threshold_hr scan not working as expected
the rough step size creates step curvers, which the fitter does not like as we don't like "0" noise.
the proposal would be to temporary decrease the step size from 10 to 5 (we are currently checking if it fixes the issue) until an alternative way to calculate (not fit) the threshold gets validated.
will open a MR in case the decreasing of the step size is demonstrated to fix the issue (at the cost of an expected 4x running time increase)