over HDMI and other low-level links - the stream is self-contained. MHAS is the preferred transport format and should be used under all circumstances, if possible.

1: ISO defines the additional sample type mha2 and mhm2 for multi-stream use-cases, but that's beyond the scope of this document.↩