Casio CZ/ VZ/ FZ - Pro Series

On 21 Aug 2008 at 23:31, synergeezer wrote:

> I'm certainly not claiming a deeper knowledge of psychoacoustics! Many
> members of the group are much, much more knowledgeable than I.
> 
> But the two questions I'm trying to answer in my work, are:
> 1. What are the minimum parameters required to re-synthesize an
> analyzed instrument sound and have it sound right?  (My first-cut
> analysis always yields a "distinction without a difference" for many
> parameters - which ones can I omit?) 2. Which parameters can be varied
> in a (more-or-less) random way in order to re-synthesize the kind of
> natural variation produced by "natural" instruments?
> 
> I think #1 addresses your comment.  

No, since resnynthesis is already some kind of a brute force method, 
analysing the atomic particles of a sound, but having all the data 
isn't the same as conceiving the true nature/essence of the sound.
To me a synthesizer that mimicrys a real instrument is boring as 
h*ll, unless it enables me to access/change certain basic parameters 
of the sound, like formants (especially their movement) 
texture/roughness and timbre. But timbre (pure waveform) is the least 
important part of regognising a sound, since human ear/brain can only 
distinguish about 50 waveforms, this parameter doesn't have to be 
very accurate, what might already answers your question...
It's rather that I think/experienced that all this synthesis methods 
all that math can be fold down to ways to manipulate those basic 
parameters I just mentioned...
You only get different results since they're providing another point 
of view and other tools to manipulate, but in general it's the same 
picture and when understanding this it's getting "easy" (still can be 
cumbersome) to get very similar results from different synthesis 
methods...

Regards!
        Summa



> 
> - synergeezer
> 
> --- In CZsynth@yahoogroups.com, "Summa" <flotorian@...> wrote:
> >
> > I wonder if any of you guys having a deeper knowledge about 
> > psychoacoustics. It might be less costly to create algorithms that
> > fits to our limited perception than trying to recreate physical
> > models or exact waveform copies...
> >   
> > Just my 2 cents!
> > 
> > ...Summa



-- 

CZ/VZ 		mailing list	: http://groups.yahoo.com/group/CZsynth
FMHeaven	mailing list	: http://groups.yahoo.com/group/fmheaven/
FS1R 		mailing list	: http://www.ampfea.org/mailman/listinfo/fss-list
Vokator		mailing list	: http://groups.yahoo.com/group/vokator
FM-Synthesis    mailing list         : 
http://launch.groups.yahoo.com/group/fm-synthesis/

http://www.summasounds.de/