Automatic update from web-platform-tests
[wpt PR 17233] - Active Processing for ConvolverNode, a=testonly
Automatic update from web-platform-tests
Active Processing for ConvolverNode
The ConvolverNode should output a single channel of silence when it is not
actively processing. Active processing is already supported and just needed
to have the convolver output a single channel of silence when it is
constructed, before anything is connected to it.
Bug: 971389
Change-Id: Icc894cd7ea6dec60a1eccb0fa4a3aea68ac5b51f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1644899
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#668149}
--
wp5At-commits: 6531eab802391a24dabca973cc4329abbeadd896
wpt-pr: 17233
bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1558611
gecko-commit: 3b9b543d1c8db5573155dc6ff44e65f871ae6ec9
gecko-integration-branch: central
gecko-reviewers: testonly
--
wpt-commits: 9e43b67f4fec9e0a1ffe2d91a5b4fd0c9b5ad705
wpt-pr: 17461
The spec says:
> An implementation MUST support sample rates in at least the range 8000 to 96000.
Differential Revision: https://phabricator.services.mozilla.com/D37435
--HG--
extra : moz-landing-system : lando
Changing the cutoff and lowering the windowing does not seem to work reliably
unfortunately.
Differential Revision: https://phabricator.services.mozilla.com/D36069
--HG--
extra : moz-landing-system : lando
Automatic update from web-platform-tests
Revert "Active Processing for ConvolverNode"
This reverts commit 499976c54a2b086bbde54ae137d0ac4adb3cd427.
Reason for revert:
Findit (https://goo.gl/kROfz5) identified CL at revision 668149 as the
culprit for flakes in the build cycles as shown on:
https://analysis.chromium.org/p/chromium/flake-portal/analysis/culprit?key=ag9zfmZpbmRpdC1mb3ItbWVyQwsSDEZsYWtlQ3VscHJpdCIxY2hyb21pdW0vNDk5OTc2YzU0YTJiMDg2YmJkZTU0YWUxMzdkMGFjNGFkYjNjZDQyNww
Sample Failed Build: https://ci.chromium.org/buildbot/chromium.mac/WebKit%20Mac10.13%20%28retina%29/15474
Sample Failed Step: webkit_layout_tests on ATI GPU on Mac Retina on Mac-10.13.6
Sample Flaky Test: external/wpt/webaudio/the-audio-api/the-convolvernode-interface/active-processing.https.html
Original change's description:
> Active Processing for ConvolverNode
>
> The ConvolverNode should output a single channel of silence when it is not
> actively processing. Active processing is already supported and just needed
> to have the convolver output a single channel of silence when it is
> constructed, before anything is connected to it.
>
> Bug: 971389
> Change-Id: Icc894cd7ea6dec60a1eccb0fa4a3aea68ac5b51f
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1644899
> Commit-Queue: Raymond Toy <rtoy@chromium.org>
> Reviewed-by: Hongchan Choi <hongchan@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#668149}
Change-Id: Ie7fedee332ca167c3dfd4fd99142de89234ff03a
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 971389
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1653924
Cr-Commit-Position: refs/heads/master@{#668243}
--
wp5At-commits: 2477f79f3c0757225d06762aeb1d3b1eea6eb52a
wpt-pr: 17288
Automatic update from web-platform-tests
ChannelMergerNode supports active processing
When a ChannelMergerNode is not actively processing its inputs, the
number of channels in its output is one, instead of the number of
inputs.
Just need to disable outputs when the node is created. Connections to
the node will adjust the number of channels appripriately, and when
inputs are no longer actively processing, the number of output
channels becomes one, as required.
Bug: 925046
Test: the-channelmergernode-interface/active-processing.https.html
Change-Id: I08e5fd2fe9c99b75af177eb85212162ecf39efd4
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1645478
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#668159}
--
wp5At-commits: bfb037131941940b9f7efa16cb1ec0829a2461cb
wpt-pr: 17273
Automatic update from web-platform-tests
Active Processing for ConvolverNode
The ConvolverNode should output a single channel of silence when it is not
actively processing. Active processing is already supported and just needed
to have the convolver output a single channel of silence when it is
constructed, before anything is connected to it.
