On Saturday, 2 May 2020 14:41:54 UTC+1, Richard Livingston wrote:
It was intermittent due to the wires shifting slightly
every time the canopy opened/closed. Shielding both the
speaker and microphone wires fixed the problem.

Setting up these experiments I was intriged by how much small relative movements in the cables when they are bundled together, even at a single point in the cable run, can change the crosstalk. Will try to show that and the effect of distance in the next video if I can pull off a clean demo.

Richard