This guy Chris Wilson, such a problem. :)
setTargetAtTime - exponential decline. Parameter - time constant:
"timeConstant is the time when a linear continuous first-time continuous system is required to achieve a value of 1 - 1 / e (about 63.2%), taking into account the input response of the step (transition from 0 to 1 value).
Thus, for each time period, the level will decrease by 2 / 3rds (assuming that the gain is 1, and you set the target to 0. At some point, the decline becomes so close to zero, it is below the noise threshold and you don’t you need to worry about it, he will never “reach the target value” - he will come close to it, although, of course, at some point the difference falls below the accuracy that you can imagine in the float.
I would suggest experimenting a bit, but here is a quick guess at what you started:
cwilso
source share