I’m fairly sure that I spotted the original Molten MIDI on the back page ads of some trade magazine. But it sounds like I’m conflating that with the 2G device & app. Either way, it was inspiring to see someone thinking along (& beyond) the same lines as my hacked-together setup. And, yes, the G-Sequencer is some device; well thought out. Probably the only reason I haven’t caved in and purchased it … I’m juggling just as many exotic control devices on my ‘board & rack as I have pedals for destinations.
I found the Whammy4 programming to have some parallel – but not the exact same – ‘challenges’ as the H9. At the time, I was stepping through all 7-bit CC values in a sequencer, with two tuners flanking the Whammy. Start / Stop on the intervals within a given range of pitch shift. Drop Tune is a small range, simple, and accurate (even up to quarter-tones and eighth-tones). Octave & 2 Octave ranges (ascending & descending) were still reasonably accurate (+/- up to 10 cents, with the occasional 18 cents off). Dive Bomb … up to 30-35 cents … deviations that you’re going to hear. I never did attempt double-precision CCs with the Whammy.
I’ve ‘graduated’ to using the Knob value from an MPX-1 (or Novation controllers in Step Mode) to discretely pace through CC values one at a time, plus a MIDI monitor or two. The H9 / PitchFactor ranges have their own oddities. Some values of MIX, you’re just not going to get, even though the 0-100 range fits nicely within the 128 CC values (7-bit). All the DELAY ranges differ, depending on the maximum delay time. And it follows that thousands of millisecond values cannot be covered with only 128 CC values. So, many precise delay values can’t be accessed directly via MIDI. And understandably so.
But things like SCALE, KEY, PITCH MIX, and the chromatic stuff are easy to nail; even if you’re within 3-4 values of the target. I have whole MIDI ‘board banks devoted to simultaneous KEY & SCALE changes for the diatonic algorithms. Those kind of templates *breed* creative approaches. The Eventide pitchbend implementation is forward-thinking, but there are quirky items to get past. ‘Return to center’ values don’t hit UNISON, so it never really returns-to-center. To be fair, that’s with 7-bit CCs, while 14-bit (and I believe a 12-bit CC) nail the landmarks quite accurately [14-bit PB supported].
I realize that you’ve probably programmed test apps in assembly for your own work, but I thought that I’d share my own experiences with the H9 / MIDI connection. There may be some limitations, like don’t bombard it with parallel LFOs and high rates. The displays give up first, as if it were an overload warning. But overall, the sky is the limit with the Eventide’s extensive MIDI implementation.