Bug: 971389
Change-Id: Icc894cd7ea6dec60a1eccb0fa4a3aea68ac5b51f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1644899
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#668149}
--
wp5At-commits: 6531eab802391a24dabca973cc4329abbeadd896
wpt-pr: 17233
Automatic update from web-platform-tests
Add test for active processing of AudioBufferSourceNode
Test that when an AudioBufferSourceNode stops, the number of channels
of its output changes (eventually) to 1 as required when a node is not
actively processing anymore.
Chrome already supports this notion (via |DisableOutputs()|), so we're
just adding a test for this.
Bug: 954545
Test: the-audiobuffersourcenode-interface/active-processing.https.html
Change-Id: I4fab646df204720cc1efe2e3ae325e60cb22e49b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1643729
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#666013}
--
wp5At-commits: c027c41a085a874404922c638387c97df74b4ebf
wpt-pr: 17169
Automatic update from web-platform-tests
Restore original tail-processing for ScriptProcessor and AudioWorklet
While the idea in
https://chromium-review.googlesource.com/c/chromium/src/+/1544880 was
correct, we forgot to consider what happens when a source stops and is
disconnected to the ScriptProcessorNode (AudioWorkletNode). In this
case, the output of the node is disabled. But that's incorrect in
general because the node's processing must continue because it has
user-defined outputs so a silent input does not imply a silent output.
Rather than adding a special case in DisableOutputsIfNecessary, let's
just revert the change. The internal properties aren't exposed to the
developer.
Bug: 959125
Test: the-scriptprocessornode-interface/simple-input-output.html
Change-Id: I0f217be24eacff6031a537086df535e6c1e8d9d6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1598427
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#657447}
--
wp5At-commits: a00eb7eb65b3372a510f25419403d7adf82c952a
wpt-pr: 16715
Automatic update from web-platform-tests
Add AudioWorklet test for disconnected inputs
Follow up for issue 959125 where we add a test for an AudioWorklet to
verify that if a source connected to an AudioWorklet stops, the
AudioWorklet still continues to call process.
Bug: 960590, 959125
Test: the-audioworklet-interface/simple-input-output.https.html
Change-Id: I3db82c29737c084c576d81bd067741ca5e1845ab
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1600406
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#659252}
--
wpt-commits: 24278ab61781de72ed363b866ae6b50b86822b27
wpt-pr: 16732
It's not complete I think, but it covers the constructor and some bits of
behaviour.
Differential Revision: https://phabricator.services.mozilla.com/D32175
--HG--
extra : moz-landing-system : lando
It's not complete I think, but it covers the constructor and some bits of
behaviour.
Differential Revision: https://phabricator.services.mozilla.com/D32175
--HG--
extra : moz-landing-system : lando
It's not complete I think, but it covers the constructor and some bits of
behaviour.
Differential Revision: https://phabricator.services.mozilla.com/D32175
--HG--
extra : moz-landing-system : lando
Automatic update from web-platform-tests
Use PFFFT for WebAudio FFT on Android
Replace the OpenMAX DL FFT with PFFFT for WebAudio on Android.
Roughly, the library is as fast or faster than OpenMAX DL FFT and is not
a giant blob of handwritten assembly code. See the bug for a link for
additional information.
The changes in the tests are due to minor changes in accuracy. Apparently,
PFFFT is very slightly less accurate than OpenMAX DL. We deem these small changes
to be acceptable for the performance gain and readability.
Bug: 917355
Change-Id: Ia3efdab06364e312608e2aa7e060c46a8c8d0885
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1590257
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Reviewed-by: Nico Weber <thakis@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#656537}
--
wpt-commits: de1466c482c99f6549317e3d202aad03cfd234a4
wpt-pr: 16664
Automatic update from web-platform-tests
Test multi-threaded ConvolverNode
Add test for ConvolverNode running on a real-tiem AudioContext. In
addition to being a general test, this also tests that the
multi-threaded convolver implementation in Chrome (and perhaps other
browsers) works correctly.
The length of the convolver imuplse response was selected to ensure
that extra threads are used in Chrome. This was verified by putting a
(temporary) print where the threads are created.
