W3C home > Mailing lists > Public > public-xg-audio@w3.org > December 2010

Re: Making Web Audio APIs More Competitive with native

From: Joseph Berkovitz <joe@noteflight.com>
Date: Fri, 10 Dec 2010 11:23:02 -0500
Cc: srikumarks@gmail.com, "Noah Mendelsohn" <nrm@arcanedomain.com>, "Chris Rogers" <crogers@google.com>, public-xg-audio@w3.org, "Doug Schepers" <schepers@w3.org>
Message-Id: <68E3DED5-66D7-47AF-B525-4FF95D9D5DE2@noteflight.com>
To: jussi.kalliokoski@gmail.com
Hi all,

After quickly looking over the OpenSL ES spec (http://www.khronos.org/registry/sles/specs/OpenSL_ES_Specification_1.0.1.pdf 
), I had a couple of thoughts to share.

One is that OpenSL ES has a set of clearly defined use cases  
("profiles"). It may be premature to think about tracking OpenSL  
functionality until this group can settle on a clear set of target use  
cases; at that point we can evaluate how close these are to OpenSL  
ES.  In OpenSL ES the cases are coupled to the mobile device arena,  
but that doesn't make them invalid for the Audio XG/WG.  I found it  
interesting to look at the profile/feature table in the spec (see p.  
10) and compare with Chris's proposal.

Another observation is that while OpenSL ES is a rather low-level API  
focusing on mobile implementers working in C, it can provide a  
valuable feature checklist for this group. Even if we do not track  
their spec closely (and I suspect that this is not very doable because  
of how low-level it is) we can look at how they have tackled parallel  
issues and examine features to see if they're worth including on their  
own merit. For example, the resource-priority mechanism in OpenSL ES  
might be worth introducing at some point for browsers working in a  
setting with highly constrained resources.

Best,

..Joe

On Dec 10, 2010, at 6:26 AM, jussi.kalliokoski@gmail.com wrote:

> Hi,
>
> I think OpenSL is more of an implementation choice than one to do  
> with the web audio API, which provides a much higher level API than  
> OpenSL. (correct me if I'm wrong, my knowledge of OpenSL is limited)
>
> Best regards,
> Jussi Kalliokoski,
> Web Developer,
> Aldebaran
>
> Sent from my HTC
>
> ----- Reply message -----
> From: "Kumar" <srikumarks@gmail.com>
> To: "Noah Mendelsohn" <nrm@arcanedomain.com>, "Chris Rogers" <crogers@google.com 
> >, <public-xg-audio@w3.org>, "Doug Schepers" <schepers@w3.org>
> Subject: Making Web Audio APIs More Competitive with native
> Date: Fri, Dec 10, 2010 8:00 am
>
>
> Hi all,
>
> I joined the public forum relatively recently and
> I've been reading this thread with great interest.
> The work on the audio API thus far is impressive.
>
> That said, I couldn't help wonder whether some of
> the initiatives in other standards groups can be
> merged into this effort - particularly that of
> OpenSL ES (http://www.khronos.org/opensles/)
> That would make the audio path symmetric compared
> to what the visual group is doing with WebGL, which
> is taking OpenGL ES and making it available via
> a JavaScript API built atop Canvas 3D.
>
> Any thoughts on this? On first glance, there doesn't
> seem to be anything particularly limiting about
> OpenSL ES (to me at least) that would make it
> inappropriate for web use. I hope to get a better
> idea of it in the coming days.
>
> Anyone here who straddles both groups?
>
> Regards,
> -Srikumar
>
> On Fri, Dec 10, 2010 at 3:28 AM, Noah Mendelsohn  
> <nrm@arcanedomain.com> wrote:
> >
> >
> > On 12/9/2010 2:12 PM, Chris Rogers wrote:
> >>
> >> I don't expect that a JavaScript developer is going to play with  
> my API
> >> for
> >> a couple of weeks and come back with Digital Audio Workstation  
> software to
> >> rival something like Apple's Logic Audio
> >
> > Indeed, but if we set off down the path that would lead to such  
> things in,
> > say 5 years, with interesting DAWs of lesser capability emerging  
> along the
> > way, that would be really wonderful.  I'm actually curious what  
> will prove
> > to be implementable in Javascript over time from a performance  
> point of
> > view.  E.g. it's very cool that Javascript is doing FFts with  
> plausible
> > performance today, but it will be interesting to see when it can  
> do 20 in
> > parallel on a many core chip.  Still, this all looks very, very  
> promising.
> >  Thanks!
> >
> > Noah
> >
> >
> >
>
>

... .  .    .       Joe

Joe Berkovitz
President
Noteflight LLC
84 Hamilton St, Cambridge, MA 02139
phone: +1 978 314 6271
www.noteflight.com







Received on Friday, 10 December 2010 16:23:39 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:37:59 UTC