As I have queried from the TS, the issue is why the Siglent, with its default probe and common (alligator) ground clip, produce the characteristic spike while the Tek 465 does not.
As noted prev., I have been able to reduce or increase the spike based on ; (a) direct BNC input from DDS ; (b) touching the cable or moving it around .
Yes, the spring ground contact can improve fidelity, as shown here:

But, in real-word circuits, such "ideal" conditions do not exist.
Maybe at this particular (low-cost) range of DSO's -- any brand -- these type of metrics are not reliable; they are "artifacty" because of price/performance trade-offs.
Nevertheless, since the spike is significantly large in the Siglent (compared to the Tek 465), the mystery remains.
As noted prev., I have been able to reduce or increase the spike based on ; (a) direct BNC input from DDS ; (b) touching the cable or moving it around .
Yes, the spring ground contact can improve fidelity, as shown here:

But, in real-word circuits, such "ideal" conditions do not exist.
Maybe at this particular (low-cost) range of DSO's -- any brand -- these type of metrics are not reliable; they are "artifacty" because of price/performance trade-offs.
Nevertheless, since the spike is significantly large in the Siglent (compared to the Tek 465), the mystery remains.