Distributed production: Distributed or disrupted?

This episode of FEED Forum sits down with Sergio Ammirata P.h.D. of the Rist Forum, and Ian Wagdin of Appear, who explore the evolution of remote, hybrid and on-prem production in broadcasting, its benefits, challenges, and future prospects, including standards, AI integration, and global collaboration.

Transcript

SPEAKER_00 0:04

Hello, it is Neil Romanic from AU, and welcome to another feed forum. Um, and today we're gonna be talking about distributed production or remote production, as you may have also heard it called, uh, or as you uh probably may have called it in uh pandemic, just a zoom call. Um there were all kinds of tools for digital or distributed production, remote production, sitting around for ages. Uh, people talked about this possibility of like, you know, you could just do a lot of the stuff remotely. You wouldn't have to like fly everybody out to to do things on location. You could have somebody doing something through a browser, you know, at a completely different location, maybe sitting at home in their living room. And those things were kind of just sitting around, kind of being used, sort of experimented with. Uh, then the pandemic came, the the height of the pandemic, and most industries, and certainly the broadcast industry, had to rethink what they were doing like by Monday, like immediately. And so, luckily, a lot of these tools were lying around and people just started sticking stuff together to see if it worked. And some of it worked. A lot of it worked, surprisingly. I think a lot of us found, and no matter what industry we were in, that a lot of these remote tools and online tools work really well. Remote production as we know it really got a kickoff during the uh during the pandemic. Um, we're gonna talk today about what happened after that, where we are with remote and distributed production and live, what that really means, um, and what um tools are available that we can use in the future and and how those can be democratized and and used for anybody doing any kind of production. So I'd like to thank my guests. If you guys had introduced yourselves, I got Ian Wagden and Sergio Amorata. Um, Sergio, tell us quickly who you are, and let's get into this conversation.

SPEAKER_02 1:49

I am uh Sergio Amurata, I'm the chief scientist of Sib Radius, and today I'm here on on behalf of the Wrist Forum. Um, one of the directors as well. It's an open standards organization uh that uh you know manages the marketing of the Wrist uh protocol, one of the ones that was born uh for remote production, so to speak.

SPEAKER_00 2:11

Amazing. And Ian?

SPEAKER_01 2:13

Yeah, hi, Ian Whitkin. Um I'm VP of technology and innovation at API, uh, and we make encoders um and decoders uh that a lot of people are using to um enable remote production workflows right across the globe.

SPEAKER_00 2:27

So both of you guys were people working on, as we were saying, the technologies that were just shutting away, being built by people like you. So that when somebody decided, oh, we need some alternative, the alternative was was kind of ready to go. Um I guess looking at the uh the pandemic, I mean we kind of appointed the pandemic as like sort of a turning point for when people really um accepted the idea that you could do uh production, browser-based production, if you will, and remote production, you know, distributed production. So that was a real thing that you could do, don't freak out. It's actually anything you could do. Um how would you say things have progressed in that time? I've heard people say that you know, people kind of you know, in the in the couple years since or a few years since people have gone back to location production, kind of abandoned remote production as a thing, uh that has kind of changed in ways that people hadn't expected. What what is your take on like where we've become over the past say five years?

