On Mon, 12 Oct 2020 15:36:51 +0100, Tim Lamb wrote:

BBC Click (last saturday) showed a solution to that, that only
needs
a few Kbytes/second thanks to something that nVidia are
developing.

I watched and failed to understand that. Perhaps you have local software
that delays and gap fills from a series of *stills*?

Seemed reasonably clear. There is a reference still image that has a
number of key points defined. The movement of those key points on the
real subject is tracked by software and that data sent to the reciver
where the key points on the still image are moved accordingly.

--
Cheers
Dave.