Skip to content

Resolve "Split rendering in VoIP mode - support for mixed formats"

Reason why this change is needed

For OSBA and OMASA bitstreams the audio output from decoding in VoIP mode was not BE between BINAURAL_SPLIT_PCM and BINAURAL output configs. The corresponding test cases in tests/split_rendering/test_voip_be_splitrend_vs_binaural.py were originally commented out.

Description of the change

This MR enables tests for decoding mixed-format bitstreams (OSBA, OMASA) to split rendering output configurations with VoIP mode enabled in the decoder.

Initially I expected that more changes are required on the C side, as some tests that were passing for single-format bitstreams failed for mixed formats. Having investigated further, it turns out that the tests were failing due to some assumptions not being true and were not a sign of an incorrect C implementation (see a more in-depth explanation in the comment below).

Also, note that SR + VoIP tests with mixed formats were already accidentally merged in !2118 (merged) (in file tests/test_be_for_jbm_neutral_dly_profile.py) and they were already succeeding there.

Affected operating points

  • N/A

Related to #1343 (closed)

Edited by sagnowski

Merge request reports

Loading