[whatwg] need a way to set output format from StreamRecorder
Kevin Marks
kevinmarks at gmail.com
Tue Feb 15 00:23:00 PST 2011
On Mon, Feb 14, 2011 at 11:52 PM, Nils Dagsson Moskopp <
nils at dieweltistgarnichtso.net> wrote:
> Kevin Marks <kevinmarks at gmail.com> schrieb am Mon, 14 Feb 2011 22:33:13
> -0800:
>
> > On Mon, Feb 14, 2011 at 2:39 PM, Ian Hickson <ian at hixie.ch> wrote:
> >
> > > […]
> > >
> > > I haven't added anything here yet, mostly because I've no idea what
> > > to add. The ideal situation here is that we have one codec that
> > > everyone can read and write and so don't need anything, but that
> > > may be hopelessly optimistic.
> >
> >
> > That isn't the ideal, as it locks us into the current state of the art
> > forever. The ideal is to enable multiple codecs +formats that can be
> > swapped out over time.
>
> Yeah, because that really worked well with <object>, the generic
> container element. Only it didn't, and with today's media elements
> people are venting stuff like “I had to encode all my sound files in
> Ogg Vorbis and MP3, just because of you, Safari. You make my life
> unnecessarily difficult.”
>
> — <http://www.phoboslab.org/log/2010/09/biolab-disaster>
>
> As someone who sometimes produces audio (and may want to use of
> browser-provided facilities, once they become available), I may have a
> reasonable interest in interoperability between browsers.
>
To be fair to Safari it supports far more audio formats than other browsers,
as it incorporates QuickTime's engine, which was designed to cope with
multiple audio, video and file formats via well designed abstractions, and
the Component Manager, which has lasted since the late 1980s itself (first
public release of QuickTime was 1991, but the codebase goes back pre-1990).
>
> (The Enrichment Center once again reminds you that codec hell is a real
> place where your web application will be sent at the first sign of
> defiance.)
>
> On further thought, “state of the art lock-in” may not be as bad as you
> might fear: First, bandwidth and storage space are becoming cheaper
> over time; second, there is something as “good enough”, GZIP / DEFLATE
> or MP3 being such examples that serve us for over 15 years each — even
> though better specifications (Vorbis, 7z) clearly exist.
>
> Yes, MP3 is the de facto standard, and somehow Mozilla and Webkit Android
still won't play them. (Bizarrely, Webkit Android 'supports' HTML5 <audio>
without supporting any codecs or file formats at all).
> Even CPIO is used by my modern desktop system and that was defined
> around the mid-80ies (or so Wikipedia tells me, I certainly wasn't
> released at that point).
>
and AIFF, MOV, WAV, AVI and MP4 are all based on IFF
http://en.wikipedia.org/wiki/Interchange_File_Format - a future-proof binary
chunk format form the mid 80s too.
Supporting playback of uncompressed audio (and uLaw, aLaw, PCM) in .au .aif
.wav and .mov should be trivial, and not encumbered by any patents. Picking
one to record in by default should be something we could agree on - which is
most widely supported at the moment? WAV?
More information about the whatwg
mailing list