W3C

- DRAFT -

Audio Working Group Teleconference

10 Mar 2016

See also: IRC log

Attendees

Present
padenot, ChrisL, joe, hongchan, rtoyg_m, jdsmith, cwilso, mdjp, BillHofmann, ChrisL_
Regrets
Chair
SV_MEETING_CHAIR
Scribe
joe

Contents


<mdjp> trackbot, start meeting

<trackbot> Meeting: Audio Working Group Teleconference

<trackbot> Date: 10 March 2016

<mdjp> I have some internet issues here, struggling to connect to webex

<rtoyg_m> Same here. Doesn't seem to accept the meeting code.

<mdjp> ah - I'm not even getting that far

<mdjp> I'm guessing it still needs the host to join

<jdsmith> I'm on webex. It shows joe and Matt as well.

<jdsmith> No audio though.

<rtoyg_m> Has the meeting code changed? It won't accept it.

<hongchan> WebEX says 'the meeting can't be found'

<jdsmith> I just ran this link: https://mit.webex.com/mit/j.php?MTID=m95d805e15fcae1a4b63eb595f048ec35

<mdjp> ChrisL_ are you joiing webex?

<mdjp> I think it will spring into life once he does

<rtoyg_m> jdsmith: I'm trying that now.

<BillHofmann> POTS dialin doesn't work because host hasn't joined the meeting.

<BillHofmann> therefore, no audio.

<padenot> curious

<BillHofmann> well, the chairs need to have the host code, that's clear.

<jdsmith> I see folks talking, but don't hear audio. Is it just me?

<padenot> jdsmith: I can hear some folks here

yes, we are live now

<padenot> I might afk for 5min in 5min

<ChrisL_> the chairs do have the host code, but it sems they need an MIT kerberos authentication key too

<BillHofmann> Fabulous.

Audioworker update

<BillHofmann> who is scribing?

<jdsmith> Looks like my local audio is broken. I'll leave and rejoin shortly.

<rtoyg_m> We're in!

<scribe> scribenick: joe

padenot: we've made progress on proxy model. lots of comments, thanks everyone. All comments have been addressed.
... we're almost there. 10 patches on worklets text since last looked at it. ready to start on audioworker.
... will be working in the alps on audioworker next week w/ no distractions.

mdjp: we've moved a long way in the last couple of weeks
... anthing blocking this?

padenot: it's all clear now. we just need to do the work

mdjp: we'll keep this at top of the agenda for the next few meetings. any other comments?

New and Uncommitted issues https://github.com/WebAudio/web-audio-api/issues?q=is%3Aopen+is%3Aissue+no%3Amilestone

mdjp: we have the usual bunch of issues to go through.

<mdjp> https://github.com/WebAudio/web-audio-api/issues/756

rtoy: this is a broad editorial thing.

mdjp: let's drop this right into v1

<mdjp> https://github.com/WebAudio/web-audio-api/issues/749

rtoy: this is a comment from Philip re SpatialPannerNode. Could we just merge the new spatial features into the existing PannerNode?

mdjp: so we'd no longer have SpatialPannerNode

clilley: I thought we wanted to separate out the "normal" L/R panning from other features

rtoy: That's the StereoPannerNode. This is about SpatialPannerNode.

<padenot> I'm back

BillHofmann: the defaults were wrong and confusing in the old PannerNode

rtoy: the fact that the X/Y/Z coords of everything are now AudioParams in the new node are significant

BillHofmann: maybe this is archaeology but if we thought that the PannerNode was hard to use, did we also make a mistake in SpatialPannerNode?

rtoy: I don't remember why we broke SpatialPannerNode out. But there were no AudioParams in PannerNode
... I don't remember why we made this change except that automation is now possible.
... we changed the deafult to eq power a long time ago

BillHofmann: so does PN still have the nonautomatable parameters?

rtoy: I think for backwards compatibility we would need to keep the old position setting methods

BillHofmann: it seems like we decided to deprecate PN, now are we asking whether to un-deprecate?

rtoy: yes

BillHofmann: we're not required to keep compatibiliy right now. Do we think this is a major impact?

rtoy: to do what?

BillHofmann: to deprecate PN. It's not marked as deprecated yet

rtoy: we could do that but because they're so similar, do we really have to break compatibility?

