Closed
Bug 1653259
Opened 5 years ago
Closed 5 years ago
Down-mix multichannel input that are more than stereo to mono when applying input processing
Categories
(Core :: WebRTC: Audio/Video, defect)
Core
WebRTC: Audio/Video
Tracking
()
RESOLVED
FIXED
mozilla80
Tracking | Status | |
---|---|---|
firefox80 | --- | fixed |
People
(Reporter: padenot, Assigned: padenot)
Details
Attachments
(1 file)
This is a hack that will work in the majority of the cases. The problem comes from the fact that if you have a mono mic plugged in the first jack socket of a professional sound card, that has lots of input channels, then we're passing in all those input channels to the echo canceler. In particular, the first channel will have sound, and the second one will be silence, and this causes unwanted (loud) noise.
The case where it won't work optimally is for professional sound cards that are stereo. The right way is to surface audio input device channel mapping in cubeb, seeing that those are discrete channels, and only using the first one (and even then it's not amazing because maybe the mic is plugged in the second channel).
Assignee | ||
Comment 1•5 years ago
|
||
Updated•5 years ago
|
Assignee: nobody → padenot
Status: NEW → ASSIGNED
Pushed by padenot@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/62abf963acfc
Downmix multichannel input to mono when applying input processing, leave stereo input stereo. r=achronop
Comment 3•5 years ago
|
||
bugherder |
Status: ASSIGNED → RESOLVED
Closed: 5 years ago
status-firefox80:
--- → fixed
Resolution: --- → FIXED
Target Milestone: --- → mozilla80
You need to log in
before you can comment on or make changes to this bug.
Description
•