Bug: 956085
Test: realtime-conv.html
Change-Id: Ia273f2e3fdba0b0ac79a03507801dc1c7b3c5006
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1586231
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#654637}
--
wpt-commits: 3058cf00eedc56bc305ea5ebad0283f11ab45784
wpt-pr: 16564
Automatic update from web-platform-tests
Test getFreguencyResponse for all BiquadFilter types
Previously we only tested the result of getFreqencyResponse for
peaking filters. Add test for all the other filters as well.
Bug: 390266
Change-Id: I7cba2a0dbcfaf3e6c0fec0a92eaedf8779f2d361
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1584368
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#654482}
--
wpt-commits: dfad25523d57c31e901c627a4cecc6bc9979e9c6
wpt-pr: 16542
Automatic update from web-platform-tests
Same events at the same time don't replace each other
Previously, if multiple events of the same type were scheduled at the
same time, then each new event would replace the existing event. This
is incorrect according to the spec. There is no special treatment
here and events are inserted in the order in which they're received.
This change only affects linear and exponential ramps. The result is
that the ramp preceeds up to the first event value, and then at the
event time, the output instantly jumps to the last event value.
Bug: 925037
Test: the-audioparam-interface/event-insertion.html updated
Change-Id: I2de32c7a71c12e5673b2db8cf81f9dc48f3ee458
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1577661
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#654242}
--
wpt-commits: 9612df0a355c80a2afde1bde6957d03cff3a99ee
wpt-pr: 16432
Automatic update from web-platform-tests
MediaStreamAudioDestinationNode has no outputs
Set number of outputs to 0 as required by the spec.
This requires some small refactoring of AudioBasicInspectorHandler too.
One Chrome test is also moved to WPT. Manually tested with Firefox (nightly) which
passes all the tests.
Spec: https://webaudio.github.io/web-audio-api/#mediastreamaudiodestinationnode
Feature: https://www.chromestatus.com/feature/5702493226926080#
Bug: 691806
Change-Id: Ic17b4f248b606701dfa3ca54892754d7ebe4e749
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1580202
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Reviewed-by: Kent Tamura <tkent@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#654503}
--
wpt-commits: 7012c4e38352fc0501409cd898a1c810f41f9aaa
wpt-pr: 16527
Automatic update from web-platform-tests
Always leave an event in the AudioParam timeline
Previously, when the last AudioParam event has been processed and is
now in the past, the event would be removed from the timeline. This
was an optimization so that we can skip sample-accurate processing of
the timeline and just returned fixed values. However, if a new event
is added after all events have been deleted, the starting point of the
event could be incorrect. For example a setValueAtTime event in the
past followed by a linearRampToValueAtTime.
Therefore, always leave the very last event in the timeline to
establish the starting event if needed by future events.
Bug: 951863
Change-Id: Iae0b550a4b497b953a10e85e603ecebb59508a99
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1565375
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#652955}
--
wpt-commits: 83d93704f0cb36dd9fa3f45dd3e4b34990e934e6
wpt-pr: 16373
Automatic update from web-platform-tests
Adjust test threshold for win10
biquad-lowpass.html test is failing on windows 10 because it looks
like the threshold is too tight. Relax the threshold a bit.
Not sure why it should start failing now, but the change in threshold
isn't huge (a bit larger than single-float epsilon).
Bug: 952780
Change-Id: Ifcf1fb083d6a758e6c76ed5c251ae89e8371ac54
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1570325
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Reviewed-by: Takuto Ikuta <tikuta@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#651495}
--
wpt-commits: dc199888f4adc90c40e3157d74678b0954beb264
wpt-pr: 16370
Automatic update from web-platform-tests
Fix flakiness in audioworklet-suspend.https.html
It is because the actual suspension takes time to stop the render loop
completely. So the current time advances a little after the suspension
and the assertion that compares two timestamps fails.
This CL fixes the problem by:
1. Suspend the context immediately after its creation.
2. Cache timestamp A after the worklet module loading is completed.
3. Check timestamp B after 500ms. If the context is not suspended
(which is the original bug), two timestamps A and B won't match.
