[vorbis] Auto Vorbis

[Firelight Multimedia]

Im quite interested in this too for pocketpc.
Someone on vorbisdev (are you on that list?) posted the following..

"

I'm currently working on converting libvorbis to fixed point arithmetic.
I've managed to get all the decoding after the MDCT reverse
transformation
into 24-bit fixed point, including the windowing.


--
Russell King (rmk at arm.linux.org.uk)                The developer of ARM
Linux
             http://www.arm.linux.org.uk/personal/aboutme.html
"

> -----Original Message-----
> From: owner-vorbis at xiph.org [mailto:owner-vorbis at xiph.org] On Behalf
Of
> josh
> Sent: Thursday, 11 October 2001 8:10 AM
> To: vorbis at xiph.org
> Subject: [vorbis] Auto Vorbis
> 
> Is anybody going to do something about the fact that vorbis decoder
won't
> run
> on ARM processors (iPaq's, etc) or any (that I know of) car music
players?
> How can vorbis expect to compete when one of the most popular uses of
MP3
> is
> in the car??   Though I would love to switch to vorbis, I am forced to
> stick
> with MP3 because of this.
> 
> I emailed a previous poster here who appeared to be working on a port
to
> ARM,
> but his response was:
> 
> "Unfortunately work has halted. It doesn't look like it will ever be
fast
> enough with floating point emulation. This puts the effort required to
> port
> much higher than the time I can affort to give to it.
> 
> Still if any thing does happen I will be sure to let you know.
> 
> Ian."
> 
> SO, I tried the brain behind the open source MP3 player at
> www.pjrc.com/tech/mp3/ , to see about the possibility of adding vorbis
> support
> to the project, but he had this to say:
> 
> "I just looked at the xiph.org pages, and there's no mention of a
hardware
> decoder. All of the important parts... in fact everything except the
> libvorbisfile library, is labeled as "Beta release". Those libraries
> claim to have stable APIs, and it looks like the bitstream format is
> now stable... but the encoder doesn't make use of several features of
> the bitstream, so there's no real way to test a 3rd party (or even the
> official 1.0rc2) vorbis implementation for correct operation with
> future vorbis streams.
> 
> Worse yet, it appears that everything is based on floating point
calcs,
> so either a floating point DSP chip or considerable work to translate
> the algorithms to integer-only would be needed."
> 
> Alas, it seems vorbis falls short of what I need.
> 
> -josh
> josh at alumni.southern.edu
> 
> 
> --- >8 ----
> List archives:  http://www.xiph.org/archives/
> Ogg project homepage: http://www.xiph.org/ogg/
> To unsubscribe from this list, send a message to
'vorbis-request at xiph.org'
> containing only the word 'unsubscribe' in the body.  No subject is
needed.
> Unsubscribe messages sent to the list will be ignored/filtered.

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.