Will BMD release a IP based video switcher?

Will BMD release a IP based video switcher? SMPTE ST 2110? This article predicts there will be an influx of IP based switchers at NAB 2018 http://www.tvtechnology.com/expertise/0003/challenges-remain-for-live-ip-production/282414

http://www.tvtechnology.com/resources/0006/smpte-st-211010-a-base-to-build-on/282238

I won't count on it in the near future.. BMD is heavily invested in the SDI market. Also SDI is still cheap and "simple" for most novice users.

There Tico IP unit is the first SDI over IP product and who knows what will come.. But the cost of conversion to IP from every cam to the cost of SDI straight or by fiber to the Atem is still cheaper.

Also TICO and NDI are 2 different approaches.. And while Vmix, Wirecast, and all computer software based products go the NDI route.. BMD goes the TICO route which is not wide spread supported..

So i think we will see some more in 2 or 3 years.. But heavily depends on what the market will shift too.. And also what camera manufacturers will do.. If a lot of camera manufacturers keep it with SDI i think IP will not pick up in the upcoming 2 to 3 years..

While IP has some advantages over SDI. I mean see Greg from Mamba fiber with it's Dante Audio network. He can drop in a microphone or whatever somewere in his IP setup.. And always comes in on the same inputs on the Audio mixer.. Which is rather cool!!

Wait and see..

Will BMD release a IP based video switcher? SMPTE ST 2110? This article predicts there will be an influx of IP based switchers at NAB 2018 http://www.tvtechnology.com/expertise/0003/challenges-remain-for-live-ip-production/282414

http://www.tvtechnology.com/resources/0006/smpte-st-211010-a-base-to-build-on/282238

Whilst it might be convenient for BMD to support IP video directly on the switcher, it would be a major leap of faith for them to completely re-engineer what is perfectly good architecture to potentially support *every* IP Video standard out there. Perhaps in the fullness of time, one IP Video format will become the obvious partner for ATEM productions - but in the short term, its pretty easy to simply use conversion interfaces at the switcher end. This gives all the benefits of an IP based cabling setup, need not cost very much and its available now.

As an example - here is a diagram of a fully IP based live production workflow which uses an ATEM Switcher, and using BMD DeckLink Duo 2 with NDI Outlet Multi on a MacMini to convert the IP Video sources from 4 of the new Panasonic NDI IP Video cameras. As you can see there is just one cable running from the 'stage' to the 'switcher' - and this carries all the video, plus the potential to control the PTZ cameras, and to send Tally etc to them. All in all, pretty slick, and we dont have to wait for BMD to re-engineer the ATEM.

And for anyone wondering how this would look with BMD cameras, here is another picture.

a Mac mini has enough juice to deal with 4 3G SDI inputs/outputs? do you need the quad/i7 or will a lesser model suffice?

some of us run much much more than this now with one piece of cable between control room and cameras...

I didn't realize that existing owners of AW-HE130/AW-HE40 cameras have an upgrade path from Panasonic to add NDI...interesting............

a Mac mini has enough juice to deal with 4 3G SDI inputs/outputs? do you need the quad/i7 or will a lesser model suffice?

some of us run much much more than this now with one piece of cable between control room and cameras...

I didn't realize that existing owners of AW-HE130/AW-HE40 cameras have an upgrade path from Panasonic to add NDI...interesting............

In a test I did here earlier, On a 2009 MacPro 3GHz, one NDI Outlet channel takes about 50-75% of a hyper thread core (this mac Pro has 8/16 cores).
a much more modern 2018 dual i7 MacMini top spec $999 should be good for at least 4 channels, probably more. I don't have any experience with the lower end minis, however.
On the NDI Source side, NDI is actually more efficient on Encode than decode, if you can believe it, so the same sort of equation follows.
NDI is incredibly CPU efficient. That is one of its most important features.

So all I need to replace 4 simple coax runs is 2 computers, 2 TB-PCI interfaces, a couple capture cards, and a managed IP network? For this "convenience" I get latency and huge quality hit? My workday just got so much simpler. This technology is interesting for sure but I'm still struggling to find a way that it's interesting to me personally.