SPEAKER_02 3:28

Well, I I think that there there are two separate um you know topics we should cover on that question. Technologically speaking, where where have things gone? And business speaking, right? You know, what what have companies actually done or not done and how the two have influenced each other, right? Because one of the things that I want to elaborate on one of the things you said, you said as you know uh literally people were you know sticking stuff together, right? During the pandemic. The pandemic broke all the rules, all the organizations, large and small, they throw, they threw the rule book out the window. All the checks and balances, this is you know, broadcast, or this is this, this is that. I can only use this type of format or technology all went out the window during the pandemic. So with that fresh mind, with that open mind, there were a lot of holes that were filled by different types of solutions. Some very good, some not very good, but they worked to some extent. And then what happens after that, right? You know, the pandemic is over. Okay, now let's calm down, let's put now the rules back in. But the rules don't fit what they were doing. So what happened next? You know, that that's a good topic to cover from a technological point of view and from the business point of view. What did the customers do? What did the broadcasters do? You know, what did the vendors do after that, you know, to to catch up? You know, vendors like you know, uh Ian or other encoder manufacturers have some of them missed the boat because they were not agile enough during the pandemic. Uh me in particular, C B Radus was the other company that I have, pretty uh nimble. So we were able to adapt and we scored some gigantic points in the broadcasting industry, and now we're very well known there. But that was all thanks to the pandemic, right? And then there's the organizations that were writing standards that realized, okay, we need to fill in this gap. How do we do it? You know, what are the protocols that we should be using? So I I think that we should elaborate on all those different aspects of what happened after the pandemic.

SPEAKER_01 5:45

I'll talk a little bit to the the the kind of technology side and actually how the technology side has has changed because I think when we went into that, there was a there was a big panic. And um there were all sorts of solutions, as as Sergio says, sort of you could either have a piece of hardware on location that you remote controlled, or you could try and run a virtual machine um somewhere and and then control that centrally. Um and I think what it did was it opened people's eyes to actually we could we can work in this way, but we haven't really got the ideal technology stack in order to enable that. And then since the pandemic, we've really started to think about how we architect our our technology so that we we give people options and we give those editorial teams choices, whether they whether they want to be on location or whether they want to be working working remotely. Um and we we know that I mean long before the pandemic, we we saw um big broadcasters sort of working around things like the Olympics and stuff like that, where it was getting expensive to fly whole big teams out and and do that stuff. So how do you bring back world feeds and maybe have you know logging teams or or certain teams not out on location? And then it's just the really the balance of how many people you send versus how many people you keep at base. Uh and and I think that's what's what's changed is the technology enablers and the way that we're thinking in order to deliver that.

SPEAKER_00 7:00

Uh that's really interesting that um that there are those those those tensions that technology and business tension. And even in the business tension of, you know, we've made some investment, we're gonna do things this way. But then the other side going, oh wow, I hadn't even really realized we could we could change this and save a whole bunch of money immediately. You know. Um and um yeah, I wonder what are your experience, especially as a vendor, maybe Ian, like what what's that been like, that kind of push and pull of like we've made an investment, but if we could, you know, if we could make a leap forward suddenly, that would save us a lot of money.

SPEAKER_01 7:36

Yeah, I mean I um my my previous role I was I was actually on on you know working for for a large broadcaster, and and we kind of we saw um that the reality is when you speak to sort of editorial people on the ground, the the the the the the point was they always wanted to be there, they wanted to feel they were part of the action, they wanted and I think what the pandemic proved was actually that didn't need to happen, and then they could choose still produce a high quality product, perhaps without having that, you know, all the editorial teams meeting for breakfast and you know having a chat and all that. That just happened in in the Zoom calls and and the stuff like that. So people just found new ways of working, and it also enabled a whole bunch of new talent to come in as well. So people who perhaps couldn't travel or didn't want to travel, so we've increased our sort of diversity in roles. Um we've you know there's a lot of people changes that go alongside just just the the the fundamental editorial changes. Um and in a world where people don't want to fly or don't want to be away, you know. If you're looking at World Cup and Olympics, you you can be away from home for for a month, you know, um uh sometimes longer, six weeks, by the time you've done the setup, the event itself, the teardown. You know, I think the the World Cup's actually two months, you know, um, all in all. So it's a big commitment from staff and young people and the talent that we want to bring into the industry are kind of a little bit more well, actually, I I I I I need a life outside of my work environment. So we it enables us to um meet some of those demands as well.

