Closed
Bug 1474470
Opened 6 years ago
Closed 6 years ago
support convolution of stereo input with a mono impulse response buffer
Categories
(Core :: Web Audio, defect, P3)
Core
Web Audio
Tracking
()
RESOLVED
FIXED
mozilla63
Tracking | Status | |
---|---|---|
firefox63 | --- | fixed |
People
(Reporter: karlt, Assigned: karlt)
References
Details
Attachments
(2 files)
https://webaudio.github.io/web-audio-api/#Convolution-channel-configurations
lists "Stereo with mono response" as one of the configurations that must be supported.
Detected by
webaudio/the-audio-api/the-convolvernode-interface/convolver-response-1-chan.html:
[convolver-response-1-chan.html]
[X 2: Channel 0 expected to be equal to the array [0,0,0.9458408951759338,0.8448333740234375,0.8210252523422241,0.8620985746383667,0.8430315852165222,0.855602502822876,0.7933436632156372,0.9865825176239014,0.3972480297088623,-0.7786127924919128,-0.9223549962043762,-0.7896472215652466,-0.8727429509162903,-0.8325281143188477...\] but differs in 1276 places:\
n\tIndex\tActual\t\t\tExpected\n\t[2\]\t0.0000000000000000e+0\t9.4584089517593384e-1\n\t[3\]\t0.0000000000000000e+0\t8.4483337402343750e-1\n\t[4\]\t0.0000000000000000e+0\t8.2102525234222412e-1\n\t[5\]\t0.0000000000000000e+0\t8.6209857463836670e-1\n\t...and 1272 more errors.]
expected: FAIL
Chrome does render stereo audio with a single response since
https://codereview.chromium.org/2732523003
but the handling of channel count changes is not nice, as
indicated in the testcase.
Assignee | ||
Updated•6 years ago
|
Comment hidden (mozreview-request) |
Comment 2•6 years ago
|
||
mozreview-review |
Comment on attachment 8992564 [details]
bug 1474470 support convolution of stereo input with a mono impulse response buffer
https://reviewboard.mozilla.org/r/257422/#review264700
Attachment #8992564 -
Flags: review?(padenot) → review+
Assignee | ||
Comment 3•6 years ago
|
||
Comment hidden (mozreview-request) |
Assignee | ||
Comment 5•6 years ago
|
||
mozreview-review |
Comment on attachment 8992564 [details]
bug 1474470 support convolution of stereo input with a mono impulse response buffer
https://reviewboard.mozilla.org/r/257422/#review269556
::: testing/web-platform/meta/webaudio/the-audio-api/the-convolvernode-interface/convolver-response-1-chan.html.ini:8
(Diff revision 2)
> expected: FAIL
>
> [< [1-channel input\] 1 out of 2 assertions were failed.]
> expected: FAIL
>
> - [X 2: Channel 0 expected to be equal to the array [0,0,0.9458408951759338,0.8448333740234375,0.8210252523422241,0.8620985746383667,0.8430315852165222,0.855602502822876,0.7933436632156372,0.9865825176239014,0.3972480297088623,-0.7786127924919128,-0.9223549962043762,-0.7896472215652466,-0.8727429509162903,-0.8325281143188477...\] but differs in 1276 places:\n\tIndex\tActual\t\t\tExpected\n\t[2\]\t0.0000000000000000e+0\t9.4584089517593384e-1\n\t[3\]\t0.0000000000000000e+0\t8.4483337402343750e-1\n\t[4\]\t0.0000000000000000e+0\t8.2102525234222412e-1\n\t[5\]\t0.0000000000000000e+0\t8.6209857463836670e-1\n\t...and 1272 more errors.]
> + [X 2: Channel 0 expected to be equal to the array [0,0,0.9458408951759338,0.8448333740234375,0.8210252523422241,0.8620985746383667,0.8430315852165222,0.855602502822876,0.7933436632156372,0.9865825176239014,0.3972480297088623,-0.7786127924919128,-0.9223549962043762,-0.7896472215652466,-0.8727429509162903,-0.8325281143188477...\] but differs in 966 places:\n\tIndex\tActual\t\t\tExpected\n\t[0\]\t2.9802322387695313e-8\t0.0000000000000000e+0\n\t[1\]\t-7.4505805969238281e-8\t0.0000000000000000e+0\n\t[2\]\t9.4584077596664429e-1\t9.4584089517593384e-1\n\t[3\]\t8.4483331441879272e-1\t8.4483337402343750e-1\n\t...and 962 more errors.]
These won't pass until bug 1475158 in fixed, but at least the differences are now small.
Pushed by ktomlinson@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/483ff2d2548a
support convolution of stereo input with a mono impulse response buffer r=padenot
Comment 7•6 years ago
|
||
bugherder |
Status: ASSIGNED → RESOLVED
Closed: 6 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla63
You need to log in
before you can comment on or make changes to this bug.
Description
•