First post here: I have a few questions to the engineers about using JP8V as a sonic replacement for my 12-bit Jupiter 8.
I have spent enough time with the JP8V product to recognize it as a formidable replacement sonically for my real Jupiter 8, in many respects. The envelopes have that same snap, that organic-sounding quality is there, and the hugeness from stacked oscillators is a real pleasure to hear in software. The level of control with cross-mod, compared to the real Jupiter 8, was also a welcome feature. I did notice some digital aisling for the highest octaves, even with the audio interface sampling rate set to 96khz, but this octave range was, for the most part, minimal and negligible compared to the remaining octaves.
There are two items however, that I would like to see added to raise this impressive level of authenticity even higher.
- Seperate audio channels for Upper and Lower:
This is an advantages of the JP8 for quick access to simultaneously add depth and thickness to a patch. Since 4 voices are isolated to Upper or Lower for any mode but Whole, panning each side hard left or hard right produces the widest and richest sound that this synthesizer can offer, since less phase-cancellations occur. This is as much true for unison mode as it is for Dual and Split modes, assuming one does not use the "MIX" output. Since the oscillators are free-running and are never quite in tune with one another, this trick works without fail on a real Jupiter 8. It is also much faster than the Oberheim equivalent of panning pots for each voice, as these are trimpots underneath the front panel, well out of quick reach.
It seems to me this upgrade for JP8V would be easy to add for any VST host. Even the standalone version allows you to select 2 channels seperately for ASIO operation, so hard-wiring Upper and Lower to one or the other seems like a viable option that could be toggled on or off. The Oberheim method of a panning value for each voice would be nice as well, but certainly not as expedient when setting up presets for Dual/Split modes. I suppose this forces the polyphony count to always be an even number, unless either Lower or Upper gets one extra voice for an odd polyphony count. - Adjustable tolerances for each voice:
First a question about the voice-stealing algorithm: When the polyphony count is adjusted or the software is initialized, does each voice live as a non-volatile module with its own array of tolerences, with JP8V's Midi interface triggering them, in the same manner that a real polyphonic synth under CPU-scanned keyboard control would? Or does the software just allocate and destroy new voice modules from templates on demand?
If the former is true, and the software acts more like a real polysynth, an ideal degree of control for presets would be to have a graphical interface where the + or - tolerances of certain front-panel voice parameters may be configured for each voice.
Perhaps a better idea is what the designers of Sonic Projects OP-X have done. There is one knob to turn that pushes specific voice parameters out of tolerance from the front-panel values, in a bipolar fashion. An increase in the rotation of the knob increases the relational difference between voices, although I cannot tell if it is randomly generated. There are switches below this knob to toggle which front panel parameters may be changed, which naturally includes pitch deviations from the front panel oscillator tuning.
I understand that when not in unison mode, JP8V uses the detune knob, at lower-left, to randomize the master tuning of the oscillators and at the extreme, the envelope decay. What I am unsure of is whether these deviations are fixed per-voice, similar to voice tolerances on a real polysynth, and whether a future version of JP8V could also add other voice tolerances, such as Oscillator 2 detune and filter cutoff.
I feel that once these have been achieved, JP8V will stand even closer to its real life counterpart in authenticity.
