[PD] circular buffer [attn. obiwannabe)

[padawan12]

This is very close to the original "resobass" from the audio example on my site.
Version 1 kinda died when I lost some abstractions. You'll have to mess about
with the controls for a while to work it.

There's 5 parts

1) Excitor is just a decaying pulse from an [ead~] generator.
2) Prefilter to set the strong components in the excitor burst.
3) The jailed delay
4) Post-filter, parallel body resonance
5) Drive - colouring distortion

The trick to taking this patch further imho is getting
the tracking of the prefilter frequency to play nice
with the note value, right now you get scales where 
some notes are completely damped and others ring out
loud as one frequency scale goes in and out of phase
with the other. 

Jailing here is just [clip~] units at 0.5, and
the method of clip, filter, clip, fliter is clumbsy to set
up but effective. Otherwise you get into a "whackamole" situation
where your clips spread out harmonics and you get other feedback
frequencies popping up from the sidebands. In this patch we're kind
of deliberately doing that, but in a controlled way. The key to 
never getting a completely run-away note is that the feedback is
controlled by an envelope which always decays away, I think 
that's labelled as "damping" or somesuch. 


On Wed, 27 Sep 2006 13:35:27 +0200
derek holzer <derek at x-i.net> wrote:

> Hi Andy, list,
> 
> On 
> http://www.obiwannabe.co.uk/html/sound-design/sound-design-audio.html, 
> Andy wrote:
> 
> > Keeping a circular buffer from saturating while filling it with signal fragments is the essence of delay based resonant modelling, one way to manage the growth or decay of signals is careful "jailing" using clipping or compression units. 
> 
> Would you be willing to share a patch demonstrating this? I find that 
> feedback-based structures are still difficult to do without saturation 
> in PD. This is especially true when filters are involved, since PD's 
> filters are quite easy to "blow up". I've tried using the Zexy limiters 
> and even a "normalizer" abstraction posted here some months ago but I'm 
> still looking for the right solution. A compressing or normalizing 
> method is far preferable to a clipping one, since I'm going for a more 
> organic and less digital sound with the stuff I'm building right now. 
> Any chance of posting the patch which made the "Resobass" sound?
> 
> thanks,
> derek
> -- 
> derek holzer ::: http://www.umatic.nl
> ---Oblique Strategy # 76:
> "Give the game away"
> 
> _______________________________________________
> PD-list at iem.at mailing list
> UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 10-resobass.pd
Type: application/octet-stream
Size: 9777 bytes
Desc: not available
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20060928/c2a14193/attachment.obj>