SPEAKER_00 9:04

So um uh I guess looking at the uh get moving away from the pandemic, which we looked at as kind of a, you know, was sort of where there was a revolution, or at least a push towards a revolution and and how we did content, um now remote production seems to be, you know, on in in everybody's toolkit. Um and I'm wondering kind of what your take is on how it's being incorporated and kind of what the upsides are really. Um I know Ian, you mentioned uh obviously the ability to like be able to stay at home, which is as a big deal for people wanting to have a life outside of their you know broadcast uh job. Um but there are there do seem to be a lot of upsides. So can you maybe get into some of those?

SPEAKER_01 9:55

Can you start on that? Um so you know, obviously there's a s sustainability aspect, um, which means we're not shipping kit um you know across continents, across the world, across the countries, however, however that is. If we can sort of send less kit on location, that's that's gotta be good for um for the planet. Um the the ability to do more than one event with a with a capital investment on on uh you know on on your hardware control surfaces and things like that's quite a lot important to to some people. You know, you might be able to cover two or three events in you know with one gallery rather than effectively what we were doing before, which is kind of shifting the gallery around um to where the events take place. Um the flip side of that is we're more reliant on connectivity than we than we ever have been. Um but connectivity is more ubiquitous now, it's better, it's faster, there's more of it. Um so a lot of people are starting to to realize that actually the balance there has kind of flipped. Um and and and so, you know, with reliable connectivity, with the tools that are now available, um, with the people who are more used to working in that way, it it feels like it's becoming more real all the time.

SPEAKER_00 11:03

Um Sugio, do you want to chime in on that?

SPEAKER_02 11:07

Yes, I mean from from a standards point of view, uh we've evolved into kind of a mess uh of uh REMI, you know, where there's 50 different ways of doing things and they don't talk to each other. And most of them are you know company specific or vendor specific. And even though there are standards uh that allow you to do all the different pieces, those standards have not evolved fast enough to have an orchestration layer that can stitch them all together. And it's it's a big problem right now because somebody that wants to uh start or get into this uh REMI production mode, we don't know where to start we don't know if to pay a fortune to company A or to try to use their team to stitch everything together with open source or an open standards or some something in between. So uh even the companies that embraced it during the pandemic and it worked have gone back and said, okay, how would we do it better? How would do we do it more reliably with more security? And uh and at the beginning, security was thrown out the window, you know, just make it work, right? And now that they have to put it in, how do they do it? You know that that's that's where we stand right now in this wild west of remote production. It works. Some solutions are more secure than others, some are more reliable than others, some recover better from packet loss or from poor connectivity. And then you know, people are confused. So you you you talk to five companies, some of them will tell you it doesn't work, the technology is not there, others swear by it and say we do everything you know on a REME based.

SPEAKER_01 12:59

I think it's interesting that actually, you know, on that standard thing, there are several different ways of doing this, yes. So so I know several companies um here in the UK who have SDI-based infrastructure and they make that work remotely. Um, we also know that the IP stuff and the transport layer stuff and the 2110 and the JPEG X's, that all works. But as we now move into to cloud, we need to start thinking about actually those compute um the data transport. And we're we're just starting on on that journey now with uh work we're doing with the dynamic media facility, kind of identifying you know how data moves around between applications, between vendors, between um different hosts, cloud solutions. Um, and and so we've we've really got these kind of three models of which there's a hybrid. You can you can either do it in a in a kind of digital way with SDI, you can do it in an IP-based way, um, using 2110 or or or equivalents. Uh NDI is obviously another one that that some people use for this. And then you when you when we move into software and the cloud, there's a completely different architecture in that in that space as well. So people really have to think about actually what does my end-to-end workflow look like? And and one thing about standards, I think we're we're we're quite guilty at just looking at one part of the workflow rather than looking at the the whole solution. Um and that's perhaps what we need to do as an industry is just take a step back and go, actually, the whole end-to-end ecosystem looks like this right the way from the lens of the camera right the way through to the the piece of glass that people are watching now.

