AEC3 sometimes fails to cancel echoes for highly saturated echoes |
|||||||
Issue descriptionWhen the echoes are so strong that they are saturated in the microphone signal, AEC3 sometimes fails to remove them.
,
Oct 25 2017
The following revision refers to this bug: https://webrtc.googlesource.com/src.git/+/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7 commit 7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7 Author: Per Åhgren <peah@webrtc.org> Date: Wed Oct 25 01:36:59 2017 Balancing the transparency in AEC3 between saturating and low echo paths This CL balances the NLP tradeoff in AEC3 to properly handle the cases when the echo path is so strong that it saturates the echo and when it is so weak that the echo is very low compared to nearend. Bug: webrtc:8411 , webrtc:8412 , chromium:775653 Change-Id: I5aff74dfadd51cac1ce71b1cb935d68a5be6918d Reviewed-on: https://webrtc-review.googlesource.com/14120 Commit-Queue: Per Åhgren <peah@webrtc.org> Reviewed-by: Per Åhgren <peah@webrtc.org> Reviewed-by: Gustaf Ullberg <gustaf@webrtc.org> Cr-Commit-Position: refs/heads/master@{#20418} [modify] https://crrev.com/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7/modules/audio_processing/aec3/aec_state.cc [modify] https://crrev.com/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7/modules/audio_processing/aec3/aec_state.h [modify] https://crrev.com/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7/modules/audio_processing/aec3/echo_remover.cc [modify] https://crrev.com/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7/modules/audio_processing/aec3/matched_filter.cc [modify] https://crrev.com/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7/modules/audio_processing/aec3/residual_echo_estimator.cc [modify] https://crrev.com/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7/modules/audio_processing/aec3/subtractor.cc [modify] https://crrev.com/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7/modules/audio_processing/aec3/subtractor.h [modify] https://crrev.com/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7/modules/audio_processing/aec3/suppression_gain.cc [modify] https://crrev.com/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7/modules/audio_processing/aec3/suppression_gain.h [modify] https://crrev.com/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7/modules/audio_processing/aec3/suppression_gain_unittest.cc [modify] https://crrev.com/7ddd46386a24b4d2f3064c3fdca0e124d8bb38b7/modules/audio_processing/include/audio_processing.h
,
Oct 25 2017
Triaging bugs... setting this to assigned since, well, it looks like it's assigned already. :) Flip it to Started or whatever other state best reflects the current status, peah!
,
Oct 25 2017
,
Oct 26 2017
,
Oct 26 2017
This bug requires manual review: M63 has already been promoted to the beta branch, so this requires manual review Please contact the milestone owner if you have questions. Owners: cmasso@(Android), cmasso@(iOS), gkihumba@(ChromeOS), govind@(Desktop) For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Oct 26 2017
M63 is already promoted to Beta and we're only taking critical merges in. This doesn't seem like M63 regression, can this wait until M64? Please provide merge justification if is critical to merge to M63. Thank you.
,
Oct 30 2017
The merge was important, but the stats still looks good. Therefore, considering the that M63 is already in beta I think it is safer to wait with that until M64. Thanks for looking into this!
,
Oct 30 2017
Thank you peah@. Rejecting merge to M63 based on comment #8.
,
Nov 20 2017
Per, is this fixed? Please set a milestone also.
,
Dec 12 2017
|
|||||||
►
Sign in to add a comment |
|||||||
Comment 1 by sczs@chromium.org
, Oct 18 2017