🚀 go-pugleaf

(I can't send you a reply, either...  This should be fun :-)

>Brian -- I'm posting this reply due to a "diode" site in your path.
>Please get back to me on this one.  Thanks, mike k.

That's odd...  I thought "@bnr.ca" would work anywhere.  Either way, this
account will be erased over the hollidays.  After that, only my SunEE
account (bcwhite@sunee.uwaterloo.ca) will be active.

>Brian -- I would like to take a crack at this.  I did some good DECB work on
>this for the Coco3, and have given out my program to record and play back sound
>on the MM/1.  And I've done UME on both machines.

I'm in the midst of packing, but I should be able to get a copy of everything
out before I go.

>I looked into the 68000 architecture enuf to decide it wasn't really as good
>as an 8-bitter for playing sampled waveforms, but you seem to have found some
>tricks I didn't know about, or maybe the CPU is just so fast that it doesn't
>matter.  ANyway I'd love to see what you got...

The only thing is that you can't do any software timing.  I just play into
a memory buffer and then issue a DMA request when done.  The next buffer is
calculated while that was is playing.

Actually, there are three buffers.  One is playing, one is queued, and the other
is being created.  If the one playing ends before the one being created is
finished, then a "stop" flag is set so the music loop can finish up and return.
If it falls too far behind (i.e. consistantly being interrupted and thus the
play buffer getting smaller and smaller) then one of the parameters is tweaked
te reduce the sampling rate.  The same buffer size then takes longer to play
so the music loop can produce more samples.  If the music loop is constantly
finishing significantly before the dma, then the sampling rate can be increased.
Of course, this is all theory.  The mechanics are in place but untested.


>Are you thinking money or more like PD here?  I'd be willing to release anythin
 g
>I did as freeware, besides integrating some of it into UME.

Well, if sold, I wouldn't mind a small cut.  If it's free, then all I ask for
is some of the credit.  Any cause is better than it dying on my hard disk.  I'd
also appreciate being consulted about changes to the current work simply
because I know it fairly well and feel I could provide a lot of helpful
insight into making it better.

Oh, I even have some converters started.  I have a "Bells & Whistles 2"
converter working and an orch-90 converter started.

The music score format is very detailed and very documented, though all
it's features would not be necessary in a first revision player!

These opinions are all                      Brian                 _ /  __/  _ /
mine and you can't have               ( studer@bnr.ca )          / / __ / __ /
any.  So there!                 ( bcwhite@sunee.uwaterloo.ca ) ___/ ___/ ___/
-------------------------------------------------------------------------------
 The difference between theory and practice is less in theory than in practice.