BBC One Cock Up 8pm 1/5/2006

I take it you were watching in England....??? I think I know what might have happened, but to explain it you need to know how the audio routing on the BBC 1 and BBC 2 symbols (idents) works.

I might have this a bit wrong... but... how I think it works in the Broadcast Centre... is that when you see a symbol intro before a programme, there are actually two copies of the symbol running simultaneously off two separate playout ports.

The symbol you see on screen is running with vision only (the director disables the sound on that event in the schedule), the music you hear is coming off the other symbol running in the background off another playout port. Sound from that port is routed to a fader in the announcer booth allowing the announcer to dip the music themselves. The audio output from the announcers' booth is faded up by the director before each junction.

So... what I think might have been the problem... is that the audio on the "symbol you see" was mistakenly not disabled. That would have resulted in music from two symbols being heard at the same time.... one of which the announcer could dip, the other they couldn't.

I could be wrong though, it could have been a different problem.

I don't think they do play from two ports simultaneously. Certainly in the old DTA the audio from the designated 'symbol' server port was hard-wired through to a fader on the announcer's audio side-car, so in the schedule the director would set the symbol vision to the correct server port and the audio to silence (actually the black and silence source which came up as 'BLK' on the SSD) so that the announcer had control. That's why if you picked up a 'mixer-out' clean-feed you would get a mute symbol. The same applied to moving pointers, and now and again if the designated 'symbol' port fell over, the director would have to play the symbol on a different port and take control of audio themselves.

Certainly under those circumstances if the director left the audio enabled through the mixer as well as the announcer's side-car you would expect to hear phasing as the same audio went through two different chains of extraction/embedding.

BUT, how this translates to the Broadcast Centre, Colossus, and dynamic source allocation I don't know - unless events to be live-voiced are flagged to force the automation to select the same server port every time.

That would explain it - the fact I noticed the phasing indicates the 2 ports output sound that is *almost* running in perfect sync, but just an increment out. Obviously the volume of the idents were louder than usual, and the phasing decreased/disappeared when the announcer cut in, leaving the mistakenly enabled audio to be dominant while the intended broadcast audio was faded down.

I remember how things worked in the DTA, but as you say, things are different with dynamic source allocation... the symbol is not always going to play off the same port.

On the Colossus Viewer screen there are two 'buttons' called Clean Feed One and Clean Feed Two. One of those, when selected, routes that event (no matter what port it is played from) to the Programme Only Clean Feed for the Nations and CBBC.... the other 'button' is used (I think) to route the symbol audio to the announcer's symbol fader.

I remember on a visit to the Broadcast Centre, one of the directors showing me the schedule, where every symbol was made up of two events, a MAIN event (with the audio button deselected) and an OVERLAY event (a bit like a guard source).

If you take a look at 'mixer out minus con' on the switchable clean feed during a junction, the symbol should still be mute (if the announcer is live)... however if you have a pre-recorded intro the 'mixer out minus con' feed will actually have music and v/o on it. I know this was resolved in the DTA, but I don't think it has been resolved in the BC.

You clearly took more notice of the minute details than I did! The CF1/CF2 buttons obviously perform the event flagging that I suspected must happen somewhere along the line.

I assume the overlay event could be anything which would run concurrently with the primary event - e.g. ECP, voiceover (of end credits or symbol) from audio server etc. That's why I suspect that they don't have to actually run two clips from two separate ports at the same time, so if (as in say mornings on BBC TWO) they're running symbols with pre-recorded voiceovers from the audio server, then the automation must have a way of ducking the main audio while playing in the overlay event, so the audio of either the main, or the overlay, must be separate rather than both being combined in the PGM out mix. Come to think of it, isn't there a bit of SSL-style 'flying faders' going on with the directors side-car?

Anyway, whatever the solution, I'm sure we'll be intimately familiar with it in due course

How come they have two symbols on the server?
Though I know nothing about the technical side of things I was led to believe that on video source 1 was the ident, audio source 1 was the ident music and audio source 2 was the announcer (live or pre-recorded). I then thought that audio source 1 was dipped when the announcer spoke. Obviously this is not the case, as has been explained, but why couldn't something like this be done?

Don't forget we've been playing with Colossus in Belfast for about 3 years now. OVERLAYS are, I believe, the method used to run ECPs etc... though I've yet to use an OVERLAY on-air.

I'm fairly sure, though of course could be wrong, that for live symbol v/o's on network, two copies of the symbol are run simultaneously. The main (without audio), the overlay (not seen on screen) the audio of which is routed to the announcer's symbol fader; the announcer's booth then being faded up on the director's audio side-car. If memory serves, the schedule I was shown had the name of the symbol on both the main and overlay event, the overlay event having a duration of 1 frame I think, which did not actually have an effect on overall timings.

The way we do it is to either... use the manual lag fader... or put an automated lag in to the schedule and dip using the master fader, letting the automation fade out the audio of that symbol event for us.

Yes, audio from a DVE (or Clarity) event would be routed to an assignable fader on the director's audio side-car, that way the music on end-credits etc can be controlled separately from the audio on, for example, an End Credit Promo.

You're quite possibly right, but if so it's just another indication of how the system is - shall we say - "Upton Park" (i.e. two stops from Barking!)



Yes, we use the lag fader, or make sure there's a second between the end of anno and start of programme to get the fader back up, and use the main TX220 PGM audio fader. Originally we had an even simpler solution. All the symbols (in the balloon days) were dubbed with pre-dipped audio at their con point!



And it's such a reliable system too!

"Upton Park"... you're being too generous



I thought the manual lag didn't work in Wales, or I'm I thinking of Scotland?

We use the manual lag fader virtually 100% of the time in the Pro-bel suites... In my earlier post I was talking about our Omnibus suite when referring to either using the manual lag (which doesn't work as well as it should) or using the automation in conjunction with the master fader.



It does have some good points erm... Oh! Time Shifting is much easier

Flicking through the BBC Nations channels over the last number of weeks, I notice a tendency for some announcers to raise the audio on the ident following their annoucement and leading into the subsequent programme. From the selection of junctions I watched, this practice seems to be restricted to a small number of announcers in Belfast and Glasgow. I'm not convinced that it works very well.

There's another group who tend to go to the other extreme and kill the ident audio completely during their announcements, leading into the next programme in almost complete silence; again, doesn't seem right.

Any views on this?

I prefer the audio to be faded back up following an announcement, rather than dipping the audio a lot during it. On some occasions during a voiceover the audio has only been dipped a slight amount and so is still very audible, but this isn't particularly common.

This just proves how, in the days of self-op announcers and people being employed to mix visuals live, things were a lot less complicated!