k y m a • t w e a k y — the kyma collective || view the current website || February 2015 Archive

/ WebHome / Sounds / Share.JX3Pv1Discussion


Share Section



All Sections


Login / Register 
Change password
Forgot password?

Symbolic Sound

Kyma Forum
Eighth Nerve

TWiki Links

TWiki Formatting FAQ
What is TWiki?
%SESSION_ENDIF% twiki.org


This Sound is based on Roland's JX3P synth. Two versions are available, JX3P and JX3P Lite, which is more DSP-wise.

Update v1.1:

Kurt has contributed a simple but very effective optimization. Now the Sound runs with 6 voices flawlessly, however I have left the polyphony to 5 voices for those who have small systems.


  • PG-200 programmer style VCS.
  • two oscillators, DCO1 and DCO2.
  • LFO and Envelope controls for each oscillators
  • VCF filter optionally controlled by the main ADSR enveloppe
  • Chorus (a simplified version of User.BenPhenix's OmniChorus)
  • HighPass filter
  • output level
  • 5 polyphony voices

Differences from the original synth:

  • the waveforms are not from the original synth. I rather borrowed GlitchPop's moog wavetables. I actually sampled the JX3P, but the resulting waveforms were not as interesting as those I am currently using in the Sound.
  • I tried to reproduce the chorus accurately, and though it doesn't sound like the real one, it really adds a nice stereo image and depth.
  • the Metal mode in DCO2 is missing. This feature is not documented and is not a simple Cross Modulator as it is stated in the synth's documentation.
  • there are 5 polyphony voices instead of 6 (it can be changed by changing the MIDIVoice sound inside the patch).
  • in the non-lite version, it is possible to crossfade the waveforms, the lite version acts more like the original synth in this sense.
  • the Pitch Follower feature is not implemented.
  • the Mode Gate/Env feature is not implemented.
  • the Tape Program Archiver feature is not implemented.

How to use it ?

  • first you have to check if your MIDI keyboard or Wacom Tablet is correctly patched.
  • the DCO1 module is the simplest to use: set the Range (16', 8' or 4'), select the waveform (Sawtooth, Pulse or Square) — if you use the non-lite version, you will be able to freely crossfade between all those 3 waveforms. The switch FreqModEnv1 activated the potentiometer ENV, which actually modulates the pitch in function of the ADSR enveloppe. In the same manner, the FreqModLFO1 switch activates the potentiometer LFO which controls the amount of LFO used to modulate the DCO1's pitch in function of the LFO.
  • the DCO2 module, has the same features as DCO1, but it add other ones:
    • this oscillator, is tunable — the Tune rotary sets the tuning in half-tone, and the Fine-Tune rotary fine-tunes from -50 to +50 cents.
    • if the Sync switch is checked, the frequency of this oscillator is matched against DCO1. Hence, the Tune parameter is discarded, but the Fine-Tune rotary acts as a static phase shifter.
    • if you check the Noise switch, the oscillator is replaced by a pink noise generator.
  • the LFO rotary controls how much of the LFO is used to modulate the pitch of either DCO1 or DCO2
  • the ENV rotary controls how much of the Enveloppe is used to modulate the pitch of either DCO1 or DCO2, the EnvNeg switch inverses the shape of the enveloppe in order to create inversed pitch modulation.
  • in the VCF pane:
    • Source-Mix controls the amounts of DCO12 and DCO2 used in the patch
    • HPF is the hgh pass filter, which acts in the final stage of the signal chain
    • CutOff controls the frequency of the CutOff filter, this frequency is also changed by the LFOMod and ENVMod parameters.
    • LFOMod controls how much the LFO modulates the CutOff filter.
    • ENVMod controls how much the ADSR enveloppe modulates the CutOff filter.
    • the VCFNegEnv switch inverses the shape of the ADSR enveloppe which is sent to the CutOff modulator.
  • in the LFO pane:
    • Waveform lets you choose the shape of the LFO between a triangle waveform and a square waveform.
    • the RandLFO switch puts the LFO in random mode. Note that the results are different wether you choose a Triangle or Square waveform.
    • LFODelayTime controls the attack time in seconds of the LFO.
    • LFO rate controls the rate of the LFO : smaller values, slower LFO.
  • in the Enveloppe pane you will find standard ADSR parameters, which stand for Attack Time, Decay Time, Release Time, and Sustain Level.

To Do:

  • implement the Pitch-Follower function.
  • use good samples of the "original" waveforms.
  • try to make the Chorus more like the original one.
  • add new functions like distortion, voice panning, phasing ...

-- User.CamilleTroillard - 28 Jul 2006

© 2003-2014 by the contributing authors. / You are TWikiGuest