[Si-Co]

This is a huge simplification, but we know that a frame of an analogue picture is made up of 625 lines. For simplicity, we'll assume that the lines are transmitted one at a time in order.

The sync pulses can be thought of as markers that identify the start of a new line and a new frame.

If, at a given point in time, source A is on line 5 and source B is on line 429 and you try to cut between them, it's going to jump - instead of line 6 you get line 429 so you don't get a full frame and it jumps. So genlocking will gently force one source to be in sync with the other by adding or subtracting a couple of lines to each frame until they are both at the same point.

The more modern technique is a digital frame store which holds the last frame of source B in memory and plays it out in sync with source A


(Again, I fully acknowledgte that's is a massive simplification for the purposes of illustration)

Thanks, Steve, for your layman’s explanation. It may be a simplification, but it makes things a lot clearer!

This can be seen with this breakdown on LWT. AIUI the vision mixer in LWT TX died. The programme returned after they patched the VT straight to air

youtu.be 

If the problem was with LWT pres, would that cause their clean feed to other regions to be lost (because in this occasion that happened)? Wouldn’t that feed normally bypass the pres vision mixer unless they needed to do an announcement over or after the end credits, or network a promo or similar - and even then they would be unlikely to flick the switch to “go dirty” so far in advance?