5G for the Music Industry
Podcast episode 46
A change is in the air for bands performing at concerts. Sennheiser is experimenting with #5G for its ultra-low latency at live audio productions. Dr. Jan Dürre tells our Michael Hainsworth why the future may be crystal clear for performing musicians, giving concertgoers a reason to get up out of their seats and cheer.
Below is a transcript of this conversation. Some parts have been edited for clarity.
Michael Hainsworth: The light at the end of the COVID-19 tunnel is coming from the stage of your next rock concert. It may be the last time your favorite band uses traditional wireless for the lead singer's microphone or the bass guitarist's axe. Industry giant Sennheiser is experimenting with using 5G as a replacement for that 1980s technology that once powered your cordless phone. 5G is more than just another G. It offers ultra high speed, high capacity and ultra low latency. It's that last feature, which has caught the attention of the head of research and innovation, Dr. Jan Dürre.
Dr. Jan Dürre: I’m part of a group that is looking at new technologies for wired and wireless audio transmissions in professional use cases. And if we are looking at new technologies, we often have to do feasibility studies and for that, we build prototypes to demonstrate if something is implementable and in what conditions, for example. This is also the connection to Nokia and 5G. Right now my group is looking at 5G as a potential wireless technology for some professional audio use cases.
MH: When it comes to 5G, we often think about it as this is something that's incredibly fast, that it can handle an intense amount of bandwidth. You can download a high definition movie, not in four and a half minutes like you do on 4G, but in 1.2 seconds like you do in 5G. But when it comes to your world, it's not so much about how many bits you can cram down a pipe at any one time. It has more to do with the ultra low latency that 5G offers. What kind of use cases does 5G support in a microphone world?
JD: I want to find out if it is feasible, if it can be applied to our use case and you're totally right. Our use cases are very different from many other use cases that are mentioned in the context of 5G. For example, what you mentioned, high bandwidth, something with video on the goal or the typical application for mobile technology of course, is a smartphone. I have Internet connectivity anywhere in the world, but there's a major difference between those use cases and the use cases that Sennheiser is looking at. This is for our applications, we require a deterministic network behavior. Latency is one requirement, but there are several others. In professional audio use cases, the user of audio equipment is used to the equipment to work and to have it working all the times.
MH: One of the key requirements then would be reliability. As I understand it, looking at this Nokia Sennheiser white paper, you're looking for reliability on six nines. It's got to be 99.9999% reliable?
JD: I mean, this is easy to understand. Think about, for example, I don't know the inauguration of Joe Biden.
Announcer: Ladies and gentlemen, here for the singing of our National Anthem accompanied by the President's own United States Marine band, please welcome Lady Gaga.
JD: Lady Gaga is singing the National Anthem and there is no second try. So the technology has to work in this environment and it is unforgivable if it does not work. So a company that is providing a microphone for Lady Gaga in this context, has to make sure that the technology it uses is valuable, it's working and it's appliable here.
MH: Well then, I guess this brings up an important point. As you stated right at the outset here, is that you're not necessarily a proponent for 5G. You're investigating whether or not this is going to be a viable technology for your industry. Why have this conversation in the first place then? Let me play devil's advocate here. Don't you guys already have a technology under the gigahertz bandwidth, a wavelength that works, that you've got an entire industry built around an existing wireless technology? Why upset the apple cart?
JD: This actually a really good question. Of course, the picture is very complex and you're right. As of today, we have working solutions, wireless solutions for microphones and also in ear monitoring. Let's think about an artist that is singing on stage, and he's using a wireless microphone because he wants to have freedom on stage. He wants to be able to move. He don't want to have cables attached to his microphone. The second important thing is that the artist that is holding this microphone also has an in ear monitoring system. This is very similar to a headphone, but more invisible. It's a small in ear headphone and this is used for the artist to give him feedback, so the artist can hear himself. He's singing into the microphone, the signal is traveling from the microphone wirelessly to a mixing console, and there other audio signals are added or effects are added. Then the signal is sent back to the in ear monitor of the artist.
This is necessary, of course. The artist has to hear himself and he has to hear other artists on stage or musicians on stage. So we have two wireless links in this scenario. One from the microphone to the mixing console, and one coming back to the in ear monitoring. Again, as of today, we have some technical solution for these two wireless links that are able to meet most of the... I'm saying most of the requirements. For example, one of the most important one is latency. For this round trip from microphone to the ear of an artist, there's a very strict requirement for the latency.
This can be explained if you think about what's happening there. The artist is singing into a microphone and the artist's signal starts traveling through the microphone to the mixing console and back to his ear. We have a round trip there, but there's a second path where an artist can hear himself. This is the one conducted in his head. In his ear, he can hear a mix of what is traveling via bone conduction through his ear, and also the one traveling through the wireless microphone and the in ear monitoring. If the time offset between those two things is too high, this becomes very irritating for an artist.
