home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Computer Club Elmshorn Atari PD
/
CCE_PD.iso
/
mac
/
0500
/
CCE_0502.ZIP
/
CCE_0502.PD
/
DSP_ASM
/
README.EFF
< prev
next >
Wrap
Text File
|
1992-11-14
|
2KB
|
44 lines
From vn Sat Aug 3 22:10:06 1991
~Subject: Stereo Chorus for DSP56001 - code sample
~Newsgroups: comp.dsp,comp.music,rec.music.synth
OK, OK, OK. After I mentioned in passing one day that I'd been working on some
effects code for the 56001 I got quite a few queries. Well, here's my story.
I have working on my 56K board reverb, chorus, flange and combinations of them.
The chorus and flange I came up with after playing with and analyzing a
commercial chorus pedal. I looked at the impulse response and delay-time
characteristics of the box on a scope. I feel OK about releasing my chorus
and flange code. The chorus code is attached.
But, my hangup with the reverb is that I used the filter configuration and
coefficients from csound. I could probably give credit to Dr. Vercoe and
release the code, but while I am quite pleased with the performance of the
algorithm, I would like to tweak it and expand on it to better match it to
my sample rate. The 56001 has suds left over (I run at 25 MHz). I remember an
article I saw in the Computer Music Journal by Dr. Moorer (I think)
where he discussed filters for reverb. I was hoping to get a hold of his
paper again to attempt to tune the code. I think I need another comb filter
or two. The sound is a bit "grainy" on highs.
But, anyhow, here is some sample code for chorus. Chorus uses a single
delay-line with time-varying delay time. Flange is very similar, with
shorter delay and feedback, building effectively a time-varying comb
filter. Reverb (as implemented in csound) uses four comb filters and
then two all-pass filters. Commercial reverb makers are very protective
of their respective algorithms.
By the way, I do not use the Motorola assembler but one that I wrote using
YACC and LEX. It was harder than I thought and I don't like the way the
grammar turned out. But it does work. Would any of you DSP hacks like
a documentationless assembler for the 56001? I've been very tempted to
post it, too, but was afraid of taking on the support issues.
I guess I'll post my comment-lean flange code as well. The comments
in the chorus code will help.
Enjoy,
Quinn Jensen
work: jensenq@npd.novell.com
home: jensenq@qcj.icon.com