Auto scaling issues after removing output disable about bodeplot HOT 5 CLOSED

Without eas the scope fails to trigger on the signal at all. Now this doesn't happen always, but only from time to time.
Unlike with eas the problem does not go away after 4 or so readings but persists throughout the entire measurement.

from bodeplot.

I connected the integrated counter of the DG1022Z via an additional inline t-splitter in hopes of getting it to snitch on one of its family members. But I was also having trouble with it triggering on the lower frequencies < 1 kHz. But as the frequency increased the reading fluctuated less and less. After trying out a couple of different triggers sensitivities I tried the "HFsuppress" option which within reasonable bounds eliminated the jitter.
In the oscilloscopes trigger menu, there is a similar setting called "NoiseReject", unfortunately, the manual does not elaborate on the options effects, functionality nor recommended frequency range.

from bodeplot.

Even with the "NoiseReject" setting and trigger coupling set to "LF" the issue persists.
Implementing fully custom autoscaling over scpi has the same problem.
The problem:
When starting a new measurement at a low frequency connected to a RC low pass filter the 2. OSC input captures only millivolts of voltage. This is not the case when there was a low-frequency measurement beforehand.

from bodeplot.

There seems to be some sort of memory effect on the DG1000Z.
On the first frequency set by a new measurement sweep, the last frequency of the last sweep is still outputted even though a new frequency has been set.
And like that the DG1000Z snitched on itself:

On successive frequency writes this issue seems to resolve itself and the DG1000Z outputs the correct frequency.

from bodeplot.

A pause of 0.1 s between the sine command with the new frequency and the 'output on' command fixed the issue.

from bodeplot.