Most if the time Sandy Heath was fed by the Radio link from Norwich .
AT some time before the opt Cambridge would signal to feed the Norwich output (rather than The contribution microwave from FRV Luton Nhampton) to it .
Often only a few seconds before the opt (if the FRV was in the west of the patch and doing a live before the opt)...
Once the incoming feed from Sandy Heath had BBC One Norwich on it, you'd then hit another button, which locally triggered the BBC Cambridge SPG to Genlock to incoming BBC One Norwich (So that BBC One Norwich and local sources were synchronous on the Cambridge vision mixer)
This was to avoid adding another frame of delay I believe - which would have been the case if Cambridge had been free-running and BBC One Norwich had been synchronised incoming to Cambridge.
Put that through the mixer and then signal to feed Sandy by the return circuit ...
and signal to feed the transmitter from Cambridge ....
Yes - the third button (which was key-locked - eventually) remotely switched the Sandy Heath transmitter between its feed of BBC One Norwich and the feed from the Cambridge gallery.
This was always a synchronous cut (no frame rolls), but the two versions of BBC One (BBC One Norwich and BBC One Norwich via Cambridge) were relatively delayed by a couple of frames.
I was told this was due to a PAL composite Tektronix synchroniser at Sandy Heath synchronising BBC Cambridge gallery incoming with the BBC One Norwich local feed at the transmitter. (PAL synchronisers could introduce more delay than component ones because of 4-field/8-field issues?)
all that switching was in the SIS domain and I think if there had been
a Tek synchroniser it did not work well !!! So the opt switch was asynchronous..
In 6 months of carefully watching the soft-opt point I never saw an asynchronous cut, but this was before the system was re-engineered for DTT
(where a relatively complex system was installed that allowed DVB-T opts in Norwich and Cambridge and put a vision circuit from Norwich to Cambridge (which previously wasn't an option) that was available when Norwich hadn't opted out (when Norwich opted out this circuit carried Norwich's output to Cambridge and triggered an opt-out in Cambridge too I believe)...
( and a relay in any case)
Then you could gracefully fade over from Norwich to local sources.
Yes - genlocking BBC Cambridge to BBC One incoming ensured all local sources were locked and synchronous with incoming BBC One, so you could mix, fade down and up without frame rolling etc.
Thus having a permanent feed that never opted at Sandy was seen
to be a great improvement
As the analogue was being fed by a digital distribution
.. the Network recall was done as it is for all digital services in coding and mux ...
its sort of how buddying works after all.
And I'm fairly certain that a sky box did 14:9 despite it being WSS not AFD.
Consumer Sky boxes were either permanent 12F12 centre cut, or switched between 12F12 and 16L12, driven by the MPEG2 header switching (Sky boxes didn't insert WSS either, though some passed on WSS (some blanked it) if it was present in the received video. ISTR that the Sky SD chains had ARCs
in their chains upstream of the MPEG2 encoder - that were fed the permanent 16:9 feed (containing 12P16 or 16F16 sources) of BBC One - and then dropped in a 12F12 ARC when 12P16 was signalled, and stayed in passthrough when 16F16 was signalled. The 12F12 ARC signalled to the MPEG2 encoder downstream that the source was now 12F12, whilst in bypass it signalled 16F16.
There wasn't a 14L12 implementation on consumer Sky boxes. (Just as there isn't a 12P16 option on Sky HD boxes for 4:3 SD output in 16:9 HD...)