Subject: Re: [linux-audio-dev] Presentation & New project.
From: Tim Orford (tim_AT_orford.org)
Date: Tue Jun 03 2003 - 01:12:11 EEST
On Mon, Jun 02, 2003 at 06:18:31PM +0100, ListBox wrote:
> Hi to everybody...
hi and welcome to the list!
> I'd like to introduce myself to you and to present you my new project.
> I'm a 26 year old programmer living on the Canary Islands. Unfortunately I'm
> new to Linux and I've been specializing on Pascar rather than on C. I'm a
like you i have more Pascal(Delphi) experience than C. You didnt say
whether you wanted to continue using Pascal, but given that any Pascal
program would really be a bit of a black sheep on Linux, i have
personlly decided to forget it, despite some of the attractions of
Kylix.
> http://laudio.freehomepage.com/index.html (my page) where I describe the
> project in detail. Comments are welcome.
it looks good, but overall i would say that it was too ambitious
and not flexible enough.
>Well, I have been trying Muse and Ardour as well as Rosegarden. These
>programs are impressioning but they all lack a certain level of
>interconnectability.
this is true and is currently being worked on. The authors are
currently discussing this on the jackit-devel list.
>it has to be possible to sychronize various different PCs, all of them
>connected via LAN or midi.
i believe you are right that interconnectivity is the key. Any serious
no-holds-barred music production is going to max out a single
machine (there is a tendency for dsp usage to expand to make use
of your processing power), and only the hardware available to an
open-source project is another pc. Midi is
too slow for most things, but ethernet and firewire hold much
promise. It would be great if you could for example write a jack
client that could communicate via these and other high latency
protocols. I dont have any references handy right now, but there are
moves afoot to do this. Another option is to take the Steinberg approach
and use digital audio channels.
If you could acheive this one (difficult) goal even, that would be totally killer!
>Let me describe my vision of the different workstations:
i'm not sure if the setup you described was just an example or
a fixed configuration, but i beleive it would be important
to allow flexibility here so that for example:
- all of the apps could run on a single machine, or
- each function could be split over multiple machines.
A major undertaking such as you have outlined, needs the
support of as many people as possible, so you should
cater for the needs of everyone from the casual hobbyist to
the hardened pro's. Even if they dont realise it, their needs are
broadly the same....
>SMPTE Clock / Midi Sequencer
why separate the Midi from audio? Dont you agree that
it is easier to work on a composition with a unified GUI?
Having said that separate audio/midi is the current state
of affairs, due to ease of development. My ultimate dream
is to have an uber arrangement client which allows the user
to control disparate clients via a single interface...
Most of your infrastructure requirements are probably already being met
by Jack and Alsa, which i beleive are *excellent* starting
points. Surely it would be easier to add the
missing parts rather than write from scratch? I will admit
that documentation is a problem with Alsa, and to a lesser
extent Jack. I am still not sure what the true extent of their
abilities is.
currently, my personal requirements can probably be met by 3 processors.
Only one machine would have a graphics card, and would control
the other machines using X :-)
total cost: 400+400+500+300+300+300 = 2200 euros/dollars
anyway, just some quick thoughts :-)
whichever way you decide to go, i wish you the best of luck!
-- Tim Orford
This archive was generated by hypermail 2b28 : Tue Jun 03 2003 - 01:42:14 EEST