BBC World News | 30th October 2017 Onwards

I know the technical issues prevented WNA from DC because it sounded like line wasn’t able to be routed to the studio. But out of curiosity how many fiber lines (I assume they are as there’s little to no delay) do the big bureaus have sending to NBH and if so are they always “on” or are they scheduled? Like would a bureau like Washington DC and Singapore (seem to be the two biggest ones) have multiple lines sent to NBH? And for shows such as WNA (that I heard is switched in DC but graphics and packages are played out) and Newsday do they only send one camera line or two views for the director in NBH to choose?

I imagine if they needed to they could set use a either a rackmount or backpack CNG unit connected to the in house ethernet connection. Dejero and LiveU are the big ones here in the US. Dejero seems to max out at 20Mbps transmitting when connected to ethernet using MPEG-4 HVEC, while LiveU’s products are also 20Mbps using H.265/HVEC. I imagine the bureaus have enterprise internet connections and can easily max out the speed and have subsecond latency.

They do use IP for the feeds from a lot of their bureaus, I think the ones from Singapore and DC are a mixture of fibre and IP (which of course is ultimately over fibre too).

However, if they couldn't route the lines to air/to a studio, then they would probably have the same issue getting some sort of LiveU or Dejero (both are used by the BBC) to air as they couldn't route anything at all. They could presumably have set up a LiveU decoder on its own interface, but could they do anything with it?

Probably the one thing they could do with the and take the preview from the control panel using a dedicated Web PC - the quality may not be great but they’d get reports. Or I assume if this was a long lasting problem they could physically take the output from each Bureau into a LiveU or Dejero and physically place receiver to the rack (or if the CAR holds equipment for ever studio a long HD-SDI cable) and connect it to the the studios router. Depending on the type of server they can output four SDI feeds, however I’ve never seen BBC use anything more than a two way live shot.

Like I said the bureaus probably have significant internet connectivity so they could easily handle multiple CNG devices. If there’s not enough bandwidth they using all the cellular cards in addition to the office connectivity. . Here in the US more and more stations are using CNG their liveshots (even when they are working out of a ENG truck that can transmit using their own frequency for free*) in well populated areas with several other stations using the same networks and as many modems as they do next to them**. However I have never seen anything less than HD quality and sub second latency with cross talks.

*One LA station (I believe KTTV the Fox Owned station) was crippled when the Dejero software controller went down for an hour for some users. Several photographers had to call in master control and be instructed how to dial in a live shot even when they are working out of ENG truck.

**Many stations during breaking news situations will use CNG when they first arrive on scene. Once they are off many will set up their ENG trucks.

Edit : How are packages from the field or bureaus handled when they arrive to NBH. Are they connected to a single server where the RAVENS dedicated to each studio could play them out? Or are all the RAVENS (love the name as I am Baltimore Ravens fan) pooled and be routed to each studio using BNCS? If they’re routed using BNCS but connected to the normal computer network (which I assume the packages would be so they could be edited) the dedicated on air PC could hot roll it.

Ravens aren't used for main content Playout in NBH studios. The HD studios in NBH are largely based around a large Quantel (now GVG via SAM) ServerQ installation for ingest, editing and playout. This is interfaced with the Jupiter system (which includes a Quantel QCut) for desktop editing. There are craft QEdits driven by picture editors for higher quality editing.

Routing line-feeds into the Quantel ingest servers for a linear feed would have been tricky without BNCS control. File-based delivery shouldn't have been impacted as it goes into Quantel in the file-domain I believe. Playout is directly from server Playout ports and doesn't go via a dynamically controlled BNCS line-switcher AFAIK - so also shouldn't have been impacted (and didn't seem to be)

(ESPN in the US had a similar Quantel+BNCS - also known as Colledia - system at one point in the 00s, but have migrated to a different set of platforms now I believe) https://www.tvtechnology.com/news/espn-strikes-digital-deals-with-quantel-bbc-technology (Colledia was effectively the BBC Technology brand for Jupiter and BNCS technology ISTR)

AIUI a lot of the BBC's bureaux now use IP connectivity (though not all of this may be over the public internet). I'd be surprised if Dejero or WMT kit was used for fixed installs. Far more likely for NTT, Ateme rack mounted H264 encoders to be used. However the decoders for these devices are likely to be on the NBH station router - and putting in agile IP decoders and the requisite control system (which is likely to be BNCS-based...) would be a non-trivial solution and distract the finite engineering resource from solving the actual problem.

