Improving slice algorithm for transient finding


I just used the auto slicing function for a digitally recorded (= noise free) drum sample collection within one sample. first, nice that there is such a functionality :slight_smile:

But I would like to suggest that the transient finding could be quite improved. It mostly doesn’t place the marker at the beginning of a new drum. In my example before the drums there is always a flat 0-line. Often the marker is either starting too early, so a delay is introduced, or a bit after the start of the sample. So you have to edit all points manually afterwards, to ensure good marker points.

Is it possible to improve this transient detection?

If not for 3.1, maybe could be moved to 3.1+ todo-list?