| ??? 09/15/07 03:29 Modified: 09/15/07 03:33 Read: times |
#144640 - No disagree... Responding to: ???'s previous message |
Hans said:
As I said before, this will remove high frequency components, but you will still see plenty, plenty of low frequency noise after averaging. The only way to get rid of that is to average over an even larger amount of samples, but this will greatly affect the speed at which the system can respond to change, resulting in sluggishness of the controls. I understood Chico's last reply... Chico said:
Sometimes, in example: it keep sending values 38 39 38 39 38 39........(making midi busy, software busy, and is very ulgy!) until I try to stop making some change in potenciometer rotation (position?) that he doesn't like that the flickering was making the MIDI transmission busy. So, the averaging and decimating would at least help in this issue. What is wrong, when he transmits a low frequently flickering MIDI signal (<=1LSB), at a transmission rate of let's say ten times a second? Also, we don't know, what Chico wants to control with his potentiometers. If the change rate has not to be high, then averaging and decimating will solve his problem. Hans said:
Deadzone constricts the range a little by introducing a nonlinear response. "loose resolution" sounds a little drastic to me. You only loose 2 or 4 steps on the entire range. Yes, if he transmits only a MIDI signal when the change between two succeeding averagings is >=2LSB (if his noise is <=1LSB), then this could be a good solution. It would also allow faster change rates in combination with a less heavy averaging. You are right. But if the deadzone is much wider, then the MIDI control can become erratic. Kai |