On the requirement here that we know from experience and from tests is that this round trip should not exceed four milliseconds. This includes two wireless links; from microphone to mixer and from mixing console back to the in ear monitoring, those two wireless links. Also, this includes some processing that is done in the mixing console. This is typically in a range of one to two milliseconds. This, again, leads to a strict requirement that we have for those wireless links and this is leaving only one millisecond per direction. This is something that we, as of today, can barely achieve. It is even that close that in many systems, we cannot do it both with digital RF technology.
So typically what you will see on stage is that the wireless link from the microphone to the mixing console is with digital RF and the link coming back to the in ear monitoring is analog RF. There's a major advantage of having digital RF and this is of course the quality.
MH: At the same time, that digital RF means that you've got to have a chip somewhere at some point, converting analog audio into ones and zeros, that takes time. And needs to be decoded from ones and zeros back to analog audio into the ear, that takes time. So we keep introducing more and more latency into the process. But as I understand it, 5G can operate not in the millisecond range, but in the microsecond range.
JD: Yes. In theory, it can.
MH: It's all about the deployment, right?
JD: I guess there are several factors involved, of course. As of today, if you're talking about 5G, can have this low latency. We are talking about the standardization. 5G is a standardized technology and the standard allows us to have this low latency. It explains how to achieve it. Of course, it's another question if this can actually be done in a real implementation and keeping all the other, well, conditions in place that are required in this scenario. For example, we do not only need this low latency, but we also have other requirements. For example, this reliability. And another one would be, for example, the efficiency. This is also very, very important because this touches the spectrum topic, always a hot iron.
But let's maybe talk one or two sentences about it because spectrum is a rare resource and this is everywhere the case in all use case. Especially in festivals, for example. Typically, today the available spectrum is a limiting factor on the number of wireless audio devices you can operate at the same time.
MH: Are you telling me that 5G, even though we still have to deal with wireless spectrum, and that's a whole separate issue, at least you could put more devices in a festival at the same time without worrying about that overlap?
JD: We have to measure, or we have to evaluate 5G as a technology with all these aspects in mind. Maybe this is coming back to what I'm doing right now and why we are talking about Nokia and 5G. It's because we want to find out if 5G is a potential technology for this environment. Of course, there are high hopes for 5G because from the standardization, this sounds all great and they are promising a lot. If we would actually be able to achieve this microsecond latency, this would mean that for example, we could maybe upgrade current systems where we can do both wireless transmissions digital. As of today, typically only one of those two transmissions; microphone to mixing console and mixing console to in ear monitoring is done digital. The other one is done with analog RF.
The reason for that is latency because in analog RF latency is no problem. The latency's not zero, but it is close to zero. But to not exceed these four milliseconds I mentioned before, today we are only able to do one way in digital. Again, doing it digital has many advantages. For example, the audio quality. As of today, well, producers decide that this direction for microphone to the mixing console is more important of course, because the signal is also used for reproducing it for the audience or transporting it for recording it and so on. It is decided to do this link in digital and the other one analog. This could be a benefit if 5G can meet all its promises, that we will be able to do both transmissions wirelessly.
MH: I have to say, I'm constantly changing out the batteries on my Sennheiser wireless microphones, out of fear they will run dead in the middle of a recording session. We've talked about 5G largely as a benefit to your industry for the low latency, but 5G in addition to high bandwidth, low latency and ultra reliability, also deals in the world of ultra low power for things like sensors for IOT, industry 4.0. Is there going to be an opportunity for me to not have to change out the batteries every time I want to use one of these microphones?
JD: Yes, you were mentioning they're doing ultra long battery run times in sensory networks. Well, what's the technology reason? How can they achieve that? And they can do that by doing very, very few wireless transmissions.
MH: Oh, it sounds like you're telling me I'm going to be changing batteries for the rest of my life?
JD: You see, there are some requirements that are not complementary, that are more opposing each other. I want a higher battery run time, but I also want very low latency, I want high bandwidth. You cannot have everything at the same time. I mean, this is typically with every technology. Question is if this technology of 5G is able to have a sweet spot for our use cases where all these requirements are met in a way that our customer is satisfied with it. Many of these questions we are just starting to investigate them. You already mentioned that we have a white paper with Nokia. In our journey in investigating 5G as a technology for these use cases, we came into a collaboration with Nokia. Nokia is one of the leading industry vendors for mobile technology.
What we did in our collaboration, is we took one of these requirements, this was the latency here. We tried to build a demonstration system to find out if 5G can really meet this latency requirement that is promised by the standard here. In this first work that we did with Nokia, we were able to show that 5G really can reduce the latency significantly compared to previous generations of mobile technology. Only looking at the latency, we are able to reduce it down to a level where it is interesting for our use cases.
Now, of course, the next question is what happens if we add a second requirement to that? For example, the reliability or something else. Are we still able to meet the latency requirement and also a secondary requirement? This is the next question we will be trying to answer I think again, with Nokia. We will continue our collaboration.
