Skip to content

MPEG H decoder example

Jump to bottom
ameci-iis edited this page Jun 21, 2024 · 2 revisions

About

The mpeghdec repository provides a command-line program as an example for MPEG-H decoding implementation. 

The example program:

  • takes an MP4 file, containing an MPEG-H audio track, as input and extracts the MPEG-H Audio frames from MP4 file format (using the mmtisobmff project).
  • decodes the extracted MPEG-H Audio frames to the PCM format according to the desired target speaker layout and optionally sets DRC and loudness decoding parameters.
  • stores the decoded PCM data into a WAV file.

How to build

To build the MPEG-H decoder command-line example program set mpeghdec_BUILD_BINARIES variable while configuring the project using CMake :

$ cmake -S mpeghdec -B build -DCMAKE_BUILD_TYPE=Release -Dmpeghdec_BUILD_BINARIES=ON

Read build instructions to learn how to configure and build the project.

If the build process is successfully finished, the MPEG-H decoder command-line tool mpeghDecoder can be found:

  • in the build/bin folder
  • in the build/bin/Release folder on Windows

Usage

$ build/bin/mpeghDecoder [options] -if <path to input file> -of <path to output file>
-if <path to input file>
relative or absolute path to an input MP4 file containing an MPEG-H audio track (in MHM1 or MHA1 format) to decode.
-of <path to output file>
relative or absolute path to an output WAV file which will contain the decoded PCM data.

Options are:

-tl <idx>
<idx> : CICP index of the desired target layout (default: 6).
-rl <level>
<level> : DRC target loudness (reference level) in steps of -0.25 LU using values [40..127].
-dse <effect type>
<effect type> : MPEG-D DRC effect type request, e.g.: 0 = None (default), 6 = General.
-db <scale factor>
<scale factor> : DRC boost scale factor, where 0 is no boost and 127 is max boost.
-dc <scale factor>
<scale factor> : DRC attenuation scale factor, where 0 is no compression and 127 is max compression.
-dam <album mode>
<album mode> : MPEG-D DRC album mode, 0: disabled (default), 1: enabled.
-y <sample number>
<sample number> : start decoding at ISOBMFF/MP4 sample number.
NOTE: The decoding will start at the nearest sync sample!
-z <sample number>
<sample number> : stop decoding at the provided ISOBMFF/MP4 sample number.
-sf <sample number>
<sample number> : seek in the bitstream from the provided sample number to the ISOBMFF/MP4 sample number provided with '-st'.
NOTE: '-st' must be set!
-st <sample number>
<sample number> : seek in the bitstream from the ISOBMFF/MP4 sample number provided with '-sf' to the provided ISOBMFF/MP4 sample number.
NOTE: '-sf' must be set!
-h
show help.

Testing

To test MPEG-H decoding with the provided command-line program example or with your own implementation please use test material from the test content repository.

Additionally you can:

Example to test MPEG-H decoding for files provided in test content repository:

$ build/bin/mpeghDecoder -tl 19 -if TRI_Fileset_17_514H_D1_D2_D3_O1_24bit2160p50.audio.mp4 -of out.wav

Example above:

  • takes the TRI_Fileset_17_514H_D1_D2_D3_O1_24bit2160p50.audio.mp4 test item as input.
    • This MPEG-H test item contains:
      • 5.1+4 channel bed
      • 3 dialogue objects
      • 1 dynamic object
      • 3 presets
  • sets the output target layout to CICP 19 which maps to 7.1+4 speaker layout as defined in supported MPEG-H decoder target layout table.
  • writes the decoded PCM data to out.wav file.

Remarks

  • The default output format written is 24-bit signed int PCM WAV (library outputs 32-bit signed int).
  • Only the first MPEG-H Audio track will be processed, other MPEG-H Audio tracks will be ignored/skipped.
  • Other tracks than MPEG-H Audio track will be ignored/skipped.
Clone this wiki locally