Genlock: keeping the signal strong when things go wrong.

Two sources have to stay in step. A feed has to survive a network it was never built for. And when that network starts to struggle, something has to decide whether the picture degrades gracefully or just falls over.

Broadcast solved this decades ago with dedicated hardware and a single reference signal that everything else followed. A one-person operation solves it with SRT, RIST, NDI, RTMP, and a consumer internet connection, which is a very different problem with the same exterior.

• Where does each contribution protocol actually stop being usable, and what does the recovery cost in latency?
• What does sync drift look like without a shared clock, and does it actually matter for the kind of production you're running?
• When a feed loses its reference, does the system hold its last good state, or does it start guessing? Guessing is worse, and it happens more often than you'd hope.
• AV control systems from Crestron, Extron, Biamp, and Q-SYS each expose their own way in. How much of a shared, read-only view of device state can be built across them?

ST 2110, PTP, and NMOS show up constantly in this research, and not because anyone here is about to deploy broadcast infrastructure in a garage. They're where the vocabulary came from. Someone already spent years working out exactly what "in sync" means down to the nanosecond, and that's worth borrowing.

The gear mostly stays in the facility. The thinking travels just fine on its own.

Genlock gets the signal there. MuxNode handles what it costs to encode and move it. Cuebus handles what survives the trip after that.