W3C home > Mailing lists > Public > public-audio@w3.org > April to June 2012

Re: a few question: gimbal lock, page visibility, outterconeangle

From: Alistair MacDonald <al@signedon.com>
Date: Tue, 17 Apr 2012 11:16:00 -0400
Message-ID: <CAJX8r2m_4YRejc3otHosgF9VRy8WQuvWE3S=oSx0rkOjmRNfbQ@mail.gmail.com>
To: Jerome Etienne <jerome.etienne@gmail.com>
Cc: Chris Rogers <crogers@google.com>, public-audio@w3.org
Hi Jerome,

Just wanted to check in and make sure you felt your questions have
been answered.

Is there anything on this thread that you feel would be good to discuss further?

Regards,

Alistair





On Tue, Mar 20, 2012 at 3:11 PM, Chris Rogers <crogers@google.com> wrote:
>
>
> On Tue, Mar 13, 2012 at 7:00 AM, Jerome Etienne <jerome.etienne@gmail.com>
> wrote:
>>
>> Hi,
>>
>> While experimenting with webaudio api[1], i hit some issues which
>> triggered questions. I am new to this mailing list, so im not sure it is the
>> proper forum. If it isnt, feel free to reroute me :)
>>
>> 1. How to handle page visibility ?
>> When the page isnt visible, what is the recomended behavior ? should i set
>> the volume to 0 ? something else ?
>> Currently i do nothing special, but it has glitches. The positional audio
>> isnt updated. And i hear "clicks" in the audio when the browser is busy on
>> some other pages (implementation issue related to scheduling?)
>
>
> Often the expected behavior would be to mute the sound (set gain/volume to
> 0).  But in some use cases, it may be interesting to keep the sound playing,
> for example if the user would like to continue hearing music while a tab is
> in the background.
>
>
>>
>>
>> 2. How to handle 'gimbal lock' with positional audio ?
>> When the sound is exactly on top of the listener, it is heard as if it was
>> from the left, then from the right. At first sight, it seems close to a
>> gimbal lock[2] issue. Is this diagnostic correct ? How to handle this ?
>> Should i ignore it and say 'well it is rare enougth, nobody will notice' ?
>> is that an issue in my code or in the browser implementation ?
>
>
> Interesting question.  I would expect the implementatation to handle
> transitions smoothly.  But, I should check the region exactly on top of the
> listener to see if I can re-create your problems in the WebKit
> implementation.  Perhaps there's a bug, or we're simply not interpolating as
> well as we could in that area.
>
>>
>>
>> 3. When using audiopannernode, coneOuterAngle seems to be ignored.
>> Let me explain. When i change .coneInnerAngle and .coneOuterGain, they
>> have the effect i expect. But no matter which value i put in
>> .coneOuterAngle, the outter cone seems to be the max possible. According my
>> understand, there are 3 states. "inside innerCone", "inside outerCone" and
>> "totally outside". Is that correct ? when the listener is "totally outside",
>> im under the impression no sound should be heard, correct ?
>
>
> Are you still having problems here?  The three zones you describe correctly,
> but "totally outside" is not necessarily "no sound" but instead depends on
> .coneOuterGain.  A quick look at the code seems to show it's doing the right
> thing, but if you are still having the problem then a small code sample
> would be helpful.
>
>>
>>
>> Lots of questions on various subjects :) Please, let me know what you
>> think.
>>
>>
>> Jerome
>>
>>
>> [1] https://dvcs.w3.org/hg/audio/raw-file/tip/webaudio/specification.html
>> [2] http://en.wikipedia.org/wiki/Gimbal_lock
>
>



-- 
Alistair MacDonald
SignedOn, Inc - W3C Audio WG
Boston, MA, (707) 701-3730
al@signedon.com - http://signedon.com
Received on Tuesday, 17 April 2012 15:16:35 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 17 April 2012 15:16:36 GMT