Never used the Arturia stuff, but apparently the capacitive grid is the technology behind touchscreens, which are as uninspiring as could possibly be.
I also tried but failed to find any research on sensing touch direction and rotation (which I guess reduces to direction, as a change in direction) – which was my initial impetus for thinking of grid techniques which could at least indirectly measure forces parallel as well as perpendicular to the touchplane, via the finger blob shape.
Of course – you can get those as well from position trajectories… but I don’t know if the measurement would be as accurate, and there are plenty of cases where one may want to press in various directions, or just swirl one’s finger around without actually moving it.
Even a collection of three “perpendicular” sensors (probably a fourth for redundancy) per position could be used to infer these parallel forces and their directions.
Also – there’s the issue of the feedback – rather than taking the somewhat detached and heavy-handed step of “estimating” direction, force, position, rotation etc – higher-level determinations that will invariably result in glitches (and perhaps latency if DSP techniques are required) – is there something simpler and more reliable, more low-level that could be used as a control. I’d think here the second-order (elliptical) moments of the finger blob – or some kind of mixture-Gaussian fit for multitouch – could be very interesting to work with here. Hence why I keep thinking of grids of some sort.