SPEAKER_00 14:28

Uh because it is, yeah, it's like so are there any like best practices established? Because you have that in in the same way, I guess, that when people were pushed into doing stuff from the pandemic, um, as you had mentioned, it was like, well, I guess it's gonna be Skype or Zoom or whatever that we're gonna do this live broadcast that normally we would have brought cameras in. And so these, I mean, I don't like the word, but some of these these processes have become democratized, if that's a word, where a lot of people have access to them, uh, a lot of uh different um parts of the sector outside of traditional broadcast start to have access to them and develop their own ways of doing things. Or if, you know, if I'm a if I'm a brand or if I'm a sports league or whatever, I might be able to grab these tools and just go running with them. But it seems like you might have people scattering in all different directions finding what works for them or what doesn't work for them. They may be wondering, how do I make this thing work and have no idea? So is there is there is it possible to develop best practices, or is everybody gonna have to just sort of sort out their own particular use case themselves?

SPEAKER_01 15:32

I I I think it is. Um I think there's some some great work going on. Um mentioned the dynamic media facilities, the EVU, around defining that. But I think we've got um certain challenges. So so one thing is production moves fast, yeah? So every single time there's a new production, somebody wants to try something new, somebody wants to do something new. So it's sort of pace of change in that environment. And that could be driven by an editorial reason, could be driven by the technology availability, could be driven by um you know a technology availability that's that's suddenly come out. So we we'll see iteration upon iteration and upon iteration of of these these kind of working um ways. And every way, every time you you take that step, you might evolve into the next step of the chain. Um and we'll see that accelerate as we move to software because you know you're no longer constrained by you know what you own as a capital investment and as you kind of move forward. But as I say, the key thing for me is connectivity. If you haven't got connectivity, you you haven't got anything. Um yeah, with an OB truck, effectively you take your own connectivity on site and you kind of have a satellite and you just think about it like that. You know, but all of these things are are moving so fast now that it's possible to bring back 50 cameras from you know an NFL game into a centralized production system. I think there's also um difference depending on the territory you're on. So if you're in the US, um you've got one big country with games happening right across the country and you're you're trying to cover that, and and therefore shipping a truck from the East Coast to the West Coast is yeah, so you've got to have dual assets and all that kind of stuff is is really complicated. In Europe, obviously we've got lots of smaller states and and federations, so it's you know it's it's less problematic to do localized production because you've probably got no B truck within 200 miles of where the event is. Um but you know, there's there's still a lot of times where you know we see these big global events, World Cups, Olympics, um just just you know, uh two that we've got this year, and then if we know virtually these things happen every year now. Um so there's just this iteration of of innovation that happens.

SPEAKER_02 17:42

From the standards point of view, and to ask you questions about best practices, there is uh a real effort to establish those. As a matter of fact, uh AMWA has, I don't know, four or five different specifications that are labeled best practice this, best practice that, you know, they exist. Uh in the BSF uh organization, we also have some of those. But like Ian says they're silos. They're, you know, the different parts, we're looking at those very well, making them work well, but they don't talk to each other right now. And with the some simple enhancements to specifications from both sides, we can have a bridge between them. That that would end up this kind of non-interoperable chaos of REMI that we have today.

SPEAKER_01 18:29

And and I think that's that's that's the exciting thing about the the work that's going on in the the MWA um and EBU work on the Joint Task Force on on network um sorry, get it wrong, Joint Task Force on Dynamic Media Facilities. It really is looking at everything from business problems through to the interoperable ways we work and then building on um some of the technologies we've seen accelerate really fast over the over the past year, like MXL.

SPEAKER_00 18:56

Um how can I mean I'm thinking about that uh not having to have a lot of infrastructure, not having to make big capital investments, that's using things like cloud-based production could be really helpful for smaller companies or more agile companies, although I guess you have your cloud bill ticking over all the time. Um but you see that as being can be could it be a big boost for small broadcasters, local broadcasters, local news organizations? Um or is it uh are you are you able to really do do that stuff with existing tech anyway?

