I just received my morph a few days ago and have been testing it out. One thing I noticed immediately when I looked at the visualizer is that as I drag a finger across the device with steady pressure, the sensed pressure varies quite a lot, the spike going up and down rhythmically, almost like a flickering flame, as if there is a lot of inconsistency in the individual pressure sensors, or a problem with how the information is interpolated or something. This isn’t a shaky hand! I was very careful to ensure consistency. I even rolled a ball bearing across the bare surface to rule out my hand.
In practice, what this means is that if I have a patch in a synth where, for example, pressure is mapped to filter cutoff or gain, if I slide across the device, that pressure reading is jumpy and makes the sound level vary rapidly. It is very audible and pretty much ruins the device for my intended application. I had hoped to use pressure mostly instead of an envelope and also want to slide around with pitch bends and the like.
The problem doesn’t show up if I just vary pressure while holding a finger in one place.
I have a feeling that part of the problem is variations in the sensitivity of the individual pressure sensors. It seems that this needs to somehow be compensated for with calibration curves for each sensor, which sounds problematic with so many sensors!
I just observed that if I hold my finger in one place and vary the pressure through the range and then move my finger slightly to one side and do the same, there is a big difference in sensitivity between the two nearby locations. It is harder to push the spike all the way up in one spot compared to another.
I decided to do a more controlled test. I set the device up as a big XYZ pad and opened MIDIberry so I could read the pressure output. I taped a weight to a drum stick and put a small soft pad on the end. Then I rested the padded tip of the stick on the surface while balancing the stick vertically using a little wire loop loosely placed around the top the stick instead of my hand so that I don’t impart any vertical force. Then I moved the tip to various points, observing differences in the pressure readings. I found that they vary considerably from place to place. I might measure 126 (on a 0 to 127 scale) in one spot, and as low as 30 in another. The differences usually aren’t this large. That’s the maximal difference I measured. Differences in neighboring locations more commonly fall in the range of 20 to 30. This is still substantial variation!
Considering that this device is a grid array of pressure sensors, I imagine it isn’t just an issue of variations in sensitivity between individual sensors, but also a problem of “pixelation” or aliasing of the “image” of the sensed finger tip. For example, a sensed object sitting on the surface might span 3 sensors wide in one position and 4 sensors in a nearby position. And the overall pressure must be computed from a number of neighboring sensors. There are probably limitations in how well this can be done.
Is this an issue that affects all of these devices, or is there something wrong with mine?
I am also having problems with false touches and releases in one spot when using the Buchla Thunder overlay in the upper region of the E, F, and G “keys” on the left. I haven’t observed false touches with the drum or music production overlays. I haven’t yet received the piano overlay and so haven’t tested that. I would imagine that the pressure variations I have observed could have something to do with these false touches.
I also notice a lot of velocity variation, but given the pressure variation, it is obvious that this likely results from hitting in slightly different spots.
Thanks for your help!