Bug: 953332
Test: external/wpt/webaudio/the-audio-api/the-audioworklet-interface/audioworklet-suspend.https.html
Change-Id: I3fc4ccf78ec9c137128f8605d502692be303b02e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1570129
Reviewed-by: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Robert Ma <robertma@chromium.org>
Commit-Queue: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#651437}
--
wpt-commits: 3f690a244306cae2276fd1e5f620f5e83096a445
wpt-pr: 16368
Automatic update from web-platform-tests
Do not resume a suspended BaseAudioContext when AudioWorklet starts
When AudioWorklet.addModule() activates the worklet thread, it
re-starts the destination handler even when the context is suspended.
This CL fixes the bug by restarting destination only when the context
is in the "running" state.
Bug: 950997
Test: external/wpt/webaudio/the-audio-api/the-audioworklet-interface/audioworklet-suspend.https.html
Change-Id: I0b44e3fca23bd069fac21f709fc665e8eacc4dd3
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1560140
Commit-Queue: Hongchan Choi <hongchan@chromium.org>
Reviewed-by: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#649303}
--
wpt-commits: 81d09351b7bd61a270d58059aba662f17fec536a
wpt-pr: 16303
This is a no-op, but is allowed per spec:
> An optional offset to copy the data to. If startInChannel is greater than the
> length of the AudioBuffer, an IndexSizeError MUST be thrown.
If the numbers are equals, this shouldn't throw.
Differential Revision: https://phabricator.services.mozilla.com/D29358
--HG--
extra : moz-landing-system : lando
Automatic update from web-platform-tests
Consolidate .gitignore files under the root (#15952)
Multiple .gitignores aren't supported by the manifest update code when running
against the filesystem, so put all the rules in the root file instead.
This also adds a lint to check for extra .gitignore files.
--
wpt-commits: 46712a196fbac08bac95203f4d0861d25268ea7f
wpt-pr: 15952
Automatic update from web-platform-tests
Add <meta name=timeout content=long> to WPT in SlowTests
This is the 2nd step in the process of banning external/wpt from
SlowTests. After cleaning up the wtp entries we're marking them as
slow by using wpt idioms, i.e.:
1) Add <meta name="timeout" content="long"> to .html files
2) Add // META: timeout=long to .js files
Bug: 765026
Change-Id: Iaf051104f419524916e19324996cc36ee37d5426
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1497005
Commit-Queue: Sergio Villar <svillar@igalia.com>
Reviewed-by: Andy Paicu <andypaicu@chromium.org>
Reviewed-by: Robert Ma <robertma@chromium.org>
Cr-Commit-Position: refs/heads/master@{#643402}
--
wpt-commits: 7287608f90f6b9530635d10086fd2ab386faab38
wpt-pr: 15630
Automatic update from web-platform-tests
Consolidate .gitignore files under the root (#15952)
Multiple .gitignores aren't supported by the manifest update code when running
against the filesystem, so put all the rules in the root file instead.
This also adds a lint to check for extra .gitignore files.
--
wpt-commits: 46712a196fbac08bac95203f4d0861d25268ea7f
wpt-pr: 15952
Automatic update from web-platform-tests
Add <meta name=timeout content=long> to WPT in SlowTests
This is the 2nd step in the process of banning external/wpt from
SlowTests. After cleaning up the wtp entries we're marking them as
slow by using wpt idioms, i.e.:
1) Add <meta name="timeout" content="long"> to .html files
2) Add // META: timeout=long to .js files
Bug: 765026
Change-Id: Iaf051104f419524916e19324996cc36ee37d5426
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1497005
Commit-Queue: Sergio Villar <svillar@igalia.com>
Reviewed-by: Andy Paicu <andypaicu@chromium.org>
Reviewed-by: Robert Ma <robertma@chromium.org>
Cr-Commit-Position: refs/heads/master@{#643402}
--
wpt-commits: 7287608f90f6b9530635d10086fd2ab386faab38
wpt-pr: 15630
Automatic update from web-platform-tests
Use matching sample rate for the context as for the reference file
The reference file has a certain sample rate. Use the same for the
context to reduce the amount of resampling that might be required.
Regenerated the webm-decode-expected.wave file with the fixed sample rate. This also
changed the SNR slightly, so update the SNR test accordingly. Also manually compared
the new and old files. These differ by at most 1 LSB (16-bit).