SPEAKER_01 19:38

Depends what you you want to do. Like I say, if you if you just want to run, be able bringing three or four sources in, run them in a software-based vision mixer on a on a local PC, yeah, that's possible. But if you when you want to do things at scale, that becomes increasingly challenging. Um so you know, I think the you know there's there's one fundamental big problem with the cloud that that we that that we talk about quite a lot, which is actually there's there's a significant cost to to bringing media in, getting it out, um storing it. Um there's a little bit of a pushback, certainly on this side of the Atlantic, about actually more and more people moving to on-prem, but that doesn't necessarily mean that they are um they just want to work in COTS, but they still want the same architecture. You know, we've learned that the cloud architectures around um you know shared memory across hosts, around containerised applications, around microservices, is a really good architecture. Um, and that and that's also where a lot of the talent are coming in and coming out of universities who understand that that kind of architecture. So we can still architect our um workflows in that way. We just need to choose whether we want the cost of running them on the cloud or we run them on our own environment and then sort of push out um uh to the cloud when when we're at capacity around larger events or you know, what however you want to operate. So I think we're this hybrid model is is where we are, and we're gonna be there for a very long time. Whether the um uh the the kind of hyperscalers will, you know, some of them will will will sort of grow and grow, but I but I do think more of the the actual owning of the infrastructure may change.

SPEAKER_02 21:16

I think there's also a a misconception uh of uh what cloud really means on hybrid workflows, right? Because you know, you described that Ian, you described a cloud where you know only the big players are the ones that can give you the service, right? Uh you know, the ones that have gigantic infrastructure that can scale from one to a million users if you need to. But that's not the reality of most of REMI. In the case of REMI, you have a limited number of cameras, flow streams go from point A to point B. You can spin up uh a bare metal machine on another cloud that doesn't belong to the big ones for$100 a month and have enough bandwidth. And their metal to do everything you want. So the cost is not an issue. The issue that we have right now is people don't know how to do it. There's not enough talent, especially in the small companies, which are the ones that need it the most. They don't have people that can stitch this together, that can make this work, that are taught from universities, from whatever it is, on how to leverage the thousands of data centers ready to give you service. Where you can actually put you know some software in there and run your own REMI for nothing for pennies.

SPEAKER_01 22:45

But they've got an SDI infrastructure and they're at capacity on that. And their kind of question is, okay, well, do I do I t make a decision to go down an IP workflow or do I kind of think about going into a software workflow? And if I go into a software workflow, well, what does that, you know, where do I host that? How do I do that? They don't get the same discounts as large broadcasters do on cloud services. Um so quite often, if you're a large broadcaster, you might have a special deal with with one of the hyperscalers. Um, but often you'll farm out your actual production work to another company who doesn't have that same thing. Um now that doesn't mean to say that the there aren't still hybrid workflows that go on with that, and even things like um, you know, contribution feeds of world feeds that are coming out from a traditionally produced, I'll call them out outside broadcast, but you know, that traditional um workflow will be via via the internet somehow more, more and more these days. And that's where RIST and and SRT are really starting to sort of change the really have changed the game um uh uh in in in the past five years. Just that acceptance actually the internet's good enough.

SPEAKER_00 23:53

So what about um uh how does AI fit into all this? Because I know there's a lot of and I use that that big AI word, you know, with some caution because it can mean a bunch of different things to different people. Uh but I know that like automation is um is something that's integrated into some products and some tech, um, being able to um uh you know, whether it's operate cameras remotely using, you know, um using machine learning or be able to automate uh different control room processes. Um how maybe what is your sense of like what is happening with AI tools now and where they might go? And is it a lot of um trying to fix problems with a little AI magic that maybe need to be fixed before you add the AI? Or what do you what's your take?