So all I need to replace 4 simple coax runs is 2 computers, 2 TB-PCI interfaces, a couple capture cards, and a managed IP network? For this "convenience" I get latency and huge quality hit? My workday just got so much simpler. This technology is interesting for sure but I'm still struggling to find a way that it's interesting to me personally.

Hi Howard !

Clearly IP Video is not something for everyone, but it's becoming more and more important in many sectors.

Its not fundamentally altering *what* we do, but how we do it, and it's the possibilities to expand and enhance workflow which make it important.

For example, in the diagram above, imagine you replace the Single Gigabit connection in the middle with a wide area network connection, via the internet - now you are doing remote production, with the same setup.
Or maybe you want to remotely control PTZ cameras over the same network, or you have other devices in the mix which are creating computer generated graphics to be fed into the mix, or you want on-demand monitoring or display of the sources *anywhere* on the network (for MAGI or other display screens), or you want a low latency monitor on an iPhone to carry around when setting up, or you want to take 'backstage' feeds from iPhone cameras.. the possibilities are endless - and that is the point.

If you are just sticking to hauling 4 cameras feeds back to one location over a fairly short distance, then you are right, maybe IP Video is simply not relevant to your workflow. However, once you try IP Video, your mind will start to revolve and suddenly all bets are off in terms of what is and isn't possible in future workflows.

Several years ago I asked Evertz what was driving their interest in IP Video. I was surprised by the reply:...
'We need to enable a bigger video router' - just that.
Apparently 1024x1024 (which was the biggest video router they made) wasn't big enough for some of their customers. So, in this case the IP video transition was primarily based on scalability. Now they make IP Video systems scaling beyond 2048x2048 if you can believe anyone needs that (they do).

Equally, other vendors and broadcasters have different motivations, but its really all about workflow flexibility, and the ability to break away from existing constraints.

To address the 2 points you made about latency and quality, some IP Video (eg SMPTE 2022-6) protocols are exactly the same as SDI in terms of quality and latency, just more flexible in terms of routing. Other IP Protocols like NDI take a different view, and use gentle compression to allow for simpler and less expensive networking, and more convenient integration with different compute devices. If you looked at the NDI images, you would not see any compression artefacts unless you looked *really* closely - its certainly somewhat better quality than most of the compression systems used in most professional video cameras. In terms of latency, NDI can technically go as low as 8 picture lines of latency, although in practice it's usually a frame or so. This is generally rather less than you are getting on your LCD monitors and projectors, and over the 100m run its many times faster than the audio you are getting back from the speakers on stage through the air.

I like to draw a parallel with MP3. 10 years ago we all had a huge stack of CDs and we had to keep changing the record and could never carry them around with us, or even search !! - now we have all accepted a *tiny* compromise in quality, to reap the huge benefits offered by having a tiny device in our pocket with all our music on it. To be able to buy an album and have it instantly on play, and so on. It will be the same paradigm with IP Video. Maybe some tiny compromises (or priority shifts as I call them), in return for a transformation of the overall experience.

Part of the challenge of IP Video is clarifying what it means to different people and this sort of discussion is really, really important. As you say, it may not be interesting to you personally......yet !

But give it a try and see. There are lots of completely free NDI tools for Windows and Mac for you to experiment with, and maybe you will be surprised....

And to return to the basic topic - clearly ATEM users will not want to be left out of the IP Video revolution, and just because the ATEM switchers don't currently have IP built in, it doesn't mean they can't take advantage of this avalanche of new ideas

Part of the reason this workflow seems overly complex is that when you have a base band switcher, it will always be simpler to work with base band sources.

However, if the switcher was IP native and the camera sources were IP native, then you wouldn't need the conversion gear in between. In fact, it would probably be cheaper since Ethernet cable is cheap and easy to find anywhere (in some cases it is already run for you), and a network switch is orders of magnitude cheaper than a video router.

Kane Peterson

And just to finish up this Hybrid IP Video workflow topic.
NDI Outlet Multi is now shipping. So you can take up to 8 channels of NDI and convert to SDI with a single box using BMD Duo2 or Quad2, to feed into your ATEM. Suddenly Hybrid IP workflows are sounding easy.
http://www.sienna-tv.com/ndi/ndi-outlet-multi.html