bbc/gst-audiodescription

Name: gst-audiodescription

Owner: BBC

Description: Barely functional gstreamer elements for decoding WHP198 audio-description data

Created: 2016-10-05 11:55:32.0

Updated: 2017-12-05 17:41:45.0

Pushed: 2017-01-28 22:03:01.0

Homepage:

Size: 42

Language: C

GitHub Committers

User	Most Recent Commit	# Commits

Other Committers

User	Email	Most Recent Commit	# Commits

README

Gstreamer elements implementing some suport for rendering Audio Description.

The elements are,

• whp198dec - extracts AD_descriptor structures from an audio waveform, encoded per BBC R&D whitepaper WHP 198
• adcontrol - consumes buffers of AD_descriptor structures and uses these to control an internal 'volume' element; used to implement the 'fading' of the audio of the main presentation as required for the audio description content to be heard clearly

               +-------------+
WHP 198 signal |             |
+-------------->  whp198dec  |
               |             |
               +-------------+
                      |
                      | AD_descriptors
                      |
               +------v------+
Main audio     |             |  Main audio (with fade)
+-------------->  adcontrol  +--------------->
               |             |
               +-------------+
                                Description audio
+-------------------------------------------->

Limitations

• Volume changes are not explicitly queued to the match audio stream, which might cause problems for some pipeline structures (untested)
• The whp198dec element hs not been generalized to support multiple sample formats and bit rates - use other Gstreamer elements to convert as required
• Ignores 'pan' information (I have no example content using the panning feature)

Example pipeline

Given a test.wav contains description in the left stereo channel, and the WHP 198 control data in the right channel, this pipeline plays the description track using a noise test signal for the main audio, as a basic demo of the control over the main audio's volume level.

gst-launch-1.0 \
  filesrc location=test.wav \
    ! wavparse \
    ! deinterleave name=d \
  d.src_1 \
    ! queue max-size-time=100000000 \
    ! audioconvert \
    ! audio/x-raw,format=S16LE,rate=48000,channels=1 \
    ! whp198dec \
    ! ad. \
  d.src_0 \
    ! queue max-size-time=100000000 \
    ! audioconvert \
    ! audio/x-raw,format=S16LE,rate=48000,channels=1 \
    ! mix. \
  audiotestsrc wave=red-noise volume=0.3 \
    ! audio/x-raw,format=S16LE,rate=48000,channels=1 \
    ! adcontrol name=ad \
    ! mix. \
  audiomixer name=mix \
    ! autoaudiosink