[bbctvtechop]

I'm hoping that one of our techy brethren on here can enlighten me.

Hello ☺ 

On 'Sunday (without) LauraK' today there was a breakdown on a DTL from Gaza. I know we've seen a quick glimpse of this screen before, but I never actually bothered reading it until today.
Apparently the interviewee can't see the studio output, just hear the audio. Yet when I use Teams/Zoom at work I can see the other parties, and we're all in HD as far as I can tell (with my weak eyes).
I am wondering why the BBC are unable to achieve the same two way video connection. It's always easier to talk to someone when you can see them and their reactions.

What you're seeing on Zoom/Teams is a automatically-generated composite of everyone on the call, or a single person talking, as captured by their webcam.

At the BBC the machines that run these calls are not located in the gallery, they are centrally controlled either by a media hub (World Service way of working) or MCR themselves (News workflow). The receiving audio and video is then patched into the internal routing switcher known as BNCS for any gallery to pick up and route to one of their Outside Source lines. The gallery also has the ability to request "2 way clean feed", which only one gallery can have at any time, and which basically pushes a mix minus from that gallery's sound desk, generated from the OS line it is on's fader, back through the BNCS network and then into that same computer in the Media Hub or MCR.

There is no technical reason why you couldn't send video down that as well, but traditionally all clean feed lines were audio only. When using more traditional technology such as satellite lines, MCR have to dial up the mobile phone of the correspondent and send clean feed down the phone line, as there are obviously no 2-way satellite lines.

The problem is you'd then need the gallery's vision mixer to send something - and what do you send? An isolated camera which may move, refocus or be stood in front of during the interview? What happens when the presenter needs to move to the next segment in a different place of the studio where the next guests waiting? Maybe you could do a locked off "ceiling cam" showing the whole studio, but this would be really poor quality.

The only alternative is to show the mix of the programme - including the guest when they are presented full frame. But now you're dealing with video delay - by the time the guest's webcam has been sent into the Zoom cloud, received into the BBC's Zoom account, processed through the internal BNCS network, processed by the gallery vision mixer, sent back down BNCS, received again by the BBC Zoom PC, sent back around the Zoom cloud, and finally received again at the guest's device, you could EASILY add 2 or 3 seconds delay. Have you ever tried talking into a screen where you see yourself back 2 or 3 seconds later? It is very distracting, to the point of stumbling over your words.

Hope this helps?