Bug: 940970
Change-Id: I979bbeb512c36d7e2eb4cba85550fa4177b8aabd
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1516723
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#640341}
--
wpt-commits: 1d62ca809d6c859a9ef5f48f82dac4a366c6a0eb
wpt-pr: 15790
Automatic update from web-platform-tests
ABSN with null buffer should output silence
If an AudioBufferSourceNode is set with a non-null AudioBuffer and is
then set to null, the output from the buffer should be silence.
Bug: 938713
Test: audiobuffersource-null.html
Change-Id: I4c5d3363f0176503b5ed8fb45f46e776b4e05b42
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1507035
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#639744}
--
wpt-commits: 26c03768165078f56415db91c78ff64bcf9cc142
wpt-pr: 15710
Automatic update from web-platform-tests
Add support for AudioContextOptions sampleRate
Optional sampleRate parameter to AudioContextOptions, if provided then
the AudioContext will run at this sampleRate, otherwise it will run at
the hardware rate.
Running the AudioContext at a lower sample rate can allow running a
heavier graph on low-end devices, for example if a given device is able
to run with X nodes at a sample rate of 48k, it can support roughly
2X nodes at 24k or 4X nodes at 12k.
Bug: 432248
Change-Id: I835f28b4625763bd6a6ddbee9b89636ef8a8a066
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1482957
Reviewed-by: Kent Tamura <tkent@chromium.org>
Reviewed-by: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Commit-Queue: Andrew MacPherson <andrew.macpherson@soundtrap.com>
Cr-Commit-Position: refs/heads/master@{#638498}
--
wpt-commits: 9b994b36003246370b352fa70c563866a8443e4e
wpt-pr: 15520
Automatic update from web-platform-tests
Relax required SNR a bit
The required SNR was a little too high for some Mac OS bots. Make it
a little smaller.
Bug: 929435
Change-Id: I5fbfe6573f4907f33999085e52bb22f4b4b9a126
Reviewed-on: https://chromium-review.googlesource.com/c/1457228
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#631309}
--
wpt-commits: 91d81bb73d5e8424e45918c84e275a9d46c51d51
wpt-pr: 15270
Automatic update from web-platform-tests
Relax required SNR a bit
The required SNR was a little too high for some Mac OS bots. Make it
a little smaller.
Bug: 929435
Change-Id: I5fbfe6573f4907f33999085e52bb22f4b4b9a126
Reviewed-on: https://chromium-review.googlesource.com/c/1457228
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#631309}
--
wpt-commits: 91d81bb73d5e8424e45918c84e275a9d46c51d51
wpt-pr: 15270
Automatic update from web-platform-tests
Round up to the next render quantum for suspend
OfflineAudioContext.suspend rounds the time up to the next render
quantum instead of down. It is sometimes confusing when suspend
occurs before the given time, which is different from how everything
else works in WebAudio.
A few tests need to be fixed because they computed the boundary by
rounding down instead of up. One test needed to be adjusted because
the times were rounded to different boundaries instead of the same
boundary.
See also https://github.com/WebAudio/web-audio-api/issues/1822
Bug: 927895
Change-Id: Ie6685c620dd38eb2a059901d13b907764ba1a2db
Reviewed-on: https://chromium-review.googlesource.com/c/1450536
Reviewed-by: Kent Tamura <tkent@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#630346}
--
wpt-commits: 1b1a627e597f15b60327191799b8081beac2a3cd
wpt-pr: 15214
Automatic update from web-platform-tests
Use oversampling to compute frame number
When computing the frame number, round-off can happen such that the
resulting frame number is not the same as the original. To work
around this, use oversampling to compute the oversampled frame
number. Round that to the nearest frame, and then convert that back
to the desired integer frame number at the original sample rate.
This won't always work, but should make the round-off errors much less
likely.
Bug: 876917
Test: the-audiobuffersourcenode-interface/sub-sample-buffer-stitching.html
Change-Id: Id3658a986490d2aec07a382497bc24a4d99e416c
Reviewed-on: https://chromium-review.googlesource.com/c/1446869
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#629697}
--
wpt-commits: 5a6917c5cb8ea8602c53fd927fe3c208bb99b463
wpt-pr: 15215
--HG--
rename : testing/web-platform/tests/webaudio/the-audio-api/the-audiobuffersourcenode-interface/resources/sub-sample-scheduling.html => testing/web-platform/tests/webaudio/the-audio-api/the-audiobuffersourcenode-interface/sub-sample-scheduling.html
Automatic update from web-platform-tests
Update cached azimuth/elevation/cone gain
Sample accurate panner automations need to update the cached value for
the azimuth, elevation, and cone gain at the end of each render in
case the automations end (because all automations are sufficiently far
in the past). When this happens, the panner switches from a-rate
sample-accurate method to k-rate method that uses the cached values
when nothing is moving. If the a-rate method doesn't update, then the
k-rate method uses the old values which are now incorrect.
