[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: New "Grand Piano"...
- Subject: Re: New "Grand Piano"...
- From: spector@vax2.concordia.ca (Mitchell Spector)
- Date: 1996/03/18
- Distribution: world
- News-software: VAX/VMS VNEWS 1.50AXP
- Newsgroups: comp.sys.apple2
- Organization: Concordia University
- References: <4i9uj4$e3e@nz12.rz.uni-karlsruhe.de>
- Sender: spector@vax2.concordia.ca (vax2.concordia.ca)
In article ug0a@rzstud1.rz.uni-karlsruhe.de (Alexander Siegfried) writes...
>
>Mitchell Spector wrote:
>
> >I see only two ways to bring the Apple IIgs up to par with
> >current PC wavetable sound boards. Either use what hardware is
> >available and find some way to make use of standard GS memory as
> >DOC-RAM (swap between 64Kb and system memory when playing MIDI
> >files). This should be possible with a very high-speed accelerator
> >and the GS doing nothing else but playing music, but programming
> >it would be quite a task. I'm not even sure if this is doable.
>
>Mitch, the technique you describe has been already implemented on
>the GS for *a couple of years*. What do you think are programs like
>Shellplay, MODZap or Beatbox doing to play those huge MOD files?
I'm talking about using this technique with sample patches
used for _MIDI_ playback. Apple's own synthLAB program (or rather
MIDIsynth it's based around) is the best tool for that, yet you
can only use one instrument bank at a time. A collection of at
the very most, 64K worth of instrument patches to use at once (and
to further limit things each instrument patch can be no larger than
32K, most being much smaller of course!).
And I'm certainly aware of MOD players swapping between
DOC-RAM and standard memory, FTA's Noisetracker GS was the first
program to do this back in 1991 (their "Delta Demo" was the first
taste of it actually :). Doing this for MIDI is a different
story.
>You don't even need an accelerator for the sample swapping, yet it
>eats over 50% of processor time...
What about sample swapping _and_ handling MIDI sequence
playback together, which is what I'm interested in seeing. Each
alone consumes lots of CPU %.
>Well, well, the 64k limitation: Another design flaw of the GS - blame
It certainly was in my opinion, especially since the Ensoniq
5503 can address up to 128K of DOC-RAM directly. Apple's using only
64K was as silly as shipping CPUs with only 256K of main memory at that
time. Yes DRAM prices were expensive back then, but certainly Apple could
have left two empty sockets to add an additional 64K of DOC-RAM for users
(which at a later point a full 128K could have been included as standard as
prices fell).
>-- Clue / Ninjaforce
Mitchell Spector
spector@vax2.concordia.ca