See also: IRC log
<trackbot> Date: 09 January 2012
<scribe> Scribe: Olivier
[meeting convenes]
Al: at earlier meeting we decided the best way to divide up the work was through the use case and requirements document
<Alistair> http://www.w3.org/2011/audio/wiki/Use-Cases
<quinnirill> hmm, I can't join, it says the call is full
<jon> same
<jon> hi joe
[meeting reconvened on another teleconference port]
<roc> er
<roc> what was the number?
<shepazu> oops
<quinnirill> Zakim: +??P8 is me
<joeberkovitz> zakim: 1.978.314.aaaa is me
<joeberkovitz> zakim: +1.978.314.aaaa is me
joe: will do a bulk transfer of known de facto requirements of our (music) world onto this document
… will drag in some aspects of UC6
… may end up talking about envelopes, samples etc in UC5
Olivier: I created the UC skeleton arbitrarily, we may end up merging some
… mostly wanting to see if any is missing
Al: is accessibility a use case?
Chris: think that's a cross-UC requirement
???: how about visualisation
… sound, music visualisation
Doug: song recognition (a la Shazam, etc)
… could that be done?
Chris: mostly happens on the server side
Al: interesting is the question of microphone input
<scribe> ACTION: Doug to find status of microphone input work [recorded in http://www.w3.org/2012/01/09-audio-minutes.html#action01]
<trackbot> Created ACTION-10 - Find status of microphone input work [on Doug Schepers - due 2012-01-16].
Chris: I actually don't like "microphone" input, can be any kind of line input
Doug: other potential UC - was wondering about background sound
Chris: like a <bgsound> tag?
Doug: was wondering if media context for a given tab has to be the same as the dom context. What if sound continued from one page to another
… without having to reload all the resources
Chris: no idea how that would work re- security or implementation
… there is a document associated with the script execution context
Al: LocalStorage is shared
<Zakim> shepazu, you wanted to ask a strange use-case question
Doug: sound effects for clicking a button?
… UI sounds
Olivier: somehow already included in the game UC, but could add one
Chris: it's interesting as a UC in itself - interaction with CSS
Matthew: can we combine UC3 (mixing) and UC6 (playing virtual instruments)
Chris: may be useful to keep them separate - high level vs low level
Suggested: change UC3 to "production tool" instead of "mixing tool"
<joeberkovitz> olivier: i was the speaker at 16:26 and 16:27
Al: would like to know who's going to be working on which for the week to come
Joe: will take 5, probably going into 6
Chris: can take on 3, maybe not this week
Al: can look into
visualisation
... should we give ourselves 2 weeks?
Olivier: sounds reasonable, the earlier the better
Chris: speaking of UC1 - anyone familiar with chat in 2nd life? seems like a fairly sophisticated example
… something we'd want to replicate in HTML
Joe: how do you abstract that?
Chris: 3D world where voices are spatialized
… how to interact with e.g webgl
Olivier: was wondering if people have requirements for latency, and ideas about testability
Chris: it is tricky to test, you could create some kind of mechanical rig (very complex to test)
… really hard to test, but it is really important (e.g in games)
… keypress to sound latency in games is going to be really important, equivalent to frame-rate with graphics
… if we want to take on media and game applications into the web, latency is a key question
Al: this issue crosses in to how many sounds you can play at once
… (and drop rate)
… agree with Chris that we need something in our requirements
… ditto for music
Chris: hard to achieve a hard latency limit. Maybe have some guidance
… can be informative
Chris: in smartphone world for instance there is a big hardware discrepancy in Android phones, for example
Doug: traditionally steer clear of performance stuff at w3c
… it's not usually a question of conformance
doug: we should optimise in the spec as soon as possible
Chris: relevant to the discussion of js only versus natively
… analogy with CSS 3D animations
… expectation that they will be smooth, and that's where implementations compete
… and no mention of frame rate in the spec
doug: could be, down the road if it became an issue
Chris: continuing on the analogy, there is nothing in CSS 3D animation impeding making it smooth / hardware accelerated
doug: [notes he's looking for hardware manufacturers to join in and help us spec things that can be accelerated easily/well]
Joe: possible to tolerate quite a high level of latency
… but it's important to know what the latency
… especially to coordinate with visual delivery
… can't coordinate if you don't know the delay
… a requirement out of UC5 will certainly be to know what the latency is
Chris: agreed
Joe: in our flash application we ask for a couple hundred milliseconds, which would be terrible for a game
Chris: 20, 30 or 50ms for a game would be acceptable, depending on platform
Al: wanted to note there will be room for competition in implementation
<Alistair> A conforming specification MUST be able to play pre-loaded sounds back at faster than a rate of 'x' milliseconds from the time the JavaScript code was executed, until the time that sound is heard through the speakers: where the audio path is not running through any external sound devices and the browser is using the default sound driver provided by the operating system.