SPEAKER_01 24:48

AI is great. Yeah, there's there's lots of things that we can do with it, but it's just kind of fast compute processing in my mind. So there's a lot of things that get badged with with AI that you know it's just the computers are now powerful enough to do stuff. Um so let's let's take a few examples. There are kind of practical editorial things you can you can do with AI, maybe things like shot logging or data management or a whole bunch of things that you can do to make finding content easier. Um there's a whole bunch of things that we may do. You mentioned orchestration, automation, and things like that. I I can see a world where I'll just go onto a website and say provision me a five camera OB and it it'll build that and it will you know put in the place the licenses and stuff like that. The problem with that is that we talk about um orchestration and automation as one thing, but it's different because you might need uh a network orchestration that's gonna provision your network layer, and then you might need another thing that's actually gonna bring in your media functions, and then another thing that's gonna manage your your UI. So um, you know, you've got all these different orchestration layers, so we need to be able to, as a manufacturer, we need to be able to provide the right control that that either a human or an AI, it doesn't really matter which one, can get in and and and provision it the equipment. And then we've got AI monitoring, and I think that's where it gets really interesting for remote production because we're now able to analyse uh uh a stream and really look in at the detail of the stream and where things are going wrong and really just identify things live. So up until this point, if you get a signal breakup, you'll switch to your your backup stream, and then you'll go into the post-mortem, you know, after the event, and you'll go, Yeah, we lost the A-leg, and they'll go, Well, why? We don't know, we've got no idea because nobody's logging that at the time because it's just the most important thing is to stay on air. Now we can have engines that are kind of constantly monitoring the the performance of the codecs, the performance of the the underlying you know infrastructure that's transporting that, give us instant feedback on on where those things are going, and then give us kind of more control uh and understanding of of you know um instant feedback and almost predict problems. So I think there's a lot of potential in that in that side of things.

SPEAKER_02 26:51

Yeah, you often see, you know, especially in the risk uh area, you know, somebody running a stream and having uh six, seven graphs with the uh, you know, as a function of time, trying to figure out what's going on and and monitoring. And you know, it's it's it's a lot of data to process, and that's where AI shines, right? They can in real time look at all those graphs, all that data, and tell you, hey, this link is getting saturated. Maybe we should do A and B before you know we go off air, that type of thing. Another uh simple example of AI, which I loved, I saw a one, a PTC camera that had an AI simple AI engine that would follow you around. You're on stage and you know you have this PTC, no operator follows you around.

SPEAKER_01 27:37

So I think I think that's very important when we talk about sort of democratisation of things. So, you know, it's it's really easy to talk about the big leagues and the big sports, but there's a whole bunch of other things. I've I've just been at a conference where people were talking about pool tournaments and and you know, um nine-ball pool and things like that. And and there's more and more demand for that content, and and we've got to produce that somehow. Um, and and and therefore, you know, there there will be engines that that enable that. And uh, you know, particularly sports. Sport is very rules-based. Um, most sports have rules, and you kind of can predict what happens in, you know, there's always that bit of jeopardy of something that happens outside of those rules, but you can be pretty certain you know where the what's going to happen next, and therefore that's another thing that's relatively easy to to predict in terms of automation and stuff like that.

SPEAKER_00 28:27

And I guess that being able to sort of say, you know, produce something remotely, and there's somebody on site who maybe no technical people on site, but you can say, all right, set up a camera. And, you know, we've got we're able to get a feed from your iPhone into, you know, into our system, whatever it is we're using, you know, uh, and then we'll take it all from here, potentially, you know. Uh and then there's very degrees of that, but it it could mean that you would be able to cover news or sports or things where potentially nobody on the scene is even.

SPEAKER_01 29:03

Yeah. I think we I mean I think we've seen we see examples of that all the time. Um it's it's you know, I I've been involved in this side of things since mobile phones could first do video, and everyone says user generated content is is the future. I think I think the reality is you know, people value the editorial input and the and the large cameras and stuff like that. So just putting you know 100 iPhones at a at a soccer game is not gonna make improve the coverage. You know, you you still want that kind of a you know directed um curated coverage. But equally well, you know, uh like I say, you know, that there are rules on how we set up cameras at a soccer game. There are you know particular shots that you go for, you kind of know if the ball's moving in one direction, where it where what the direction of travel is. There are things that you can do in in that space that are um uh achievable, but you're probably not going to want to do it on your your big flagship event. But analytics, I think, is is really important, being able to take sort of those broadcast feeds and then the governing bodies and what's really interesting, again, remote is more and more sports governing bodies are bringing production in-house using remote models because they want to own the content, they want to own the the the have access to it for more social media, more analytical stuff and things like that. So we're we're seeing a kind of shift perhaps from you know um uh rights holding sports governing bodies, sort of just letting the broadcasters produce tally and then charging them for for the broadcast rights on that to them actually wanted to produce and own that content.

