Tx. This is fixed now. The “6 seconds per data point” for this run triggered a bug in the moving average calculation that caused the original power to be kept.
Hi again @david ,
It seems that since your correction last week, some old activities (still with long timelapse between date points, sorry about that) are now missing their 30s power (it only appears at the very beginning of the activity).
Here is an exemple of the problem which occurs on many activities.