On 24 f=E9v, 16:19, "Jeff Blakeney"  wrote:
> =A0 To: Michael J. Mahon
>
> On Mon, 23 Feb 2009 17:33:35 -0800, Michael J. Mahon wrote:
> > Jeff Blakeney wrote:
> >> But the Ensoniq is able to deal with 128 KB of RAM connected to it. =
=A0I'm
> >> surprised no one has done a hardware hack to piggyback the RAM and
> >> double the size of the sound buffer. =A0Imagine the MIDISynth wave ban=
ks
> >> that could be made with twice as much room for samples. =A0:)
>
> > I wonder if the toolset allows for 17-bit addresses, or the equivalent
> > bank-selection bit...
>
> I'm not sure if the toolsets allow it but it might be possible to patch
> them to do so. =A0I just did a quick look online and in the IIgs Hardware
> Reference and it looks like the 5503 only has 16 address lines but does
> have a Bank Select on pin 40. =A0In the IIgs schematics, pin 40 looks lik=
e
> it isn't connected. =A0The only way to set that pin's state is through th=
e
> Wavetable Size Register where bit 6 is always zero on the IIgs at
> present because it doesn't have another bank of memory to deal with.
>
> The problem is, like you implied above, putting the samples into memory.
> I'm not exactly sure how the Sound GLU, auto incrementing address bit
> and the bank select bit in the wavetable size register will all work
> together. =A0There might need to be more done to be able to put samples
> into the high memory bank.

If one of you is able to perform the hardware hack, I will update the
sound/music toolsets of which I have my own source code.

antoine
--- SBBSecho 2.12-Win32