SPEAKER_00 30:37

Um so I guess looking ahead, like we're I mean, you said that there are a few things that probably need to be ironed out. Um but I mean, do you see things being radically different in five years? Or is it is it really just kind of ironing out some of the wrinkles we have now?

SPEAKER_02 30:57

I I envision you know the the having an interoperable standard for REMI and an orchestration layer that you know everybody can agree on so that we don't have as many silos of uh vendor-specific solutions, especially when trying to you know bridge uh the internet and bring a stream of from point A to point B, uh, or bouncing it off a cloud. You know. From my point of view, I would like to see also, like you said, democratization of these services. So that not only the companies that can afford AWS use it, right? You know, there's so much uh so many possibilities of all the types of uh REMI that are not happening because of it. That I hope we get there in five years.

SPEAKER_01 31:49

I mean, uh, you know, fundamentally I think um uh the way we produce content, it's it's linear signals in, um, it's probably asynchronous stuff in the middle and then it's linear linear out. And um and so we're always gonna have a hybrid model by by design. Um I think what we'll see is a uh this shift from hardware-centric to software centric to more and more um challenges um coming along in and then it just it some of them aren't just technology changes. Um some of them are really um, as I said earlier, business challenges. You how do how do you license people want paper use licenses, but that doesn't always necessarily work for um you know, if you've got uh there's a cost to producing software and and so vendors need to sort of recover that cost and and do things, whereas there's a tension with with people just wanting to pay for stuff as they're using it. Um so how do how do we resolve that? There's there's questions about how do we um uh manage software that's running on on different environments and and kind of move between different hosts and in different environments. But all of these things are solvable and have been solved by other industries. So we just need to kind of in some ways grow up as a as a broadcast industry and stop thinking that we're we're super special. Um we used to be because you know we needed that bandwidth, but the reality is now bandwidth and compute power is more than capable of doing what we do. So that's the now's the right time to really start pushing in in that direction. Up until now, I think we've we've almost been held back because our requirements have been so um demanding that it's not been achievable. But we're seeing more and more um people just proving that wrong now.

SPEAKER_02 33:26

Yeah, we've also seen the the gap between broadcast and pro A V closing to a point where you cannot tell the difference anymore. You know, when one one type of production versus the other.

SPEAKER_01 33:40

Yeah, but which if I was going to change, I think one thing it'd probably be um not thinking about legacy broadcast workflows and and uh things that stop us being innovative. So fractional frame rates, you know, we do fractional frame rates because we've got a legacy um uh transmission system. And actually that you know, if everything's on the internet, then let's just work at sensible, you know, um uh uh integer frame rates because the maths is easier, the the whole thing is just so much simpler. We can work at sixty. What about in interlaced? You know, yeah, interlaced should have gone ages ago. Why doesn't it exist today? Yeah. Yeah. So there's a whole bunch of legacy that we're just clinging on to that that actually somebody just needs to be brave enough and go, you know what, I'm producing this for um, yeah, the web is now on my first client and and TV is a little bit secondary, and it's hard to do a framework conversion on my you know 60p content for the TV rather than the other way around.

