See also: IRC log
<rikrd_> hi f1lt3r
<rikrd_> the telecon starts in 20min, right?
<scribe> scribe: F1LT3R
<chrislo> Is the UK bridge number working today? I'm getting "the number is not in service".
MikeSmith, shepazu: any ideas on the UK bridge?
Conference code : 28346
Is anyone else getting a lot of noise?
<joeberkovitz> yes
is it still going on?
<joeberkovitz> just stopped
hmmm prehaps it was my phone
<rikrd_> ok, that seems to have been me, then
<Dick> yes there is a bit of static on the line this end
<joeberkovitz> i hear nothing
did everybody get on to the call?
<rikrd_> yes
<chris> yes
<joeberkovitz> yes
<Dick> yes
<chrislo> No
<chrislo> UK Bridge line is not working.
chrislo, we'll try and keep you up to date on irc, and i will look into the uk bridge later today
<chrislo> Thanks.
trackbot, start telecon
<trackbot> Date: 04 October 2010
<chrislo> F1LT3R: The only update I have is that we are running a XG Audio workshop for generating use cases at the BBC on Wednesday.
chrislo, that's great! :)
Standing Wave: http://github.com/maxl0rd/standingwave3
<chrislo> F1LT3R: I'll try and make notes in the form of http://www.w3.org/2005/Incubator/audio/wiki/Audio_API_Use_Cases
chrislo, perfect
rikrd: I think from the API point of view, I think filters like IIR and QUAD could be the same thing node?
Chris Rogers: So we could give different settings to the same node? That's possible, but one of the things that might be more difficult to implement, if you give it a 100 pole IIR filter, do you intend to change those dynamically?
rikrd: That's a good point. I think in most cases that would be constant. But we should discuss this on the email thread.
Chris Rogers: I would think for the convolver it should be unchanging, but for the biquad nodes you could change the params dynamically. Maybe there is some way we can put it into a single filter node, but we will have to consider the use cases.
joeberkovitz: Can you dynamically change the parameters of any node in this API?
Chris Rogers: I have been considering using parameter control curves that run as part of the compiled code, that are programmed in JavaScript.
joeberkovitz: I can tell you what we found to be useful. I think we found that the cubic spline interpolation and linear interpolation allowed us to add a good set of functionality to our application.
Chris Rogers: What about looping automation?
joeberkovitz: In our software, we allow several loop points.
Chris Rogers: Lets talk about these things on the email list.
Chris Rogers: It might be possible to get rid of the RealTimeAnalyzer. We could use the JavaScript processor node to do the FFT.
Al: What are the performance concerns with FFT in JavaScript?
Chris Rogers: You would definitely have problems running say, 100 FFTs in JavaScript. So perhaps we should leave the RealTimeAnalyzer in.
rikrd: If we are going to have a RealTimeAnalyzer we should have a basic FFTNode.
joeberkovitz: Do we need complex output in the frequency domain?
rikrd: In my work, I need it for use with things like spacial analysis.
Chris Rogers: We can try just doing it in JS and see if it is effecient enough, if not perhaps we can revisit this later.
Chris Rogers: I will do some testing and see if the graphics are as smooth as they were without processing in JavaScript.
rikrd: These node we have are really useful to process all kinds of signals, not just for audio.
Chris Rogers: What types of processing and what type of data?
rikrd: You use convolver nodes or fft to process data. For example: finding the energy of a cymbal.
Chris Rogers: I am going to have to program something like this for testing purposes and I think it will be useful in allowing you to work along the lines of your suggestions.
shepazu: Could you use the GPU to do some of the processing for audio?
Chris Rogers: The GPU tends not to be that much faster in terms of processing, and adds latency which is a major issue in audio. So I do not see much of an advantage in doing this.
rikrd: It is also not as parallelizable.
<shepazu> http://lists.w3.org/Archives/Public/public-html/2010Sep/0228.html
shepazu: I know Robin did follow up on this, but I am not certain what the current status is right now. I do not thinkit has gone any further since our last telecon. But the expectation is that the DAP WG will be the group working on this.
<shepazu> http://www.w3.org/2009/dap/
<shepazu> http://dev.w3.org/2009/dap/api-perms/
<shepazu> http://dev.w3.org/2009/dap/device/
Chris Rogers: It would be nice if there was a standard for this, but we can still put this in the API for future considerations.
shepazu: I will do some more research on where this is, and find out if progress is still being made with <device>.
shepazu: The incubator group
should be focused more on audio in the web, and it is more
appropriate to be working on the API within the working group.
I am proposing that this group should draw up a charter for a
Working Group and get it launched.
... Is there any decent on the idea of drawing up a charter and
launching a Working Group?
Room: *long pause* (no decent registered)
Resolution: this group will draw up a charter to start a new Audio Working Group.
Chris Rogers: we can pull in which ever high resolution timers the OS provides. What is more important is fleshing out how the API would integrate with this.
This is scribe.perl Revision: 1.135 of Date: 2009/03/02 03:52:20 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Found Scribe: F1LT3R Inferring ScribeNick: F1LT3R Default Present: +1.978.314.aaaa, +1.617.584.aabb, +1.510.334.aacc, +1.206.819.aadd, +1.978.314.aaee, +1.617.584.aaff, Shepazu Present: +1.978.314.aaaa +1.617.584.aabb +1.510.334.aacc +1.206.819.aadd +1.978.314.aaee +1.617.584.aaff Shepazu Found Date: 04 Oct 2010 People with action items: WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option.[End of scribe.perl diagnostic output]