- From: Matthew Paradis <matthew.paradis@bbc.co.uk>
- Date: Thu, 8 Jan 2015 20:24:16 +0000
- To: Audio WG <public-audio@w3.org>
- Message-ID: <D0D497ED.1CC5D%matthew.paradis@bbc.co.uk>
Dear all, Please find below the minutes of our teleconference on 2015-01-08: http://www.w3.org/2015/01/08-audio-minutes.html Summary 1. Proposal for separate multiple-I/O AudioWorker API and single-I/O AudioWorker API (action from previous meeting) ChrisW talked us through the proposal for the audio worker http://cwilso.github.io/web-audio-api/#h3_the-audioworker and explained the new input/output definition. General concensus was that this was a good solution but we agreed to post this to the list for discussion and evaluation before the next meeting. ChrisW will try to produce a proposal for the factory model by the next telecon if his time allows. ACTION – Start a discussion thread on this list. 2. Audio Worker "bypass" Attribute #456 (In the "needs WG Decision" queue) Unanimous agreement to move this issue to V.Next 3. Inter-app audio (Consistent request from ISVs, a la VSTs, Audio Units, Rack Effects) #358 (In the "needs WG Decision" queue) General concensus that this issue represents more fundamental questions around the api and the low level interaction with audio i/o and device enumeration. It has been suggested that we review use cases and split the problem away from just being a feature to allow interoperability with VST and 3rd party code but rather to concentrate effort on more fundamental issues. If this does not make it to V1 we still need to be sure that it can be supported in the future. 4. AOB As time had run short we skipped other business. For the record requests were made to revisit 2 issues; rtoy - "Can we get clarification on issue 419 about biquad filter formulas? The currently implemented filters don't match the cookbook formulas. Since Chrome, Firefox, and Safarai all implement the same set of formulas, we may consider dropping the cookbook formulas and just use the actual implementation." rtoy - "In issue 455, we discuss removing the doppler effect from the PannerNode. Since the PannerNode is going to be replaced with new panner nodes, would it make sense to leave the old PannerNode as is (with doppler) for backward compatibility? When PannerNode is finally removed, the doppler stuff will go away at the same time." These items will be added to the agenda for the next meeting. 5. GitHub Issues to mailing list Doug commented that we no longer get notifications from github on the mailing list and will get it reinstated. 6. Next meeting 22nd January noon Boston time. Best Matt -- Matthew Paradis Senior Software Engineer (Audio), BBC Research & Development 030304 09889 | matthew.paradis@bbc.co.uk | www.bbc.co.uk/rd/sound
Received on Thursday, 8 January 2015 20:24:47 UTC