Bug: 917757
Test: the-panner-node/automation-changes.html
Change-Id: I7289c278b1cc9dc31d9e3f5698ac467cb1474ca5
Reviewed-on: https://chromium-review.googlesource.com/c/1427795
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#627129}
--
wpt-commits: 48e44386b5c07758ddd22c8227701c37beebe76f
wpt-pr: 15005
Automatic update from web-platform-tests
Sub-sample accurate start for ABSN
Implement sub-sample accurate start for AudioBufferSourceNode.
Previously, if the start time was between sample boundaries, we would
start the output at the frame before the start time. This is actually
incorrect because we haven't actually started yet. We should start
the output at the next boundary, and interpolating the value based on
the true start time and the sample boundary.
Many tests needed to be updated. Basically for each test that needed
to be changed, the sample rate is set to a power of two and all ABSN
sources are updated to make sure the source starts exactly on a frame
boundary. We also took the opportunity to adjust the error thresholds
for the tests in case lower values could be used.
Some additional notes for the tests:
audiobuffersource-playbackrate-zero.html: Add new test to make sure
sub-sample accurate start handles a zero playback rate.
audiobuffersource-loop-points.html: add some code to save the actual
output. This is needed because a new reference file is needed since
the sample rate has changed.
Also manually tested all of the modified tests with Firefox nightly.
They all pass still (except for the new sub-sample test because
Firefox doesn't do sub-sample accurate start/stop).
Bug: 876917
Test: the-audiobuffersourcenode-interface/sub-sample-scheduling.html
Change-Id: Ib13ba30eaa160cfd10739feabac961bf074ee309
Reviewed-on: https://chromium-review.googlesource.com/c/1212270
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#626753}
--
wpt-commits: 00fa50687cab43b660296389acad6cc48717f1d1
wpt-pr: 13061
Automatic update from web-platform-tests
Compute azimuth correctly according to the spec
The azimuth calculation is incorrect if one of the vectors is the
zero vector. The spec handles this case correctly by just using the
dot product alone with "normalized" vectors instead of trying to
divide the dot product by the lengths.
Bug: 893983
Test: panner-azimuth.html
Change-Id: I7f890ae3353a71ffccfc3138038572a8a18b2379
Reviewed-on: https://chromium-review.googlesource.com/c/1277666
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#626665}
--
wpt-commits: f2b7361daff0d555c271f8ffa50dc22ac8e948c5
wpt-pr: 13477
Automatic update from web-platform-tests
Don't overflow when event times are very close in time
If a linear ramp starts very close to 0, the computation of the output
can sometimes overflow and produce NaN. Clamp the time difference so
thia doesn't happen and internally treat it as if the difference were
zero.
Add test that NaN doesn't occur and also verify that the clamping
doesn't change existing interpolation.
Bug: 899507
Test: the-audioparam-interface/audioparam-close.html
Change-Id: Ic91cec577fdb4e0774d51fff66da2947baa98263
Reviewed-on: https://chromium-review.googlesource.com/c/1411651
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#623394}
--
wpt-commits: 75318c963f97a674d9e171be761d90e67517a076
wpt-pr: 14897
Automatic update from web-platform-tests[media] Treat cross-origin redirect as TAINTED only for no-cors requests
With https://crrev.com/a9cbaa7a40e2b2723cfc2f266c42f4980038a949,
WebMediaPlayer blindly treats a resource experiencing cross-origin
redirects as TAINTED. In fact, it should be treated as TAINTED only
when its request mode is "no-cors".
The added tests are provided by hongchan@chromium.org.