BillHofmann: if we meant to deprecate and didn't that sounds like a mistake. Did we strongly feel we should do that?

rtoy: yes we did want to deprecate. but we forgot to mark PN as deprecated. Now that we look at SpatialPN though it's really just become an "enhanced" PN

joe: if we didn't realize taht SpatialPN was an enhancement of PN, why not reverse a bad decision

mdjp: this buys us some compatiblity by not deprecating something
... I don't see an issue with this and I think I'm in favor

BillHofmann: what happens if people use old-school setPosition() in conjunctoin with AudioParams

padenot: we could define setPosition() in terms of actions on AudioParams to unify the definitions

BillHofmann: yes, that does seem a bit clunky

rtoy: we could add a deprecation msg to the console
... how about everyone post their comments to the bug

<mdjp> https://github.com/WebAudio/web-audio-api/issues/748

<mdjp> https://github.com/WebAudio/web-audio-api/issues/741

rtoy: because we're automating SpatialPN now, you would need to break up automation into X/Y/Z changes to do something like spin around
... so using spherical coordinates makes this much easier

BillHofmann: at Dolby we think about this in terms of whether you're doing user-based changes or environment-centric
... so there's no single way to do this that makes "the most sense"

rtoy: I'm fine with closing this as just an idea. One can make things reasonably smooth by piecewise automation of Cartesian coords

mdjp: how does this work in other common libraries? do they mostly use Cartesian?

rtoy: yes they do
... we can close this w/ no action

mdjp: that would be my feeling

_1

+1

<BillHofmann> ChrisL - I can never remember it :)

<mdjp> https://github.com/WebAudio/web-audio-api/issues/740

mdjp: makes sense to do this simple ed. change

<mdjp> https://github.com/WebAudio/web-audio-api/issues/739

rtoy: not sure what to do here. ChrisW was concerned that on mobile in particular the delay in loading HRTF database was pretty noticeable an dbad

BillHofmann: I love the idea of being able to load one's own

mdjp: there were a few requests on this. some came from my team a long time ago.
... we need to do a bit of digging to see if there's a ticket anywhere
... what I would suggest for this is bring it up again with Chris and I'll look and see if I can find any reference to our agreeing to do this

<mdjp> https://github.com/WebAudio/web-audio-api/issues/737

rtoy: this basic term isn't defined anywhere

joe: isn't "block size" actually defined?

padenot: we can link to some text I added in the Processing Model

rtoy: I'll wait for the Proc Model to land and we can take it from there

<mdjp> https://github.com/WebAudio/web-audio-api/issues/730

ok this sounds like something that just needs to be done. what are the compatibiliy issues here?

padenot: it's got a patch, I'll have a look in gecko

mdjp: shall I just assign this to you Paul?

padenot: yes thanks

<mdjp> https://github.com/WebAudio/web-audio-api/issues/729

padenot: this is not like a state that is stable.
... if you are in the same ... wait I think I got confused ...
... I'll follow up on this

<mdjp> https://github.com/WebAudio/web-audio-api/issues/713

rtoy: this is about scheduling param automation events in the past

mdjp: didn't we discuss this last call?

rtoy: I don't remember that

joe: I dno't remember discussing this
... for congruence w/ rest of API, clamping to currentTime feels like the right thing to do

rtoy: yeah, we could do that

mdjp: that just leaves the many "property bag" issues

joe: I've held off on my ctor PR because of all of these dangling questions about ctor signatures and property bags

mdjp: I think that's it for item 2

<mdjp> https://github.com/WebAudio/web-audio-api/issues/673

rtoy: we need to look at nominal ranges for all the AudioParams to expose the min/max r/o attributes

<mdjp> https://github.com/WebAudio/web-audio-api/issues/667

rtoy: one other item from last week was the playback latency category-or-number issue

<rtoyg_m> https://github.com/WebAudio/web-audio-api/issues/348

mdjp: question still seems to be about numbers vs categories

rtoy: original proposal used categories, some wanted more fine grained control

joe: is there a precedent in Web APIs for signatures that allow String or numbers?

rtoy: I don't know but WebIDL does let you allow this sort of thing

<rtoyg_m> https://github.com/WebAudio/web-audio-api/issues/348#issuecomment-190341633

joe: nice to be able to specify both but need to be able to introspect what the UA actually gave you

padenot: don't we have this on the dest node

joe: that is proposed to be latency in ms not the batch size

mdjp: we need to come back to this

rtoy: I'll propose something on the bug and folks can comment

mdjp: we're out of time. one more meeting before F2F

Next Meeting

mdjp: next call is 24 March

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.144 (CVS log)
$Date: 2016/03/10 17:59:47 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.144  of Date: 2015/11/17 08:39:34  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Found ScribeNick: joe
Inferring Scribes: joe
Default Present: padenot, ChrisL, joe, hongchan, rtoyg_m, jdsmith, cwilso, mdjp
Present: padenot ChrisL joe hongchan rtoyg_m jdsmith cwilso mdjp BillHofmann ChrisL_

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 10 Mar 2016
Guessing minutes URL: http://www.w3.org/2016/03/10-audio-minutes.html
People with action items: 

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.


[End of scribe.perl diagnostic output]