MH: As that collaboration continues, if we meet all of those KPIs, those key performance indicators, the checks are in the boxes for all of them, I can imagine at some point there's going to have to be a conversation about cost. The cost benefit analysis that works in favor of 5G over existing infrastructure. You've built an entire industry around digital wireless that would be upended by this proof of concept, proving that it is in fact more than just a concept.
JD: You mentioned a factor that's also very, very important. Our industry is also increasingly cost driven. Many of our customers want their devices to become cheaper. This is really also a hope, I think by many people that if you are able to use standardized technology in a use case, it probably is cheaper because maybe then there will be very highly integrated chips available because they can be produced for the mass market if it is a standardized technology. Maybe they are a bit cheaper then. But again, this is yet unclear. It's not sure if this will really be the case. It could also be a scenario where 5G, there will be several 5G chip sets. Some of them supporting what's done as of today in smartphones and some of them supporting what would be required for our use case. Then maybe this argument cannot hold.
MH: That's an interesting thought. That hadn't occurred to me. My assumption had always been that a 5G chip, is a 5G chip, is a 5G chip. Whereas if you stripped out the need for the systems to focus on high bandwidth, because you don't need high bandwidth, you need low latency, you could customize these chips. That although, further to what we were just talking about, seems expensive. So it strikes me that this is much like what we would see with any other new technology where the wealthy, deep pocketed component to any given industry will adopt it first. Those first adopters will be the ones who help drive the economies of scale that make it viable for everyone else down the road.
JD: A chicken egg problem here. Some of these features that are required for our use cases, these are not mandatory necessarily in 5G. We want to have this latency of, I don't know, microseconds, let's say. But this feature is not required for the mass market use case of 5G chip set, which is a smartphone. The question is, why should this be implemented in these chip sets? You see what I mean by chicken egg problem? It could be the case that these features are not implemented, thus these features are not cheaply available and thus leading to our industry not seeing the benefit of 5G as a technology here.
MH: As we work our way towards solving those bigger picture issues down the road of economies of scale, of making these chips available widely within any given microphone, or any kind of wireless gear, what's the very next series of questions that you need answers to before we can even get to that point? What's next for Sennheiser in your relationship with Nokia?
JD: The first question is always, does it meet our technical requirements? This touches the question of latency, reliability, efficiency, battery run time and so on. These questions have to be answered and checked to even look further, to look at business questions, or scaling questions, or spectrum questions. If we cannot check these first technology questions of are our requirements even met, it doesn't really make sense to look further. Of course, this can be paralyzed a bit, but these checks have to be made anyways. Maybe I want to emphasize the point where we are at right now, we are only at this first step at finding out if our technical requirements will and could potentially be met with 5G. Even with this first step, we only made the first step within this first step by looking only at one requirement right now. I expect that this is a long journey to go because there are still a lot of open questions.
MH: And it's a long journey for 5G as well. Many of the technologies that you're probably going to require once proof of concept is in fact proven, would be things like 5G network slicing, which doesn't come out until 3GPP release 17. That's two releases away. There is a journey that's taking place within the telecommunications industry simultaneously as you are going through that process, as well.
JD: Yeah, and here the picture becomes really much more complex. I mean, this touches the topic of operating mobile networks. Traditionally, or in the systems that are available today, the world is very simple because the musician is, or somebody that is directly associated to the musician, is the owner and operator of this wireless audio equipment. All these devices as of today are link based. This is fundamentally different from a mobile technology where you have a big, large base station that serves a lot of clients in today's wireless audio devices. Again, we have a link based approach typically, so you have one very small base station that is linked to one wireless audio device. This scales very, very simple. If you want to have two wireless microphones, you add a second small base station. If we think about mobile technology for these scenarios, this could mean that well, the operation of such networks becomes very different than before, but this is always the challenge to transition from one approach to another one.
But of course this also holds opportunities. If you think about a 5G system where we have this well, this base station approach where one base station could serve a lot of clients, why not use this one base station for several use cases or several different wireless devices? One section of this could be wireless audio devices, but you could also perhaps, this is maybe a vision, also use the same base station to operate wireless cameras. Then this might become very interesting because you only would have to operate one wireless network and could use it to operate several different use cases in it.
MH: It's going to be a concert scenario where in the past you'd have these giant transport trucks roll up to the stadium and they'd be loaded with audio gear, loaded with guitars, loaded with musicians. They would fill out that auditorium or that stadium with all that gear. Now in the future, you're telling me there's going to be another transport truck that's going to roll up. It's going to be a mobile base station that provides not only the 5G wireless for the microphones that the musicians are going to use in their guitars and all that kind of stuff, but the fans are going to be able to connect to it so they can stream their version of the concert to their friends over social media. The cameras can be used to provide that 360 degree experience for people who can't arrive and the security system's going to be tied into it too. You're going to have this whole separate gear to handle the 21st century concert.
JD: Yeah, this is maybe a vision on how this could look like if all the technical requirements for all of these use cases are met.
MH: Then maybe what we need to do is say thank you for your time, so you can get back to work to creating that future. It's been fascinating. Thank you so much for your time.
JD: Thanks for having me, Michael.