Subject: Re: [linux-audio-dev] EVO spec 0.0.1, linuxsampler 0.0.5 released ...
From: Richard A. Smith (rsmith_AT_bitworks.com)
Date: Thu Jul 20 2000 - 01:46:06 EEST
On Wed, 19 Jul 2000 17:17:05 +0200, Benno Senoner wrote:
>On Wed, 19 Jul 2000, Richard Smith wrote:
>> Last Friday I was over at Mike Bailey's house discussing what EVO's feature set needs to
>> be.
>
>BTW: is Mike reading linux-audio-dev ?
>(a mailinglist dedicated to EVO would be useful (not yet), in order to avoid
>boring LAD subscribers too much)
Yeah the posts from "AudioNexus" and "larry d byrd" are his.
I will try to set up the sourceforge account sometime next week while I am on vacation.
>But beforestarting coding on the GUI you have to define the specs of both
>the engine and the communication protocol in every detail.
Right but some of the engine features and comm protocol depend on what you want to do with
them which is why I posted the spec.
I am going to try wvHtml tonight and see how it does on the word docs.
>I will focus my efforts mainly on the engine part, and leaving the GUI issues
>to others.
>Another point could be that EVO could be the ideal testbed for
>the much criticized rtsoundserver model.
>(sooner or later we will need it to run a midi sequencer/harddisk recorder
>in parallel to EVO which need to be perfectly in sync)
Works for me.. For Mikes application a dedicated soundserver is no big deal.
>- are the algorithms/math publicy available ?
> (if yes, URLs with docs would be handy)
Mike found a webpage or 2 with some algors on them but I don't know if the algorithms are
non-patented.
>- will the realtime implementation be fast enough to still
> allow 20-30 voices on PII CPUs ?
I am not really that concerned about this. Because by the time we get this cranked up IA64
will be rolling out to the mass market. In the very near future we will be able to get a
whole lotta horses for not too much $$. I would much rather have software waiting on
hardware to run it rather than the otherway around.
Plus there are always optimizations and so forth that can be make once a stable code base
is reached. Perhaps we have to use lots of assembly or MMX and 3DNow to get to the
performance level desired.
I wouldn't let that deter us from trying to hash out a framework that will allow 20-30
voices plus all these nifty effects.
-- Richard A. Smith Bitworks, Inc. rsmith_AT_bitworks.com 501.846.5777 Sr. Design Engineer http://www.bitworks.com
This archive was generated by hypermail 2b28 : Thu Jul 20 2000 - 02:17:57 EEST