Different traces when single capture vs loop

Hi there!

Today I noticed something new, a huge difference in the traces when I got a single capture vs when there is a loop making several captures in a row for posterior analisys.

This is done with CW Nano, I’m with the LCP1114 Fuse Bypass notebook…

This is how single capture looks like (very clean to be honest):

But when I do a few ones in a row the plot changes A LOT:

As Nano does not have gain configuration is something I can do to improve this beyond executing scope.reset_clock_phase() that looks to have no effect at all?