Quote:
Your example essentially says that 100% of the dB-changes are corrected per second.
No, that's what happens per sample. However since the difference is reduced at each sample it slows down quickly!
Quote:
However that would be much too fast because speed increases (linearly?) with greater distance from target.
That is correct. But as I said, the difference decreases with each step. So the speed drops rapidly. Which is usually exactly what you want (you don't want to HEAR it drop over a longer period of time).
Quote:
So maybe I should divide by the maximum distance to make it the maximum speed or by half the maximum distance to make it fit on average?
That's too much. Have to think a bit here....