Pedro Batista wrote:
> 
> If I have a wav file of the sound I want to reach, I can use a HETRO/adsyn
> combination to recreate it in the context of a csound instr, but its sort of
> like cheating, if you know what I mean. Cant those utilities be used to
> retrieve info from the soundwave itself, that would allow us to recreate it
> using just sound generation/shaping opcodes?

Cheating? If sound analysis/resynthesis in general is cheating, then 
surely hetro is cheating. Using hetro is cheating from a purist 
synthesician's viewpoint (which I'd like to take, but it's too frustrating), 
but the info it uses is certainly derived from the sound itself. 

What you are looking for is perhaps to be able to tap into that info
with more detailed control, to reshape the material. HETRO data is 
as you may know a set of time/pitch and time/amp lines for a bank 
of oscillators, and you can only use them wholesale in adsyn. 

But you can do several analyses of the sound: like 5-10 files, 
each with just 10 harmonics, analysis starting at different pitches,
like 50, 150, 250, 450 etc. This will make the spectrum denser,
so you lose less from the analysis, and you can use these as
different layers which are resynthesized at different speed/pitch.
(Or indeed, 80 files with just one harmonic. Must try that now.)

Also, do try pvoc. With the newer opcodes (pvread et al) you can 
get the kind of control you're looking for (or at least that which 
I'm talking about). 
If you're having problems with it, post the orc/sco, and the command 
line you use in analysis, and hopefully someone can tell what's up.


Cheers,

        re