An experiment with pitch synchronous delays

(74 Votes)
1.4 (Updated 12 years ago)
October 22, 2006
Reaktor 5 or lower


An experiment in harmonic manipulation using pitch synchronous regenerative delay lines applied to oscillator waveforms. Good for making shimmery pads and off-centre leads.

Note: This is a CPU hog (30% on my machine), but well worth the extra cycles in my mind.

(I shamelessly stole the mod wheel image and delay and chorus modules from Stephan V.'s great Tritium synth. Thanks Mr. V! And thanks to Bernd Keil for the knob images)


Thomas Clement
5 years ago
Love this. Very organic in a 50s/tube kind of way. Did I mention I love it?
Patrick Adorati
7 years ago
Hi James Clark....I'm new to the online library and I tried to write you but I couldn't because I don't know if it's possible.... I am building a special type of synthesizer that combines different kinds of synthesis. I found your Derelayer FANTASTIC...and I would like to use this sound engine in my synth,with other engines produced by me. Can I do it?? In case your answer is "yes" I'll add a logo "powered by J.J.Clark" in front ("A") of the synth and a very big logo in the "B" panel. Excuse me for my bad english...I hope you'll answer me.
Christopher Soulos
11 years ago
A Work Of Art
Evan Schiller
12 years ago
stellar. thx!
Michael Doherty
12 years ago
David Horsley
12 years ago
Really useful for me. Thank you for this.
Mark Bonnington
12 years ago
Gustav Holst would have loved this ensemble!
ned Bouhalassa
12 years ago
Space-y in the best way! Another cool synth is added to our solar system... ;-)
Florian Schwehn
12 years ago
I am sorry, I like it, really!
Stephan Becker
12 years ago
You're welcome, Mr.Clark ;-)
James Clark
12 years ago
Hi Florian, thanks for the feedback. If I get some more, maybe I will saturate! As for your suggestions: - I put in a HP in the feedback path as you suggested, although sometimes the nonlinearity the offset accumulation gives can be interesting. - I can't strip down the 3X ladder filter as you suggested, since the feedback path for the resonance comes from the 4p output of the filter. Even though I send the 1p output down the chain, the 4p output is being used in the resonant peak. - I changed the P input of the osc to event, although this changes the cpu only a little bit. - if you really want to reduce CPU you can change the delay lines from 4p to 2p. This makes the sound not as good, however, and I will always trade CPU for sound quality.
Florian Schwehn
12 years ago
Nice, but here's some constructive criticism: - if regeneration is set fullscale the oscillator produces some DC offset that grows over time. To avoid that either limit regeneration to 0.99 or something or add a highpass filter (frequency about 10 Hz) to your feedback line - that will remove the offset - to get your cpu load down, through out everything that isn't needed - escpecially in core cells. for example there is a 3x ladder filter but only output 1 is in use. so remove everything that only affects outputs 2-3. Another example is the saw/pulse osc: switch the P-input from audio to event - because you just put in signals at event rate. Sorry - I know that I'm a backseat driver - you built a beautiful synth!!!