additive

[McIntosh, Malcolm]

Yes, it sounds very difficult.  But I was thinking of an analog module
which uses additive synthesis theory in a practical way to create new
sounds and timbers.  Not imitative at all.  I think you would have to
limit the number of envelope generators to a reasonable number like 8 or
16.  Maybe you could use 2 or 3 VOCs to feed 8 or 16 frequency
multipliers each with a VCA and ADSR.  Whew!  Maybe by de-tuning the
VCO's just a little you might avoid the problem that Bill Layer
mentioned:

<SNIP>
>I think the limitation here is that you would be limited to natural
>harmonics; most complex sounds contain much enharmonic information as
well.
>You are right in that freq. multiplication could provide a palette of
>harmonics...
>
>Perhaps if one employed eight oscillators, set to different microtonal
>increments, and then created a palette of harmonics from each of those.
<SNIP>

I don't know.  I haven't finished or supplying my prototyping Lab with
enough stuff to try and I think that's the only way to see if the idea
has merit.  Combine the above scheme with filters and waveshapers and it
just seems like it may sound, well, very interesting (or as Spock would
say, "Fascinating Doctor").

Malcolm

PS  "thnaks" again Bill

> -----Original Message-----
> From:	Steven Curtin [SMTP:sdcurtin at lucent.com]
> Sent:	Thursday, August 13, 1998 2:40 PM
> To:	McIntosh, Malcolm
> Subject:	RE: additive
> 
> At 08:32 AM 8/13/98 -0600, you wrote:
> >Hi,
> >
> >This is a subject I have been thinking about with no idea how to
> >implement.
> >
> >Would it be possible to use frequency multiplication and one really
> good
> >sine wave VCO for limited additive synthesis?  Would it be necessary
> to
> >have 64 or 128 harmonics if you just wanted to make some really nice
> >sounds but not be totally imitative?  Could you get away with 16 or
> >maybe 8 (!) harmonics to make really interesting non-imitative
> sounds?
> 
> There's been some research done on this subject by James Beauchamp and
> Andrew Horner- they gave a paper at an International Computer Music
> Conference a number of years ago.  It's theoretically possible but
> rather
> hairy technically, since they used genetic algorithms for spectrum
> matching
> between different c:m ratios and the expected spectrum, etc.  Additive
> is
> such a control monster that it's hard to build an interface to it
> that's as
> elegant as modular analog synths, which is why it's been hard to get
> lots
> of people to buy into using it.
> 
> Steve C
> 
> --------------------------------------------------------------
> Steven Curtin  
> http://www.emf.org/people_curtin.html
> Lucent Technologies - Bell Labs Innovations
> rm. 1A-229, 200 Laurel Ave S
> Middletown, NJ 07748-4801  U S A
> ph: (732)957-2996   fax: (732)957-6878
> --------------------------------------------------------------