Karl Tomlinson c714e843ba Bug 1605894 support AUDIO_FORMAT_S16 chunks and Chunk::mVolume in AudioSegment::Mix() r=padenot
AUDIO_FORMAT_S16 is still used by MediaPipeline and fake microphones.

The shadowing of offsetSamples, which would lead to incorrect offsets when
down-mixing multiple chunks, is removed.

The code now checks for down-mixing after up-mixing, as intended by the
up-mixing design, even though this is not necessary for the current up-mixing
tables since
https://hg.mozilla.org/mozilla-central/rev/7ed8524e54f5c3d740780d52cc73510ae6e80337#l1.18

Depends on D193522

Differential Revision: https://phabricator.services.mozilla.com/D193523
2023-11-15 09:27:53 +00:00
..