Live video sounds like one service. It is two, and they fail in opposite directions.
A broadcast feed must not drop for ninety seconds while a presenter is crossing to it. Nothing else matters as much — not resolution, not archive, not coverage. It has to be there, now, with latency low enough that a studio can talk over it.
A site monitoring feed is judged the other way round. Nobody is watching most of the time. What matters is that when someone does look — or looks back at last Tuesday — the picture is there and the record has no hole in it. A ten-second buffer is irrelevant; a missing afternoon is not.
Designing one as though it were the other is the usual mistake. A broadcast rig left on a site produces no archive. A monitoring camera pointed at an event produces a feed a broadcaster cannot use.
On air
For live transmission we fly with a licensed pilot and a dedicated link, deliver a zero-latency Full HD feed, and hand it to the broadcaster or straight into a streaming platform. Ground and aerial feeds can run together so a director can cut between them.
The work that actually determines whether this succeeds happens before the day: the flight approval, the crowd overflight permission, the frequency coordination, and a plan for what happens if the aircraft has to come down mid-transmission. We arrange the permits ourselves, and on event work that lead time is usually the binding constraint rather than anything technical.