SPEAKER_00 34:40

Right. Well, I wonder also how that and and maybe this is a question for how vendors can work together, um, because it um just it's it's in the it's in the word, distributed broadcasts that means that uh different disparate areas of the globe, really literally, can be working together on a broadcast that might not have had any connection previously. Um, you can have a team that's in uh you know, it's another continent, uh, and I can form a relationship with them uh for whatever reason. Maybe it's you know they're we're doing news or we're doing sports, or you know, it could be anything in terms of live broadcast and those collaborations. So I wonder what that that invites in terms of those collaborations among different broadcasters who may be very different ways of doing things, and then also vendors and how vendors work together that might not have been exposed to each other before. It kind of brings up that standards question, but I think there's also something bigger about like how do you find ways of learning and working with each other.

SPEAKER_01 35:45

So so I think you know, the the broadcaster's always been really good at it because they're all trying to solve the same problem. And you you go to an IBC, uh, you know, the Olympics or or or a World Cup, and you you know, everyone's in the same boat. They're all sort of talking, looking at how people have done it and sharing ideas and putting those demands out there. On the vendor side, um, you know, there's always been that that competition about our product is best. But what we've seen recently, particularly around things like MXL, is actually vendors coming together going, actually, look, we you know the great thing about SDI, as Sergio said earlier, is a standard that everyone had. And I could be pretty certain that if I took an SDI, let's ignore level A, level B stuff for now, but if we if I took an SDI out from one box and plug it into another, it was just going to work. Um, and we have we've kind of lost that a little bit in um in in some instances, we've got to have this interoperable layer. And I think what MXL has done is it's just said, right, hang on a minute, let's have a an interoperable, understood set of um specifications that we can all work to. Now they might not be what we're doing internally within our product, but it does enable me to to to bring content into um from from our software product and send it over to somebody else who's a graphics engine or a replay engine or or s or something like that, um, without having to pay the penalty of the read-en code, transcode, transport stream thing. I can just move stuff from compute to compute. Um, and I think that's gonna be a game changer. But that's just one thing in this whole stack of other things we need to worry about as as we move into Right.

SPEAKER_02 37:16

Because MXL solves uh a very important problem, but it's still an uncompressed format. So it's it's still only going to work within that compute module that you have. Yeah, it's great. Let's all share this. What happens when we want to move it across the internet or to another place?

SPEAKER_01 37:32

Yeah, yeah. How do we compress it? Why don't we use what it's the same? That's the same problem we got with SDI. SDI is effectively an uncompressed format. But you know, if I'm running it in a gallery or in one location, I've got SDI, but if I want to move it somewhere else, I've got to put it on a transport stream.

SPEAKER_02 37:45

So maybe MXL can replace SDI that uh refuses to die.

SPEAKER_01 37:50

I'm not going to be the person who says that. But yeah, I mean, you know, we've we've got to we've got to understand our architectures, rethink how we we do things as we move from dist into distributed production. You know, there are things in in the compute world like RDMA and uh you know all of those things that will enable us to do more of that. I'm not suggesting for one minute that you'll have a compute at a location with uncompressed, you know, 50 uncompressed streams coming in that you'll shift over memory exchange to to a host out the back now. You know, maybe if I look back on this in 50 years, that may be possible, but the reality is now that you're still gonna have that compression. But if we you know MXL is only uncompressed because that's what we've agreed to do in in stage one. There is no reason that you know, if we all agree we're gonna share a you know a particular codec, you could still share that. It's just data, yeah. Um so so there's no reason we've we couldn't in some at some point in the future have compressed MXL as well. It's just that's what we've chosen to do first.

SPEAKER_00 38:51

Amazing. Well, I think that may wrap it up for us now. But um, you know, we've kind of we've done our own remote broadcast here, along with so many other podcasters around the world. Um but it's uh yeah, I mean it's I think it's a a a fascinating topic because it really does mean that, you know, I could do a broadcast in partnership with somebody anywhere in the world. Um and and that uh that leads to I think more connection and you know more information getting to more people, and I think that's it's uh really something pretty amazing that if you told somebody 50 years ago, they would be uh thrilled and shocked. Um but yeah, thank you, Sergio, thank you, Ian, and we hope we can talk to you again in the future at um feed forum and um make sure you at home stay subscribed to feed uh so you get more great content like this. Thanks very much.