The lossy part needs to go into the feedback loop of the delay. Each time around the loop the distortion function shifts some of the energy away from it's original position, some up and out band and some into the capture of the bandpass. As it goes round and round (recurses) it gets focussed more and more into one or two places. Good dub delays are always pretty close to exploding in certain frequencies close to the 1-2kHz mid. That's the trick. You need to experiment with dangerously unstable settings to get it to sing. Distortion followed by bandbass. Any kind of tape and amp simulation in the loop usually works quite well. Some of the best old dub delay IMHO is from a "bucket brigade delay" circa '76/77 that used leaky capacitors to simulate a tape, a kind of analogue data buffer.

http://www.youtube.com/watch?v=XYepvarTsVM @3.44

can get a subtle, spooky, slow timbral shift if you catch it and ride it.

http://www.youtube.com/watch?v=MgAfsUdQc_A

@2.40 +

On Mon, 25 Oct 2010 18:55:17 +0100 Marco Donnarumma devel@thesaddj.com wrote:

Hi all, how would you suggest to synthesise dub chords in Pd?

I tried some waveshaping applied to phasor chords + some resonance filters but it resulted in something totally different.

thanks,

-- Marco Donnarumma aka TheSAD Independent New Media Arts Professional, Performer, Teacher Ongoing MSc by Research, University of Edinburgh, UK

PORTFOLIO: http://marcodonnarumma.com LAB: http://www.thesaddj.com | http://cntrl.sourceforge.net | http://www.flxer.net EVENT: http://www.liveperformersmeeting.net