Bug: 899745, 901383
Change-Id: Idb66407552085b053818f3e4a9d8d5ff3ddeaf45
Reviewed-on: https://chromium-review.googlesource.com/c/1325281
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Reviewed-by: Fredrik Hubinette <hubbe@chromium.org>
Commit-Queue: Yutaka Hirano <yhirano@chromium.org>
Cr-Commit-Position: refs/heads/master@{#606681}
--
wpt-commits: d5be80a86d4f938250c075ac12414ad47516969c
wpt-pr: 13980
Automatic update from web-platform-testsRemove some suggested reviewers who have asked to be removed (#13796)
Part of https://github.com/web-platform-tests/wpt/issues/13702.
--
wpt-commits: 75b0f336c50105c6fea47ad253d57219dfa744d3
wpt-pr: 13796
Automatic update from web-platform-tests[run_web_tests] Remove extra baselines after crrev.com/c/1277654
After crrev.com/c/1277654, we no longer generate text result of many
tests, thus these tests no longer need the text baselines.
This CL removes all the extra baselines, including the extra text
baselines after the above CL, and also other kind of extra baselines
(e.g. extra -expected.png for reftests).
Bug: 703899
Cq-Include-Trybots: luci.chromium.try:linux-blink-gen-property-trees;luci.chromium.try:linux_layout_tests_layout_ng;luci.chromium.try:linux_layout_tests_slimming_paint_v2;master.tryserver.blink:linux_trusty_blink_rel
Change-Id: Iadc647d0393e3f61a7d983bc3946042602de70c0
Reviewed-on: https://chromium-review.googlesource.com/c/1292599
Commit-Queue: Xianzhu Wang <wangxianzhu@chromium.org>
Reviewed-by: Ned Nguyen <nednguyen@google.com>
Cr-Commit-Position: refs/heads/master@{#602195}
--
wpt-commits: fd1e161dd0529bb189eac76250bd4d8718d15150
wpt-pr: 13693
Automatic update from web-platform-tests[run_web_tests] Check for extra baselines
After https://chromium-review.googlesource.com/c/chromium/src/+/1277654,
we no longer generate text result of many tests, thus these tests no
longer need the text baselines.
This CL checks for extra baselines in the following cases:
1. Extra -expected.png (existing logic) for ref tests;
2. Extra -expected.txt, -expected.png, -expected.wav if the test doesn't
generate the result;
3. Extra -expected.txt for all-pass testharness tests.
For 2, the extra baseline will be treated as a mismatch failure (which
can be interpreted as that the test doesn't generate the result, but
the baseline expects the result).
For 1 and 3, there are just warning messages because for now we don't
have a good way to present them in the layout test result page.
We have already had a similar logic in pre-submit for 3 and the new
check in run_web_tests is more accurate. For other types of extra
baselines, it's hard to implement the check in presubmit because it
needs more knowledge than static pre-submit knows.
Bug: 703899
Cq-Include-Trybots: luci.chromium.try:linux-blink-gen-property-trees;luci.chromium.try:linux_layout_tests_layout_ng;luci.chromium.try:linux_layout_tests_slimming_paint_v2;master.tryserver.blink:linux_trusty_blink_rel
Change-Id: I9016d5583791f179f1d51561245602af615d21de
Reviewed-on: https://chromium-review.googlesource.com/c/1286894
Commit-Queue: Xianzhu Wang <wangxianzhu@chromium.org>
Reviewed-by: Ned Nguyen <nednguyen@google.com>
Cr-Commit-Position: refs/heads/master@{#602378}
--
wpt-commits: 07d9f452d740d98ded5b11c176ba783fe9d80e5b
wpt-pr: 13698
--HG--
rename : testing/web-platform/tests/webaudio/the-audio-api/the-audiobuffersourcenode-interface/audiobuffersource-multi-channels-expected.wav => testing/web-platform/tests/webaudio/the-audio-api/the-audiobuffersourcenode-interface/resources/audiobuffersource-multi-channels-expected.wav
Raymond Toy
Paul Adenot
shindli
Razvan Maries
Findit
Oana Pop Rus
Cosmin Sabou
Hongchan Choi
jgraham
Sergio Villar Senin
Sebastian Hengst
Andrew MacPherson
Brindusan Cristian
Yutaka Hirano
Philip Jägenstedt
Xianzhu Wang