Compute tail time from Biquad coefficients

Compute tail time from Biquad coefficients

Instead of using a fixed tail time of 200 ms for all biquad filters,
we compute the actual tail time from filter coefficients.  Many
typical filters in applications have a tail time of 10 ms or less.
This allows biquads and downstream nodes to stop processing much sooner
than what currently happens, reducing CPU usage and allowing GC to
collect nodes sooner.

BUG=612215
TEST=

Review-Url: https://codereview.chromium.org/2862373002
Cr-Original-Commit-Position: refs/heads/master@{#475632}
Committed: https://chromium.googlesource.com/chromium/src/+/0be695dcb495cade0ed5c4b70aba95eb3dce49bd
Review-Url: https://codereview.chromium.org/2862373002
Cr-Commit-Position: refs/heads/master@{#476113}
Committed: https://chromium.googlesource.com/chromium/src/+/ea0745f9549a61349c37d5e853f8337f4ee79d5e

[Raymond Toy, 2017-05-22 21:10:50]

rtoy@chromium.org changed reviewers:
+ hongchan@chromium.org

[Raymond Toy, 2017-05-22 21:10:51]

PTAL

[Raymond Toy, 2017-05-23 16:57:43]

Description was changed from

==========
Compute tail time from Biquad coefficients

BUG=612215
TEST=
==========

to

==========
Compute tail time from Biquad coefficients

Instead of using a fixed tail time of 200 ms for all biquad filters,
we compute the actual tail time from filter coefficients.  Many
typical filters in applications have a tail time of 10 ms or less.
This allows biquads and downstream nodes to stop processing much sooner
than what currently happens, reducing CPU usage and allowing GC to
collect nodes sooner.

BUG=612215
TEST=
==========

[hongchan, 2017-05-24 22:21:35]

I won't go into detail on the mathematics of tail-time computation.

lgtm with nits in layout tests.

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
File third_party/WebKit/LayoutTests/webaudio/BiquadFilter/tail-time-allpass.html
(right):

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/webaudio/BiquadFilter/tail-time-allpass.html:44:
taskInfo:
I would rename this to: description

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/webaudio/BiquadFilter/tail-time-allpass.html:46:
testInfo: {
and this to: parameters

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/webaudio/BiquadFilter/tail-time-allpass.html:96: 
Why an empty line?

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
File third_party/WebKit/LayoutTests/webaudio/BiquadFilter/test-tail-time.js
(right):

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/webaudio/BiquadFilter/test-tail-time.js:76:
console.log(filtered);
This needs to be removed.

[Raymond Toy, 2017-05-25 18:57:14]

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
File third_party/WebKit/LayoutTests/webaudio/BiquadFilter/tail-time-allpass.html
(right):

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/webaudio/BiquadFilter/tail-time-allpass.html:44:
taskInfo:
On 2017/05/24 22:21:34, hongchan wrote:
> I would rename this to: description

Done.

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/webaudio/BiquadFilter/tail-time-allpass.html:46:
testInfo: {
On 2017/05/24 22:21:34, hongchan wrote:
> and this to: parameters

Done.

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/webaudio/BiquadFilter/tail-time-allpass.html:96: 
On 2017/05/24 22:21:34, hongchan wrote:
> Why an empty line?

Done.

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
File third_party/WebKit/LayoutTests/webaudio/BiquadFilter/test-tail-time.js
(right):

https://codereview.chromium.org/2862373002/diff/100001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/webaudio/BiquadFilter/test-tail-time.js:76:
console.log(filtered);
On 2017/05/24 22:21:34, hongchan wrote:
> This needs to be removed.

Done.

[Raymond Toy, 2017-05-30 17:57:01]

The CQ bit was checked by rtoy@chromium.org

[Raymond Toy, 2017-05-30 17:57:01]

The patchset sent to the CQ was uploaded after l-g-t-m from
hongchan@chromium.org
Link to the patchset: https://codereview.chromium.org/2862373002/#ps120001
(title: "Address review comments")

[commit-bot: I haz the power, 2017-05-30 17:57:26]

CQ is trying da patch.

Follow status at:
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-05-30 19:56:05]

CQ is committing da patch.

Bot data: {"patchset_id": 120001, "attempt_start_ts": 1496167021027250,
"parent_rev": "4523afedd7271dcdd8a2d310e219ecb8c63fff63", "commit_rev":
"0be695dcb495cade0ed5c4b70aba95eb3dce49bd"}

[commit-bot: I haz the power, 2017-05-30 19:56:14]

Description was changed from

==========
Compute tail time from Biquad coefficients

Instead of using a fixed tail time of 200 ms for all biquad filters,
we compute the actual tail time from filter coefficients.  Many
typical filters in applications have a tail time of 10 ms or less.
This allows biquads and downstream nodes to stop processing much sooner
than what currently happens, reducing CPU usage and allowing GC to
collect nodes sooner.

BUG=612215
TEST=
==========

to

==========
Compute tail time from Biquad coefficients

Instead of using a fixed tail time of 200 ms for all biquad filters,
we compute the actual tail time from filter coefficients.  Many
typical filters in applications have a tail time of 10 ms or less.
This allows biquads and downstream nodes to stop processing much sooner
than what currently happens, reducing CPU usage and allowing GC to
collect nodes sooner.