Al: asking for opinions about that blurb
Chris: that would be setting a hard limit
… was quoted earlier as saying 50ms would be acceptable… 10ms is better!
… platforms are getting better
… for rhythm based games you really do want better than 20ms
Al: around 30ms seems to be threshold
Chris: jitter impacts as well
…you will feel the 30ms if the jitter is high
doug: consistency sometimes more important than latency
al: and probably harder to achieve than low latency
chris: agrees
Joe: proposal to try to put in what the requirements should be for jitter and latency
… and then survey and see if they deserve a special place in the spec
doug: another possible
criteria
... need to think about what criteria we are going to be
judging what we put in the spec
olivier: would it make sense to have a list of design principles such as hardware acceleration capability, low latency and jitter, consistency of interfaces, etc
chris: my goal is to achieve similar level of performance as a native app
olivier: not sure we can achieve same level of performance as pro software on desktop, but agree with goal in general
Joe: we have flash apps/games as a benchmark
chris: will vary between platforms
… for games we can get there
… for digital audio workstations, will be tricky
[general agreement]
<scribe> ACTION: olivier to send details of f2f logistics [recorded in http://www.w3.org/2012/01/09-audio-minutes.html#action02]
<trackbot> Created ACTION-11 - Send details of f2f logistics [on Olivier Thereaux - due 2012-01-16].
Olivier: invite all to discuss ideas for f2f agenda on the mailing-list
[adjourned]
<scribe> ACTION: joe to work on Use Case scenario 5 [recorded in http://www.w3.org/2012/01/09-audio-minutes.html#action03]
<trackbot> Created ACTION-12 - Work on Use Case scenario 5 [on Joe Berkovitz - due 2012-01-16].
<scribe> ACTION: olivier to copy Use Case scenario 1 from RTC document [recorded in http://www.w3.org/2012/01/09-audio-minutes.html#action04]
<trackbot> Created ACTION-13 - Copy Use Case scenario 1 from RTC document [on Olivier Thereaux - due 2012-01-16].
<scribe> ACTION: Alistair to work on Use Case scenario for visualisation [recorded in http://www.w3.org/2012/01/09-audio-minutes.html#action05]
<trackbot> Created ACTION-14 - Work on Use Case scenario for visualisation [on Alistair MacDonald - due 2012-01-16].
<scribe> ACTION: ChrisR to work on Use Case Scenario 3 (music production) [recorded in http://www.w3.org/2012/01/09-audio-minutes.html#action06]
<trackbot> Sorry, couldn't find user - ChrisR
<scribe> ACTION: crogers to work on Use Case Scenario 3 (music production) [recorded in http://www.w3.org/2012/01/09-audio-minutes.html#action07]
<trackbot> Created ACTION-15 - Work on Use Case Scenario 3 (music production) [on Chris Rogers - due 2012-01-16].
This is scribe.perl Revision: 1.136 of Date: 2011/05/12 12:01:43 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Found Scribe: Olivier Inferring ScribeNick: olivier Default Present: [IPcaller], +1.978.314.aaaa, +1.650.253.aabb, Doug_Schepers, quinnirill, olivier, joe, chris Present: [IPcaller] +1.978.314.aaaa +1.650.253.aabb Doug_Schepers quinnirill olivier joe chris Found Date: 09 Jan 2012 Guessing minutes URL: http://www.w3.org/2012/01/09-audio-minutes.html People with action items: alistair chrisr crogers doug joe olivier WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option.[End of scribe.perl diagnostic output]