I've got several different drum VSTis that have "complexity" options.

The first is RealiDrums. It's got a number of different styles, and each style has sliders on each kit part (kick, snare, toms, hi-hat, etc.). Each slider can be moved to a setting between low and high. To implement this, a drummer played all the parts at various complexities, and then the data was tediously hand-edited to work.

The next is EZDrummer. Each pattern can be editing by picking a part of the kit (kick, snare, toms, hi-hat, etc.). An "amount" slider will show the number of hits in that pattern for that part of the kit, and that value can be dialed up or down. The velocity can also be adjusted. My recollection was that some sort of AI training was used to teach it where a drummer would put the hits.

I've also got JamStix, which is just too complicated for me to figure out how to make it work.

PG Music already has a MIDI engine in place for drums, keys, bass and guitar.

It seems to me that, given a "complex" drum pattern, converting it into a "simple" drum pattern is just a matter of deciding in what order notes should be removed. wink

Obviously, nothing is ever that simple. But I would imagine that some sort of beat importance hierarchy could be developed, and then applied to add/remove notes from instrument patterns. Or have a "simple" and "complex" pattern, so the algorithm would know what was essential to the pattern as it reduced complexity.

UJAM has a different approach. Each of their drum styles comes with the following patterns, triggered by a MIDI keypress:

  • 2 introductions
  • 5 verse patterns
  • 3 fills
  • 5 chorus patterns
  • 3 breakdowns
  • 3 specials


The verse and chorus patterns have increasing levels of complexity. Even so, I find I end up editing the MIDI data.

UJAM guitars have a similar approach with each style having 11 patterns of increasing complexity. They also come with 24 "global" patterns. With these many options, there's a lot more flexibility.

But an approach like this either requires sampling all the different chords with different strum types (upstroke, downstroke, muted), or a playback engine that can create a convincing MIDI strum, which is no small feat.

The big question is how well a new approach would integrate into the existing BiaB implementation.

-- David Cuny

My virtual singer development blog
Vocal control, you say. Never heard of it. Is that some kind of ProTools thing?

BiaB 2025 | Windows 11 | Reaper | Way too many VSTis.