[sdiy] Resonator type filters

[rburnett at richieburnett.co.uk]

I think vocoder bands must be a lot more sharp (higher order) than a 
"string filter" so the effect is a lot more pronounced with a vocoder as 
Mattias Rickardsson originally said.  When I passed an impulse through 
my vocoder consisting of a bank of thirty 1/4 octave bandpass filters 
all set to the same level it yielded a very definite chirp.  Clearly 
audible, and you could also clearly identify the change in direct of the 
"sweep" if that sound was recorded and played back in reverse.

My point... If you pass a signal through a bank of bandpass filters and 
then re-combine the outputs, the signal undergoes dispersion.  If 
there's lots of filters with very narrow passbands (high-Q) then that 
dispersion can become quite extreme and very noticeable as MR had 
stated.

-Richie,



On 2017-07-03 09:36, ijfritz at comcast.net wrote:
> Richie --
> 
> All true enough, but I'm not clear on what your point is.  I don't
> hear any chirp when a note is attacked on a violin.  I just tried
> pulsing my string filter: I can see the output is dispered, taking 100
> ms or so to totally die out, but I don't perceive anything like a
> chirp.  It just sounds like someone beating on a drum. Perhaps at a
> much higher Q setting you could perceive the chirp, but nothing when
> in normal operation, as far as I can tell.  And why would you call it
> all effed up when it is emulating a violin? :-)
> 
> Ian
> 
> Sent from XFINITY Connect Mobile App
> 
> ------ Original Message ------
> 
> From: Richie Burnett
> To: Mattias Rickardsson, Tom Wiltshire
> Cc: Synth DIY
> Sent: June 30, 2017 at 9:41 AM
> Subject: Re: [sdiy] Resonator type filters
> 
> If you pass a transient through a whole bank of high-Q analogue
> bandpass
> filters in parallel and then recombine their outputs what you end up
> doing
> is dispersing the energy.  Even if their responses combine in equal
> proportions to give a more or less flat magnitude response, different
> frequency components incur different delays unless some measures are
> taken
> to compensate for this.  As you said the more bands you have in your
> vocoder, for example, the high the Q-factors, and the more rapidly the
> phase
> changes across the bandwidth of each individual filter, and therefore
> the
> more severe the dispersion you get when you feed transients through
> the
> parallel filter bank as a whole.  For example, a click gets
> transformed into
> a chirp...  Or f####d up as you put it.
> 
> -Richie,
> 
> -----Original Message-----
> From: Mattias Rickardsson
> Sent: Friday, June 30, 2017 2:48 PM
> To: Tom Wiltshire
> Cc: Synth DIY
> Subject: Re: [sdiy] Resonator type filters
> 
> On 30 June 2017 at 10:46, Tom Wiltshire <tom at electricdruid.net> wrote:
>> Is there some reason why string filters use a large number of fixed
> bands,
>> rather than fewer variable ones?
> 
> Isn't the goal also to make the phase as messy as possible?
> 
> I always think it seems to sound more wooden the higher the number of
> resonant bands you have in a vocoder-type filter... even if all the
> bands are set to middle position. And I guess the phases of the sound
> is what's mostly modified then. Steady-state sounds are not as
> spectacular, but every sudden transient will be fücked up.
> 
> /mr
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy
> 
> ---
> This email has been checked for viruses by AVG.
> http://www.avg.com
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy