Voices of Video
Explore the inner workings of video technology with Voices of Video: Inside the Tech. This podcast gathers industry experts and innovators to examine every facet of video technology, from decoding and encoding processes to the latest advancements in hardware versus software processing and codecs. Alongside these technical insights, we dive into practical techniques, emerging trends, and industry-shaping facts that define the future of video.
Ideal for engineers, developers, and tech enthusiasts, each episode offers hands-on advice and the in-depth knowledge you need to excel in today’s fast-evolving video landscape. Join us to master the tools, technologies, and trends driving the future of digital video.
Voices of Video
Pixels, Profits, and Processing: The Hidden Economics of Streaming
Ever wonder what happens behind the scenes when your favorite streaming platform adds thousands of classic films overnight? The challenges are far more complex than most realize.
Joe Waltzer, CEO and founder of Arcadian, pulls back the curtain on the intricate world of media workflow engineering in this enlightening conversation. With over five years of experience solving delivery challenges for major Hollywood studios and global media companies, Joe offers a rare glimpse into the technological puzzles that must be solved to bring content from production to your screen.
"When you look at everything that goes into delivering an asset, there's a lot to it," Joe explains, detailing how seemingly standard video workflows can quickly become engineering nightmares. From misaligned audio tracks to subtitle files from multiple decades ago, the back catalog monetization challenge represents both massive opportunity and significant technical hurdles for media companies.
What truly stands out is Joe's perspective on the shifting economics of media delivery. As studios look to monetize vast libraries through ad-supported channels, the premium workflows designed for blockbuster releases become financially unsustainable. This has triggered a significant industry shift, with many companies reevaluating their cloud-first strategies in favor of specialized hardware solutions that can slash encoding costs by orders of magnitude.
The conversation also explores the balancing act between maintaining pristine quality and controlling operational costs. “I have never had a conversation with someone who said 'I don't care what this looks like, just give me low cost,” Joe shares, highlighting how technical teams must make intelligent trade-offs that preserve viewer experience while enabling sustainable business models.
Whether you're a video technology professional, a content creator, or simply curious about how your favorite shows reach your screen, this episode offers valuable insights into the rapidly evolving media technology landscape. Subscribe now and join us at IBC in Amsterdam to continue the conversation with Joe and the Arcadian team.
TL&DR
• Arcadian positions itself as a services integrator rather than pushing proprietary solutions
• Studios face significant challenges monetizing back catalogs due to technical complexities with older content
• Modern platforms often build workflows for single use cases that become economically unsustainable when applied to other content types
• Maintaining quality while reducing costs requires understanding where to make trade-offs that won't impact viewer experience
• The industry is experiencing a major shift away from cloud-only solutions as specialized hardware offers dramatic cost savings
• Video Processing Units (VPUs) can deliver 10-40x efficiency improvements over CPU-based encoding
• AI technologies show promise but haven't yet delivered turnkey solutions for media workflows
• Simplified workflows and reduced complexity often yield the greatest operational efficiency gains
Stay tuned for more in-depth insights on video technology, trends, and practical applications. Subscribe to Voices of Video: Inside the Tech for exclusive, hands-on knowledge from the experts. For more resources, visit Voices of Video.
Well, welcome to another edition of Voices of Video. Thank you, as always, for spending your precious time with us. You know, every week, sometimes a little more frequently, sometimes a little less frequently, we bring conversations from those who are building some of the coolest, the most advanced, the most innovative and just the special workflows in video streaming. And today I'm joined by Joe Waltzer, who's CEO and founder of Arcadian Joe, welcome to Voices of Video.
Joe Waltzer:Thanks, Mark. Thanks for having me. It's good to see you again. Always appreciate you letting us talk.
Mark Donnigan:Yeah, absolutely. You know, we've recently our two companies started working together and you know you're doing some some, some really cool things. I I don't know that everyone in our audience knows who Arcadian is, but they soon will, so why don't you, you know, give give that three minute elevator pitch, or-minute, or however long an elevator pitch is about Arcadia?
Joe Waltzer:Absolutely so. We are a company that helps content creators deliver content to their customers. So, whatever that entails, we have an engineering team that will build applications. We have a QA team that will validate the quality of the content. We have a media services division that will help you transcode and set up custom workflows to get that content from your post-production all the way through the pipeline, all the way to your customers. So you have a wonderful experience. They love what they see. So that's where we sit in the space, and we have lots of different customers in different parts of those little areas I talked about. And yeah, that's who we are, that's what we do.
Mark Donnigan:It's a big job. You know the media workflow end to end for a lot of us. You know, myself included, and NetEnt. You know we play in this very, very narrow little sort of niche. You know the video encoding portion, but when you look at everything that goes into delivering an asset, there's a lot to it, right.
Joe Waltzer:Oh, it's incredibly complex. Yeah, and it can be incredibly simple as well, because it's all very commoditized. At this point, I mean, the technology has sort of been laid out, but putting it all together in a way that works for customers and clients is that's where the secret sauce is. That's the secret sauce, yeah. I mean, when you start looking at all the things that go wrong with a subtitle file or an audio file or even a video file and how to fix those in an efficient way. Yeah, we built some pretty complex solutions that that solve a lot of crazy problems. I mean, we've seen we can tell stories. I could tell stories forever about all the wacky content.
Mark Donnigan:Yeah, yeah, yeah. Well, let's you know, I think some of those stories will come out even in this, uh, in our interview together today, um, but now you are working, uh, primarily, or even exclusively, with Hollywood studios. Or when you talk about content creators, who is that primarily? Who are they primarily?
Joe Waltzer:Biggest client, hollywood Studios, for sure, that's where most of our work comes from. But we're branching out. We work now with regional carriers in Southeast Asia. We've got a couple of clients out there. Southeast Asia We've got a couple of clients out there, from a channel-specific company to conglomerates in some of the countries out there. That are the big players, like TV stations, places like that. So most of our work is for studios. That's where we started and that's sold to our bread and butter. But we will deal with anyone who has content and needs it to be delivered to clients. We're branching out and trying to help everyone. The expertise that we learned with the studios applies to everybody. Yeah, anyone who has content needs to get it out.
Mark Donnigan:Yeah, absolutely yeah. So you know, in the context of solving all of these problems, maybe a good place to start is tell us how you approach building your platforms, especially when it's on the distribution side of content, so encoding it. You know you obviously have to encode it. There's a transcode function, there's different resolutions that platforms require, there's different bit rates, there's different codecs. So I'm curious are you building these as bespoke solutions, you know again, for each client to meet their very specific needs, or do you bring a product or do you have sort of a platform that then you extend based on requirements?
Joe Waltzer:One of the things I wanted to do when I started this originally was stay away from solutions that we would resell.
Joe Waltzer:I had a lot of experience in the space, and a lot of my experiences coming from the other side were that when I was dealing with a partner, they were always pitching what they had. It wasn't always what I wanted and it felt like I was never getting all the things I needed from. I always had to pick a partner that was doing 90% of what I wanted, but I had to go figure out the other 10% or 15%. It was great. There's a lot of great solutions out there. I position Arcadian as more of a services integrator, so we have a lot of contacts in the industry, a lot of tech solutions, so we work with all certainly with AWS and Azure and all the cloud providers, and there's hybrid, and I could go on forever. So our role is to help you make the best decision based on what you need, and we're not trying to sell you anything more than just the services to make that happen. So our team sits down with you. What do you need? Like you said, what transcodes are you looking for? So we have the expertise to say who is your target audience.
Joe Waltzer:Are you streaming to high-end 4K TV? Are you in the US? Are you overseas? Are you in Europe? Are you in Asia. Who are you trying to get to? And then how do we get your content to those people in the best way possible? And then we build custom solutions. It's all bespoke, it's all it's yours. You want it to work for hire, so at the end of the day you walk away and you know that's your system, that we built for you and you can do what you want with it. We can help you expand it, or we don't have anything that we're trying to license ourselves.
Mark Donnigan:Yeah, yeah, yeah, it's, it's. It's very I think it's good positioning. I like the whole idea of the services integrator because it definitely is true. It rings true to me that video workflows it's amazing. Everything's a standard, right. So that should be like simple. But there's literally a hundred variations of every single standard and and, and I don't think there's a single standard that we deal with that is implemented the same way across all the various platforms. And so no, when you bring up a ready-made product or solution, there is always this component to it of like we have everything, but you know. And then you go to the next. You know the next customer, the next platform, the next, you know product requirements and you know it's like we got that piece covered that we didn't have for the, for the, for the last customer, but now there's a new thing and you're just like, oh yeah.
Joe Waltzer:It's. It's the thing we see probably the most that a lot of people, when they come to us, there's somebody at the company who's thought this through. They're the product owners, if you will, and they understand what they want and what they're trying to deliver and they usually have commitments in some direction. So they're committed to AWS or committed to Azure. They have sort of these high-level requirements and they've gotten to this point where they realize that putting that all together is way more of a challenge than they thought, because it seems like, hey, I have all these solutions in place, I just got to turn some knobs and do a couple of things that should start working. But you don't realize that your IMF files are formatted in a way that doesn't work for the you know the transcoding partner you're using. They have to. You've seen everything and they just the devils are the details. Because, as everyone who uses tech knows, that the one little thing that goes wrong will stop the whole process. And it's all got to work seamlessly and it's all got to be robust. It's got to, you know, survive issues and faults and tallies. It's got to be well built, otherwise you're going to spend more money chasing down bugs and crashes and production issues than you realize, and I think that's sort of the expertise we bring is that we can help you work through. This is a great tool and this is a great tool and this is the best way to connect them so they work together properly.
Joe Waltzer:Because that's usually what's missing at the end of the day is everyone has slightly different requirements, slightly different needs, slightly different partnerships, and there's no one-stop shop solution. Everyone will try to sell you that, and that's why I wanted to get away from that, because you'll go to a partner like, hey, if you license our product, it solves all of your problems, and then you get down to implementation. It's like, wait a minute, this doesn't work, and now we end up having to build a custom solution or not doing what we want, or spending more money than we thought, and it just doesn't quite get into the end line, to the goal line. So we've stepped into quite a few places Okay, here's how we get you there, here's what it'll cost, very transparent about upfront pricing and all that stuff and time, and trying to be as clear as we can about the complexities of the situation, and so that's, that's the service we bring to our customers and it's it's been great.
Mark Donnigan:We've been doing this for five and a half years now. Yeah, it's a real, it's an important service that you provide and that's why you've built a nice business around it. And you know, and that's even how I think we ultimately engaged is that, as the market is coming around to a realization that OpEx matters and that some of these and I say that with a chuckle, right, because it should be like, well, obviously, but yet I think there was a period of time, especially as some of these platforms and services were being built out and didn't have the utilization they have today. You know what, just spinning up servers on the cloud was probably the right choice actually at that time, you know, because it could be done very, you know, very easily. You didn't have to bear the burden of the CapEx and everything we know about the cloud, right.
Mark Donnigan:But I know that you're seeing from your customers and you know we're certainly seeing from our customers and, again, that's partially not only what brings us together, but you know, what's brought us together is people are saying this isn't sustainable. You know, like just running CPUs, more CPUs in the cloud, buying more instances. You know, renting more instances and running software is just simply not sustainable. And so then that's bringing in the need for dedicated hardware, and as soon as you do that, you now need new system architectures, now need new system architectures. And then that's where people go hang on. Either I don't have the engineering capacity to build it, or I have the engineering capacity, but I want them. I want my in-house team working on other things, and that's where they bring in you and Arcadian.
Joe Waltzer:Yeah, that's it exactly We've had. I can't tell you how many times we've talked about the cloud and how amazing it is and then how much of a burden it can be. And don't get me wrong, I love AWS, I love Azure. They have revolutionized everything I've done.
Joe Waltzer:I mean it's an incredible service, but if you aren't looking at it, the it's quick and it's easy and it's very helpful and it gets you started really fast and any startup should be using AWS. Don't buy your own hardware if you're starting up. Just get it in the cloud and get it working that kind of thing. And then, as you start to scale, if you're not thinking about what the next steps are or the next use case, you get stuck. And this is the big conversation we have is, a lot of people have built these platforms and these workflows for a single use case. So they launched a streaming service and they used AWS or Azure and that's their partner and things have gone well and it cost them a lot of money and they got it up and running. And now they're like this is exciting, let's do another streaming service, let's make a fast service, let's do some AVOD one-off, let's talk to one specific market in Latin America and let's use this tech we have. And then they're like this is not, it doesn't work.
Mark Donnigan:This way it doesn't work. The economics don't work, I don't. My CPMs aren't high enough you know, to pay for the infrastructure.
Joe Waltzer:And it's not built for all of the nuances that we talked about for that particular market. And re-engineering it to do that, that wasn't a thought. So they're just like this is it's going to cost more money to adapt the current system than it would to just be to build a new system. And we see that a lot. We see a lot of companies just rebuilding the same system over and over again because you know they just this is how they went and that's very normal. I think that's just how business works.
Joe Waltzer:It's hard to have that foresight, but that's where we step in and can help and say wait, well, we can salvage this. We put a lot of modular parts. So when we build you a service, there's a transcoding service that is just for transcoding. It's got nothing to do with this being a 4K US-based service. It'll transcode anything from high-end Marvel movies all the way down to your back catalog, and so when you pivot and you want to do fast and you need something built for that, this is built for that. It's built to be efficient at all levels. So that's how we come to the process. We're trying to help think about what's your problem now, but also how will this be useful in the future for you, so we don't have to come back in four years and build it again, and build it again. I spent a lot of time in the industry doing that and just reinventing the wheel and we're trying to sort of solve that problem for people. Sure, sure.
Mark Donnigan:Yeah, exactly. Well, you know, I think it's fascinating. I, a number of years back, spent a considerable amount of time selling into the studios and dealing with all of the technology heads of the studios. And this, you know, this is back in, like you know, 2012, 13, 14, 15, you know, which is now more than 10 years ago my how time flies. But that was for some of the studios. That was exactly when they were dipping the toe in the water of going direct, taking their content direct. So, you know, they had been licensing, licensing out content onto even netflix, for, you know, five, six years at that point. So, um, it wasn't the streaming itself was new, but many of them were just. You know, hey, we're in the, you know we're in the licensing business. You know we license to the distributor and then it's on them to make money.
Mark Donnigan:Um, and I, I know you and I had talked um the other day, as we were preparing for our conversation, about the fact that you know what you build for you know, you know a major tentpole movie, let's just say Spider-Man, you know the infrastructure and the technology you build to deliver Spider-Man and all those kinds of movies suddenly just doesn't work. Not that it doesn't physically. You know technically work works fine but it doesn't work economically. When you say we've got this massive back catalog, you know that we can monetize with fast or with some other, you know free, or you know maybe with a with a subsidized type subscription fee. You know very low subscription with ads kind of an offer. So I'm curious if you can comment about. You know what you're hearing, seeing and even you know maybe tell us some examples of what you're building to help studios you know, take their back catalog and actually make money with it.
Joe Waltzer:That's probably becoming our, our specialty, I would say, and it's mostly because it's it's easy to make movie off, easy to make money off Spider-Man right, Everyone wants to watch Spider-Man.
Mark Donnigan:It's a major franchise.
Joe Waltzer:You could spend quite a bit of money building some infrastructure and streaming Spider-Man and Marvel and all the big movies and you'll be doing just fine. As we see, there's lots of people who are doing great with all this new content. But when you take that same infrastructure and you didn't think about Bat Catalog, it just becomes this whole other world, because the source files are very different, the quality of the inputs is very different, you get terrible audio quality, you get misaligned audio, you get subtitles that are just who knows what happened there. There's been so many cuts made to the movie over the years that they're all mixed together. So a movie will have seven cuts, seven alphas, and the audio is all just in one spot and they all sort of and it's a mess. It's a lot. Again, I chuckle.
Mark Donnigan:Sorry for interrupting but I chuckle because most, a lot of our audience are, um, just never have had to encounter this world and, um, I think, have a view that um well, if it comes out of a major studio, it's just pristine, the files are all, and when you get in and you see how the sausage is made, you say holy cow.
Mark Donnigan:And again it's not because there aren't incredibly capable people working in operations at the studios. It's just, over the years, exactly what you said. It's like there's different cuts, there's different versions, there's files that you know. Originally it was a 5.1, uh, you know, uh, you know, simple dolby digital mix. And then suddenly someone came along and said, well, let's try and turn this into atmos, and there and you begin to compound.
Joe Waltzer:People who've seen it know Making sense of it is really hard. Yeah, I mean, there were so many people listening who were like I know the pain because they've come back and they've seen it, and then for a long time the studios were sort of handing that off a little bit too, and we get a lot of customers coming to us and asking questions because they're distributors right, they're not content creators, they're sort of the middleman. They're licensing content and the studios are just sending them this pile of stuff.
Mark Donnigan:Yeah, and they're going. What do I do with all?
Joe Waltzer:these files, how do I? And they can't. They're like. Well, this is a problem, because now the problem is on us.
