Watch the video here.
David Allen: Welcome to the Learning Lab. I’m David Allen here with Daniel Routman and on today’s Learning Lab we will be discussing placing audio and video over your data network or “AV over IP.” Now, we’ll be covering a lot of information today and that may bring up some questions. You may email those questions to email@example.com and we’ll try to answer all of your questions as quickly as possible. Now, Daniel, AV over IP, why don’t you wrap it up really quick, just a quick little sound byte to describe what we can talk about today.
Daniel Routman: Alright, so that means putting audio and video onto the network very simply put. What we’ve seen recently is this term “convergence of IT and AV.” We’re putting audio and video equipment onto the network, not really talking about the audio and video itself but just the equipment, connecting it to the Ethernet port.
DA: Like a projecter or something.
DR: Can you think of reasons why we might do that?
DA: I know that we put a projector on the network so that we can turn it on and off sometimes.
DR: Exactly. Control. Conventionally, we use RS-232 to control devices. Now we’re moving more towards Ethernet. Another reason too is for monitoring, so it’s something we do a lot especially for larger systems.
DA: Describe what you mean by monitor.
DR: So where we might want to keep track of how often something is used or be able to see if the room is on or off, and it also works very well for troubleshooting and for alerting people that, “Hey, this device fell offline. Go fix it before the customer needs to use it.”
DA: For instance, if those projectors that have the data network have a web interface, you can set up an email address. If your lamp hours get to a certain point, it will email a tech to go change the lamp or to order a lamp.
DR: Right, it’s very useful. So, we’re seeing more and more that the IT department is in charge of the AV equipment. Because it’s all in the network, they now need to be accountable for all that equipment. It’s part of the network. So, it only makes sense to also just put the audio and video on the network so that it’s centrally managed; they know how to manage the signals.
DA: So, over the last few years that’s what we’ve been moving into, not just controlling and monitoring but actually putting the signals on the network?
DR: Right, so let’s talk a little bit about putting the video signal itself onto a network. The term for it is streaming, you’re digitizing a video signal frame by frame, turning into ones and zeros, and so this is not new technology. It’s been around since the early 90s, and you probably used it today or maybe in the past few days. Can we have some examples?
DA: Netflix or YouTube?
DR: Exactly, so even the TV that you watch at home was, at some point, streamed. It was streamed across the globe, across a satellite that’s all using AV over IP. It’s also been in use and pro AV for a very long time too. Some examples being digital signage or IPTV, distributing TV signals across the network to receivers. Church live streaming is very big where people at home can watch the service from their computer, and also stadium video distribution is another example where maybe the game is being broadcast to all the concourse displays. Some of these examples are mostly multi-point meaning one source to many destinations. When you put this multicast signal onto a network it gets broadcast everywhere and that’s what professional ITs are concerned about. They’re worried about that bandwidth being on their network.
DA: Yeah, I’m sure there are a lot of our viewers out there that may be IT managers or IT directors that are saying, “You’re not putting that on my network!” So, what do we do to alleviate that?
DR: The big question is: Can you replace your AV switcher with the network switch?
DA: Can you?
DR: Of course we can. Why would we want to do that though? That’s the question. So, let’s talk about the network switch. I think this is a good reason why is that how reliable it is. These are some words to describe a network switch: reliable, widely used, expandable. It can be infinitely expandable really. If you need more ports, you add another switch and you connect them together.
DA: Okay, that makes sense.
DR: It’s secure. We depend on it and trust it on a day-to-day basis.
DA: Our data networks are secure.
DR: …right, within businesses or personal emails. It’s fast. We can send a lot of data across a network.
DA: And they’re just getting faster and faster too.
DR: They sure are. And they’re versatile. They’re not just used for your printers and your computers. You can, like we mentioned, use it for audio and video, and there’s so many other use cases that keep popping up every day. So, it only makes sense that, if you have such a powerful device that we know and trust, why not use it for AV?
DA: Okay, so we have a fast network switch here. It’s reliable, it’s expandable, but how do you actually get the audio-video signal onto the network?
DR: That’s something that’s been around for a long time: encoders and decoders. But now what we’re seeing are these specialized products that work for pro AV in a conference room environment. For example, you can do certain things like third-party control. That’s very important for us. We want to be able to control it from our Crestron, AMX or EXTRON control system. They’re fast switching. Maybe in the past you didn’t really need to quickly switch from one source to another, but, if you’re using it in an environment where you want to switch between sources, you need to make sure that it switches quickly. It does scaling. That’s something that we depend on in the AV world where you might have different resolution displays within a room, so you want to make sure that they look correct on each display. Or maybe someone brings in a laptop that’s not a standard resolution, you want to make sure that it scales up or down to fit your display correctly, so we’re seeing devices that have scaling built into them. And then, of course, RS-232, IR, Ethernet ports are built into the box so that you can control your endpoint or your source just like we do in conventional systems.
DA: That ties into the third party control and you can control all your devices — your endpoints, your displays and such.
DR: It does, yeah, and it’s all done over a single twisted pair cable that’s connected to your network.
DA: That’s pretty great! Encoders, decoders, what else do you need?
DR: Really that’s it. You plug it into your network switch. You can use the existing infrastructure.
DA: That’s fantastic. Well, it sounds like, though, there’s still that bandwidth issue that we have to address because I would assume that if you take your signal, especially high definition signals, and code them into your ones and zeros that’s going to be a lot of ones and zeros on the network.
DR: It is and that’s what IT professionals are scared of, so that’s what we want to we want to go over real quick. Look at this example here. We’ve got four encoders and four decoders plugged into a network switch. Encoders are what we use to encode signals. These are your video sources, like a computer or a TV tuner, and your decoders may be connected to displays like projectors and five panels. I’m visualizing the streams as these envelopes that are color-coded where…because IP data is really just packets of data, right? They’re just chunks of data, so each envelope is a packet of data, and they’re color-coded. So, what would happen if you put this on a network switch?
DA: You would have all your signals go to everywhere.
DR: That’s right, so all four signals go to all four destinations. Every decoder is getting the same four signals. That’s a lot of data. Imagine if this had had 30 or 40 encoders. That would be a tremendous amount of data and go into every decoder. All right, so what if in a real world there’s more than just decoders? You might have a printer and a desktop PC connected to it. Those devices would also get all that traffic. Your printer may be getting dozens of streams of traffic that might bog it down and shut it down, the network port.
DA: How do you fix this?
DR: That’s where there are a couple ways. One is is through VLAN’s. This is something that a lot of people use on corporate networks where you can virtually segregate the network. VLAN stands for “virtual local area network.” Through software we are segregating, in this example, the first eight ports from the rest of the network switch virtually, so there is no way for this traffic to travel down onto the printer.
DA: Okay, that makes sense, so you’re separating them virtually even though it is the same physical switch.
DR: That’s correct, and this, of course, requires a managed switch (a layer three switch) in order to do this.
DA: But you still have all those packets going to your destinations here.
DR: You do, yeah.
DA: But I guess it’s not coming here?
DR: That’s correct, so how do we manage this? There’s a lot of traffic going on in a decoder, so we use something called IGMP snooping.
DA: IGMP snooping, okay.
DR: “Internet group management protocol,” and this is a standard protocol that’s built into most network switches, virtually all of them, and this is what the encoders and decoders are using these days. What it does is it allows the network switch to know what channel the decoder wants to tune into. It’s listening in on the traffic between the decoder and the encoder, so that’s why they call it snooping. It’s listening in and it knows that that decoder only wants to tune into this channel, for example. Same with this second decoder is only tuning into that channel, and this third decoder is not getting any data at all because the network switch knows that it’s not tuned into any channels right now.
DA: So, obviously, you’ve really reduced your bandwidth based on that just by implementing the IGMP snooping.
DR: Exactly. There should only be one stream going to every decoder.
DA: That’s all it could handle at one time, right?
DR: Well, it can handle more than that.
DA: Ok, so you can tune into multiple channels if you wanted to.
DR: You could. Some decoders have the ability to be daisy-chained where there are two network ports on a decoder, so you could have two decoders coming off of this. You could have several.
DA: And they could be listening to different channels.
DR: In that case, your networks would intelligently route just the channels that need to be tuned into.
DA: All right, that makes sense.
DR: You’ll notice here that I’m not using VLAN’s, so this is still one VLAN, and the switch knows not to route any traffic to the printer and the computer. If you use this IGMP snooping along with VLAN’s, that’s two layers of protection that we can put in place that ensures that this video traffic is not going to interfere.
DA: Makes sense. Let me ask you a question. If we have the PC traffic coming into the switch, it also uses IGMP or the switch uses IGMP snooping to not have that PC traffic go to the decoders as well, right?
DR: No, not necessarily, because most traffic that comes from a PC is going to be unicast. It’s going to be point-to-point communication. If this computer wants to print a document, it communicates directly to that printer so that traffic won’t go anywhere else anyway.
DA: Just when you’re multi-casting, like we do with encoders, that’s what happens.
DR: That’s just the nature of an encoder. It’s always going to be multi-casting everywhere. This is what stops that from propagating.
DA: Ok, so what’s next?
DR: So bandwidth is still something we…
DA: Ah, the bandwidth question, so how do we deal with bandwidth?
DR: In the previous slide, we covered how we manage the flow of traffic, but there still may be cases where there’s a tremendous amount of traffic going through there. A network switch that we use typically in a corporate environment is a gigabit switch where every port can handle a gigabit of data. Any kind of encoder that you use today is going to be able to go on that switch without any problem. Some examples of encoders that are used today are H.264. That’s a very high efficiency encoder, low bandwidth, but somewhat high cost sometimes. There’s also JPEG 2000, which is not as efficient, it uses a lot more bandwidth, but the plus side of that JPEG 2000 is that it’s low latency.
DA: Okay, and what you mean by latency is the time it takes to encode and decode, right?
DR: Exactly. The time it takes to go from point A to point B. There might be cases where, if you’re doing live video of a stage that you’re showing on a projection screen, you may not want to have that large amount of latency.
DA: You wouldn’t use an H.264 encoder in that case. You would use the JPEG 2000. Latency is the time it takes to encode it and then decode it at the other end, so that time it takes at the beginning into the end.
DR: Right, there’s a delay to encode a signal and then to decode a signal. It takes time to do that.
DA: So, H.264, you would want to use that type of encoder if you’re watching something live and seeing that decoded signal on a monitor or projector at the same time.
DR: Right, or another good example is when you’re using a mouse and a keyboard within a room. If you’re looking at the projection screen and you move your mouse, any kind of latency is going to really throw you off.
DA: Gotcha, are there any other kind of protocols for encoding and decoding codecs?
DR: We’re seeing a lot of proprietary codecs being used today where they claim near zero latency.
DA: Those are specific to manufacturers, I assume, meaning you have to have the same encoder and decoder from those manufacturers.
DR: That’s correct. Typically, what you see is that the lower the latency the higher the data rate (the bandwidth) that’s required. We can put all this onto a gigabit switch without any problem. You could populate a 20 port switch full of encoders and decoders and through the use of IGMP snooping and VLAN’s it can handle it. But now what if you want to expand that? What if you’re going to a second switch that’s maybe located on a different floor or a different building? That’s when you have to design it correctly to make sure that it works. For example, if you used a copper connection at one of those gigabit ports to go from one switch to the next…
DA: …you only have a gigabit connection between those two switches.
DR: Right, so if you’re using JPEG 2000 and coders, for example, you may only be able to get five of those streams from one switch to the next.
DA: Okay, so it’s very important then that the system is designed properly based on how it’s going to be used.
DR: Right, so in that case we would design it using H.264, or, better yet, we would use fiber connections between the two switches.
DA: Just increase the bandwidth between those two rooms or floors. Let’s talk about advantages and disadvantages. There are obviously advantages and disadvantages in everything that we do, so let’s go ahead and talk about the advantages first about using AV over IP; or are these encoders and decoders?
DR: Sure, the first thing that we can talk about is expandability. We already touched on that earlier where networks, which can be infinitely expandable, right? The same can be said about AV over IP. If you want to add additional sources to your tier system, you just add some encoders. Or if you have displays you add decoders. In a conventional system you may have to upgrade your switcher to a larger size.
DA: If all your inputs or outputs are full.
DR: That’s right, very cost prohibitive. It’s future proof, so I think we can both agree that the network switch is here to stay. A lot of the infrastructure that’s in place today and businesses are going to be here for a very long time. The network jacks that are in the walls are probably not going to change for a long time as well, so we can take advantage of that and use that technology that’s already in place. Anytime, let’s say in the future, they come up with a replacement for HDMI then it’s just a matter of replacing your encoders with a different box that handles the new type of signal.
DA: As opposed to replacing your whole infrastructure.
DR: Right. Another one is low cost. You could argue this either way really, but we’ve done some price comparisons with conventional systems, and we’ve seen that, especially when you’re dealing with systems that are a few sources to many destinations or many destinations to few, is really where you see a huge cost savings.
DA: As we actually interview our clients and discuss their use cases and the types of solutions they’re looking for, we determine what’s the right fit based on their existing infrastructure or their data network and things of that nature to determine is it going to be a low cost advantage to them or not.
DR: Exactly, and, of course, you see a huge cost savings when we can use the existing infrastructure.
DA: Absolutely, I could see that.
DR: We just want to make sure we talk to the customer and know that their network can handle it. Talk to their IT department. Pooled resources is an interesting one that we’re seeing a lot lately.
DA: What does that mean?
DR: It means sharing one resource among many rooms. A good example of that is video conferencing where you could have multiple rooms equipped with cameras and microphones, but they only share one video conferencing codec that’s located in a rack closet somewhere. It makes sense for customers who only use video conferencing occasionally. A room can book that video conferencing unit. They can schedule it ahead of time to use it at a certain time of day.
DA: Maybe even clients that are looking to just get into video conferencing and they don’t want to buy one for every room yet. That might be a good solution if you have the infrastructure to be able to support it. What’s another advantage?
DR: System-wide AV. I mean that comes naturally, right? Because the network is system-wide, it could be across an entire floor, a building, campus-wide or from a city to city. As long as those two locations or more are all interconnected, then you can send AV from one room to another.
DA: On the same data network it’ll go there.
DR: Right, so with a conventional AV system you’d have to home run everything to a central location. Now you just use the existing ports that are there.
DA: Perfect, any others?
DR: Centrally managed. We talked about that when we talked about convergence of IP/IT and AV. Now we can add audio and video to the mix. The IT department can now manage the flow of AV. They can monitor what’s going on. They can actually see and listen to the circles.
DA: It makes it so much easier to be a tech support for your company or industry. Let’s talk about the disadvantages now. What are some of the disadvantages?
DR: On the previous slide we mentioned that low cost is an advantage, but the same can apply the opposite way.
DA: So a solution could be higher cost depending on their scenario. For instance, if the client did not want to use their existing data network and needed to put it in a completely new network, obviously you’re paying the cost to do that.
DR: That’s a great example.
DA: Or even if you want to highlight and see low latency, things of that nature, it could affect the cost. I think that’s where it’s a benefit to have a really good designer on the team to be able to help the client work through those decisions as to what would be best for them.
DR: Yeah, I agree.
DA: What are some other disadvantages? Oh, network outage…
DR: Right, so, of course, you depend on the network. The network goes out, so does your AV, but what we’re saying is that we typically put more attention into our network and keeping it running than we do our AV equipment.
DA: That’s the worst nightmare for any IT person regardless if there’s AV on the network or not. And you’re right, that’s when they invest a lot of money to make sure that does not happen.
DR: What we typically see are the network switches on battery backups whereas you may not see that on your AV switcher.
DA: What’s another disadvantage?
DR: Quality and latency. Whenever you encode a signal you’re compressing it. You’re losing quality. And then, when you decompress it, you’re trying to return it back to the original state. Some manufacturers claim zero quality loss, but, when you do some high compression to bring that bandwidth down, you’re going to lose quality. In most cases, though, it’s not noticeable to the human eye. There are cases where maybe it’s a poorly designed system and you have to cut down on the bandwidth, so you’re turning down that quality a little bit.
DA: So, again, it’s very important that you have a good design and it’s executed well. I think a lot of us are probably familiar with this, especially those who have digital satellite systems. They experience quality issues whenever it rains for instance. You can see the drop out, so, obviously, you can’t control the rain, but you can control your infrastructure, and we have design when it comes to your network.
DR: That’s right, and then we also touched on latency, but you want to design it correctly. You don’t want to put in the wrong kind of encoder when you need that low latency.
DA: Depending on the scenario, yes.
DR: And then security.
DA: What do you mean by that?
DR: It’s as secure your network, right? You’re putting video on the network and you’re broadcasting it, so anyone who has access to that network may be able to see what you’re presenting for example.
DA: As long as they have the right equipment.
DR: Exactly, they’d have to know what they’re doing. You don’t want to be broadcasting sensitive information on the network if it’s, let’s say, the executives are on one floor, you may put that on a separate VLAN so the rest of the building can’t tune into it. There are ways around that but, externally, people can’t really come into your network. It’s as secure as your network, as I said.
DA: So, again, it goes back to proper design. As a matter of fact, all of these disadvantages can be alleviated with a proper design, so that’s one of the issues that we obviously like to deal with with our client.
DR: Right, it doesn’t have to be a scary thing at all.
DA: Daniel, thank you, you’ve given us a lot of information about AV over IP and this exciting technology that isn’t just converging on to the IT network, it’s actually converged. AV has converged onto the IT network, which is fantastic, and it makes us excited because of how the technology continues to grow. We’ve covered a considerable amount of information today and you may email any questions that you have to us at firstname.lastname@example.org and we will try to answer those questions as quickly as possible. This is all the time we have for today, and we hope to see you next time on The Learning Lab.