[sdiy] Why Bands for String Filters and Narrow and Numerous

[Edward Schultheis]

But what fun is that!? I for one love the challenge of doing things in the analog domain. I even steer clear of some mutable stuff simply because it makes it too easy for me!
 The satisfaction in getting sounds that are more easily achieved in the digital domain, is so worth it! 

Edward Schultheis

Sent from me!

> On Jun 30, 2017, at 3:36 PM, Richie Burnett <rburnett at richieburnett.co.uk> wrote:
> 
> If the aim of a string filter is to produce a complex uneven frequency response so that it translates frequency modulation of source harmonics into amplitude modulation, then there are probably more efficient ways to achieve this goal digitally.
> 
> For instance, a handful of comb filters with carefully chosen feed-forward and feed-back taps can produce a very uneven frequency response with hundreds of peaks and dips scattered across the audio spectrum.
> 
> -Richie, 
> 
> Sent from my Xperia SP on O2
> 
> ---- Bernard Arthur Hutchins Jr wrote ----
> 
> 
> Tom Asked Fri, 30 Jun 2017 : Is there some reason why string filters use a large number of fixed bands, rather than fewer variable ones?
> 
> 
> 
> Simply, the filters are sharp (narrow), so you need a lot of them (say 40) to span (overlap) a wide enough processing bandwidth. So (you rightly demand!) why do they need to be sharp? For achieving dynamics, either the filter frequencies or the input frequency typically needs to move.  (Else you are merely waveshaping one boring fixed shape to another.) Moving all the filter frequencies is of course impractical while moving the input is basically just FM of a VCO.  If you have only a few broad filters, AND FM depth is small (typically achieving an animated dynamic sound like tremolo), very little equivalent AM (harmonics moving up and down BP slopes) is achieved. With lots of sharp filters, lots of relatively large motions are achieved (some up, some down, some flat).
> 
> 
> 
> The other application is as a formant filter.  (That is, not FM, but note-to-note spectral shape changes.)  The ability to get lots of detail (with adjustable depth) is useful, and variable Q, ALL 39 filters with one control (overall feedback), was demonstrated in my Electronotes EN#115 filter bank. I will see if I can post this if anyone is interested.   
> 
> 
> 
> -Bernie      
> 
> 
>   
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20170630/b23abd27/attachment.htm>