Sky News use a lot of Open Broadcast Encoders for their fixed IP links, and now run their MCR in the IP domain significantly I believe (I assume transcoding from H264 compressed to 2022/2110 though I may be wrong - so the routing is done at IP not HD-SDI level. I know BBC News have OBEs they use for temporary installs at Glastonbury etc. (they use OBE and Raven on a common hardware platform I think) and BBC Scotland are big users of OBE too.

Q
I was just talking about using Dejero or WMT as a backup type solution. A few market stations in the US actually put a receiver for the CNG equipment at their transmitter site in case they have to evacuate or if for some reason their microwave link to the transmitter is down.

What’s NTT?

I'm not sure how I see that helping in this situation? You are just swapping one IP receiver/server with HD-SDI output for another? You still need to bypass the router to get it onto the mixer/switcher (and audio onto the sound desk)?


Nippon Telegraph and Telephone Corporation - a formerly state-owned Japanese telecoms company. They are one of the market leaders in IP connectivity codecs. Their codecs are used by broadcasters for IP-based contribution circuits. NTT or Ateme encoders are used for many fixed IP paths AIUI.

NTT, Ateme, Harmonic are the three main broadcast IP encoder/decoder manufacturers I see in daily use. Dejero, LiveU and WMT are more newsgathering systems for bonded cellular or low bitrate WiFi-style working. They aren't high up the list for point-to-point fixed connectivity over guaranteed connectivity.

https://www.ntt-electronics.com/en/products/video/codec_system_solutions/index.html

https://www.ateme.com/solutions/dsng-contribution/

https://www.harmonicinc.com/products/encoding-transcoding-multiplexing/ (Harmonic are more geared to emission coding but do contribution coding products too AIUI)

Could they plugged it directly into the mixer crate and deembed the audio that way? I’m not sure how audios handled at the BBC but I think many US setups audio is embed via HD-SDI and sent to audio desk.

As for bypassing the router I would assume each studios router has the multitude of ports set up for OS that would normally be connected the “house router” (not sure what to call the one that takes all the OS and studio outputs) and in the case of it being down they could plug it into one of those ports.

Again this is coming from a layman.

Yes - for a source like an IP video decoder the output will almost certainly be an HD-SDI with embedded audio, which would be fed into the station or studio router (depending on whether the decoder is shared across the building or local to the studio)

If you were to install an IP decoder locally to overplug it and replace an output from the station router, whilst it's possible to do clever things with MADI for multiple outputs from some routers, I expect that a standard studio would take an embedded HD-SDI source in and de-embed it to AES audio for the sound desk (whilst routing embedded for other destinations). This would then be likely hardwired to an I/O device like a Calrec Hydra, that would then fibre into the sound desk processor. (Most broadcast sound desks no longer have I/O integrated into them, and delegate that to fibre stage-box type devices like Hydras - which can have analogue and AES baseband inputs)



The station router will have allocated destinations designated for each studio (if a studio has 10 outside sources from the station router, these are likely to be 10 destinations on the station router). These 10 outside sources are likely to then be fed into the studio router BUT in a system designed to cope with a local router failure, they (or at least some of them) are also likely to be hardwired to the vision mixer and sound desk for resilience.

Whether the interconnects between the station router and the studio are easy to re-patch is a different matter. It's also questionable whether you are not creating more problems for yourselves re-wiring a studio to bypass a fault, when you could be diverting effort from diagnosing and fixing the underlying fault.

And of course you could be creating more faults in the future as everything would have to be put back afterwards.

Is it common to have jackfields where things can be plugged around or temporary modifications put in place? Or is that a concept that belongs in the realm of analogue radio studios?

The audio rooms of UK OB trucks still tend to have a lot of jack-fields (as well of loads of electronic routing too ) European far less so. Old habits die hard in the UK. Video jackfields in the UK and Europe are getting rarer and smaller.