Background
During the “Last Night of the Proms” on BBC 1 HD in 2014, the pictures were easily 3 or 4 frames late compared to the sound: Freeview showing BBC SD was better. The following programme was in good sync on HD.
Pamela Barnes of NEP Visions (this being one of the companies taking over outside broadcasts for the BBC) has told us that it was a box somewhere along the system as everyone had been ‘happy’ at line up! It seems that someone didn’t really check somewhere downstream. All the sat comms, incoming OBs, outgoing lines to parks, all worked wonderfully!
Mike Jordan
It seems a lifetime ago now but when OBs were involved in a Christmas linkup between Aldershot/Germany and the Gulf, I was in TC5 (I think) overseeing all the Comms circuits.
We had a band in Aldershot terrestrial to TC, a regiment in Germany via EBU circuits, a regiment in the gulf by 2 hop satellite via Cyprus, all with reverse audio so everyone could sing in sync with the Aldershot band.
We had a Designs Dept very clever video/audio sync tester to set up one of the very first video delay lines (used in Italy to make their "network" delayed for feeding to cable so it was not truly "live") to get all the various timings right.
In the end however, the method we used and which worked superbly, was a real person just speaking! The eye is so very good at determining non-syncness.
We managed it then but now? (bring back CAR!).
Mike Giles
It was originally planned to use Pres B as the TVC mixing point, but the amount of kit required, sound and vision, ruled it out at the last minute and I’m pretty sure we went to TC3, but it could have been TC4.
We were using one of the essentially prototype Hot-lips, with TIM (speaking clock) and the canteen manageress from Kendal Avenue, on VT from Aldershot, giving us the clap as alternatives. In essence Hot-lips was giving us a figure for the overall loop delay for the overseas venues, including the new-fangled variable vision delays which had already been set to get the picture timings to match. But when we dialled the Hot-lips derived figures into the relevant delay lines, any improvement in sync was transitory, to say the least. Heads were scratched and TIM confirmed that Hot-lips was giving realistic figures, but the eye and ear told a different story when we sent the Kendal Avenue manageress round the loop. So we started again with no improvement, different results but still miles out. A degree of panic was setting in in the SCR as the due recording time approached, as it looked for all the world as though all the planning and theory was coming to nought, (or in fact to many milliseconds!) It was being suggested (by others) that sync would have to be achieved in the edit.
At that moment, a sheepish studio engineer poked his head round the SCR door and asked "Who is meant to be driving this into sync, you or us?" Because of the late move from Pres B, the relevant guys had not been involved in the initial planning and had got hold of the wrong end of the stick. They were trying to establish S/V sync by eye, by tweaking the vision delays again! So every time we put in sound delay, they tweaked the vision, putting us all into an impossible, tail-chasing loop, as we tweaked again, so did they!
So at the eleventh hour, they reverted to the numbers they first thought of, we put in our Hot-lips figures and things suddenly slipped into pretty near sync, but the final tweak was done using the clap and speech tape again. The essential problem being that whilst you could measure the overall loop delay, the figures for the send and return legs could only be derived from a calculated guess, based on the known elements of the various routes. I recall that the Kendal Avenue canteen manageress was outside in the dark, (beautifully lit, of course) and, as it was Christmas, looked very seasonal in a fur coat!
Nowadays, with a modern version of Hot-lips at each end, it should be possible to measure everything to the last millisecond.
David Denness
One would think in this day and age of embedded audio it is extremely difficult to screw up the synchronisation to any extent.
Mike Jordan
Surprisingly not that difficult.
One of the last shows I did for BBC OB Comms as a freelancer was a terrestrial link for Sky via Crystal Palace.
I was told that the vision had Dolby E encoded in it and I had never heard of it!
The last time I had seriously met Dolby used anything like this, was in about 1967 when working in BH lines department, we used Dolby on GPO lines to improve performance for stereo on OB temporary lines.
Thanks to the wondrous internet, I found E was digits in frame blanking. However the OB scanner routed the vision from the sub-mix via the main unit matrix which wiped the Dolby E and TC (en route to Sky) had no suitable test equipment.
Having lost 10 minutes of Sky3 audio, it was sorted. Unfortunately I had also had the only contact number Sky had available so it was down to me to sort out!
I later went down to Sky MCR (knowing an old BBC mate there) who informed me that when using E, they synchronised the syncs separately from the video so as to be able to carry the Dolby over multiple switching/mixing routes.
So even with digits and embedded audio, there can still be problems. It could have been a similar problem with the Proms routing.
Alan Stokes
I’ve spent time sub-mixing, in ICR at TV Centre (RIP). We’d set up and check sync with International Sound and vision, only to find it out of sync on TX. After much shouting and questioning it seemed that some PTT somewhere in Europe had swapped a circuit!
Rex Palmer
As one of the track cameramen on the Formula 1 GPs (though not in either Austin or Sao Paolo, but I will be in Abu Dhabi for the 2014 championship deciding race), I believe that the combined vision and sound signals (sound in syncs?) are beamed back via satellite from wherever the race is in the world.
The track sound, in stereo, is mixed in the Formula One Management (FOM) compound and then fed to all the broadcasters who add their own commentary (FOM do not provide any commentary on the F1 sessions) and then feed the combined signal on to the satellite.
F1 take their own power supply units to these events and work on the European standard. We shoot in PAL 16 x 9 hi-def but try to keep the ‘action’ within the 4 x 3 area, as the pictures go to some countries where they are still on 4 x 3! (we have a box outline in the viewfinder showing the 4 x 3 area). This also means that there are 2 sets of the on-screen graphics – one for the 16 x 9 feed and the other for 4 x 3. On the track cameras we get the 4 x 3 graphics on our viewfinder return feed, but the cameras on the pit wall get the 16 x 9 version for some odd reason!
Jeff Booth
Sound in Syncs was only relevant in PAL. Modern SDI (SD and HD) has audio embedded within the video data stream.