BUG=612215
TEST=

Review-Url: https://codereview.chromium.org/2862373002
Cr-Commit-Position: refs/heads/master@{#475632}
Committed:
https://chromium.googlesource.com/chromium/src/+/0be695dcb495cade0ed5c4b70aba...
==========

[commit-bot: I haz the power, 2017-05-30 19:56:15]

Committed patchset #7 (id:120001) as
https://chromium.googlesource.com/chromium/src/+/0be695dcb495cade0ed5c4b70aba...

[tyoshino (SeeGerritForStatus), 2017-05-31 06:04:40]

A revert of this CL (patchset #7 id:120001) has been created in
https://codereview.chromium.org/2912293002/ by tyoshino@chromium.org.

The reason for reverting is: The added layout tests are failing on "WebKit Linux
Trusty MSAN" bot.
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Linux%20Trusty....

[kjellander_chromium, 2017-05-31 13:16:26]

A revert of this CL (patchset #7 id:120001) has been created in
https://codereview.chromium.org/2919503002/ by kjellander@chromium.org.

The reason for reverting is: Breaks WebKit Linux Trusty MSAN:
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Linux%20Trusty...
WebKit Linux Trusty MSAN.

[Raymond Toy, 2017-05-31 18:50:53]

Description was changed from

==========
Compute tail time from Biquad coefficients

Instead of using a fixed tail time of 200 ms for all biquad filters,
we compute the actual tail time from filter coefficients.  Many
typical filters in applications have a tail time of 10 ms or less.
This allows biquads and downstream nodes to stop processing much sooner
than what currently happens, reducing CPU usage and allowing GC to
collect nodes sooner.

BUG=612215
TEST=

Review-Url: https://codereview.chromium.org/2862373002
Cr-Commit-Position: refs/heads/master@{#475632}
Committed:
https://chromium.googlesource.com/chromium/src/+/0be695dcb495cade0ed5c4b70aba...
==========

to

==========
Compute tail time from Biquad coefficients

Instead of using a fixed tail time of 200 ms for all biquad filters,
we compute the actual tail time from filter coefficients.  Many
typical filters in applications have a tail time of 10 ms or less.
This allows biquads and downstream nodes to stop processing much sooner
than what currently happens, reducing CPU usage and allowing GC to
collect nodes sooner.

BUG=612215
TEST=

Review-Url: https://codereview.chromium.org/2862373002
Cr-Commit-Position: refs/heads/master@{#475632}
Committed:
https://chromium.googlesource.com/chromium/src/+/0be695dcb495cade0ed5c4b70aba...
==========

[Raymond Toy, 2017-05-31 19:06:04]

PTAL. The only change is initializing tail_time_ in the BiquadDSPKernel
constructor.  Locally verified with my msan build that this no longer causes the
msan build to crash.

[hongchan, 2017-05-31 19:37:31]

lgtm

[Raymond Toy, 2017-05-31 19:38:58]

On 2017/05/31 19:37:31, hongchan wrote:
> lgtm

Thanks.  I'll wait for the msan bot to finish before attempting to land this
again.  But it did work on my local msan build.

[Raymond Toy, 2017-05-31 21:51:41]

The CQ bit was checked by rtoy@chromium.org

[commit-bot: I haz the power, 2017-05-31 21:52:30]

CQ is trying da patch.

Follow status at:
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-06-01 00:41:19]

CQ is committing da patch.

Bot data: {"patchset_id": 140001, "attempt_start_ts": 1496267501052300,
"parent_rev": "a94a9b3c6b33bebe664967c1f63c0ecdcaa053bd", "commit_rev":
"ea0745f9549a61349c37d5e853f8337f4ee79d5e"}

[commit-bot: I haz the power, 2017-06-01 00:41:28]

Description was changed from

==========
Compute tail time from Biquad coefficients

Instead of using a fixed tail time of 200 ms for all biquad filters,
we compute the actual tail time from filter coefficients.  Many
typical filters in applications have a tail time of 10 ms or less.
This allows biquads and downstream nodes to stop processing much sooner
than what currently happens, reducing CPU usage and allowing GC to
collect nodes sooner.

BUG=612215
TEST=

Review-Url: https://codereview.chromium.org/2862373002
Cr-Commit-Position: refs/heads/master@{#475632}
Committed:
https://chromium.googlesource.com/chromium/src/+/0be695dcb495cade0ed5c4b70aba...
==========

to

==========
Compute tail time from Biquad coefficients

Instead of using a fixed tail time of 200 ms for all biquad filters,
we compute the actual tail time from filter coefficients.  Many
typical filters in applications have a tail time of 10 ms or less.
This allows biquads and downstream nodes to stop processing much sooner
than what currently happens, reducing CPU usage and allowing GC to
collect nodes sooner.

BUG=612215
TEST=

Review-Url: https://codereview.chromium.org/2862373002
Cr-Original-Commit-Position: refs/heads/master@{#475632}
Committed:
https://chromium.googlesource.com/chromium/src/+/0be695dcb495cade0ed5c4b70aba...
Review-Url: https://codereview.chromium.org/2862373002
Cr-Commit-Position: refs/heads/master@{#476113}
Committed:
https://chromium.googlesource.com/chromium/src/+/ea0745f9549a61349c37d5e853f8...
==========

[commit-bot: I haz the power, 2017-06-01 00:41:30]

Committed patchset #8 (id:140001) as
https://chromium.googlesource.com/chromium/src/+/ea0745f9549a61349c37d5e853f8...