W3C home > Mailing lists > Public > public-audio@w3.org > July to September 2012

Re: Detecting time-averaged levels

From: Joe Berkovitz <joe@noteflight.com>
Date: Fri, 3 Aug 2012 18:36:42 -0400
Message-ID: <CA+ojG-YPhca9xsRLLbupfkSBSgfAnEGPmfFNGe3ReWoP5RiTVQ@mail.gmail.com>
To: Chris Wilson <cwilso@google.com>
Cc: Audio Working Group <public-audio@w3.org>
Sounds clever and workable!
On Aug 3, 2012 6:29 PM, "Chris Wilson" <cwilso@google.com> wrote:

> Another way I'd considered doing this was using a DynamicsCompressor node,
> cranking up the ratio and cranking down the threshold and knee, adjusting
> the attack and release as desired - then the "reduction" property becomes a
> convenient sampling of level.
>
> On Fri, Aug 3, 2012 at 12:00 PM, Joseph Berkovitz <joe@noteflight.com>wrote:
>
>> A number of scenarios in the Use Case document require a time-averaged
>> level for some audio pathway -- maybe something like an RMS using a rolling
>> average of squared signal values.
>>
>> Anyway, I'm wondering if I have missed any simple way of calculating such
>> a level, or whether one must fall back to a JavaScriptAudioNode to do this.
>>  It kind of seems natural for a RealtimeAnalyzerNode to do this even though
>> it's not frequency-domain data.
>>
>> Regards,
>>
>>        ... .  .    .       Joe
>>
>> *Joe Berkovitz*
>> President
>>
>> *Noteflight LLC*
>> Boston, Mass.
>> phone: +1 978 314 6271
>>        www.noteflight.com
>>
>>
>
Received on Friday, 3 August 2012 22:37:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 3 August 2012 22:37:13 GMT