Joe Waltzer:You, know, Netflix has famously made subtitles for hundreds and hundreds of languages that just weren't available. So when you go to Netflix and see a subtitle for a language, it's probably made by Netflix and then it's owned by Netflix and the studios are now having conversations like wait, why aren't we making that subtitle so that we can then leverage that and license that, that? Why are we handing off all these problems continuously and causing all these headaches for our partners, when we should be upstream fixing these and helping utilize our back catalogs more? So we're having a couple of conversations now with people that were interested in how do we do this from the beginning, how do we do it right, how do we store it in a way that's accessible and easy to get to? And then, how do we distribute it internally and then externally? What does that look like? And we get really excited because I think there's a lot of potential there If some people in the studios can get together and say, hey, there's a better way to distribute movies in this new day and age.
Joe Waltzer:It's not all new stuff. The Bat Catalog is very valuable. I hear that the Game Show Network is doing gangbusters and running reruns of old TV shows from the 80s right Soap operas. There's so much content out there that people haven't been exposed to that is basically new to them, and it's sitting in a vault. It's sitting in a folder somewhere. It's probably in a bucket in S3. And it's a mess. And that's what we've been doing for five and a half years is undoing these messes and building workflows and custom solutions, like we talked about, to unravel what we have and then get it to you as efficiently as possible. So it's probably 80% of what we do is back catalog stuff.
Joe Waltzer:The new stuff is fairly straightforward and people don't usually have a lot of questions and they'll have partners and they'll go to them, and building websites is again fairly commoditized now, so we don't have any special sauce there for anyone. I don't have a licensed product. It would be a custom solution. So most of the business we get are people like wait, how do we fix these bad catalog issues? Like I said, I feel like we've seen it all. We've seen lots and lots and lots of problems and they're solvable. It's something you can work with, but you know, you just got to fight through what you're getting and where it's supposed to go and just understand the process. Yeah, exactly.
Mark Donnigan:Exactly so, yeah, so let's talk about, you know, then, the new architectures that are required, because what is probably a given and would make sense is OK, even though this is back catalog. Okay, even though this is back catalog, um, uh, you know, I think, even when we were talking the other day, you know, you, uh, you, um, maybe this is your favorite movie or you just grabbed it out of the air, but it's a great movie, uh, you, you know, um, uh, let's see what was the movie actually. Now I'm blanking out on a few good men.
Mark Donnigan:Yeah, a few good men.
Mark Donnigan:That's right. Yeah, because it's funny. When you said that, I went oh, I haven't seen that movie in years, but it is awesome. I made a mental note. You know which I? I obviously forgot, but I made a mental note. I need to watch that movie again.
Mark Donnigan:But there's a lot of those movies and and there's many that are Oscar winners and that are a part of quote unquote the back catalog. And yet the studios are, you know, maybe willing to. You know, they obviously do need to cut cost in terms of what it costs to prepare it and distribute it. And you know, just because you have to be able to do that economically, for what you can make on advertising largely, but at the same time they're not willing to completely cut the quality, you know. So there's a little bit of attention, right, you know you need to cut costs really considerably, but you still have to deliver high quality.
Mark Donnigan:So tell us what the workflows, you know. Where are you then having to say, hey, we really can't touch anything here, because you know what we're doing today is, you know, like if we do that, we're going to affect quality. And then where are you finding you can bring efficiencies? You know where you? You know the 80, 20 rule right. You know where hopefully you could save 80%, but you're really only impacting quality. You know, by by a much, much, much smaller percent.
Mark Donnigan:What did that look like for you as you're building these systems.
Joe Waltzer:Well, you nailed it. Every studio, everyone we talk to, any content creator, wants the highest quality. I have never had a conversation with someone who said I don't care what this looks like I don't care about quality, just give me low cost.
Joe Waltzer:No one has ever said that to me. I don't think I'll ever have that. Yeah, everyone wants a good experience for users, because that's what the users demand and that's what they should want. So we always start there. This is how we maintain quality. This is how we even increase quality. So you know, like doing translations and adding more content to make it more accessible um, we've, we've looked in dubbing and other things like that, and you know we're always trying to be on the cutting edge of dubbing, if that's something we can offer. And, spoiler alert, it's not quite there yet, but it's getting there. So we're always trying to help not only maintain the quality but make it a better experience for the users, create subtitles and translations for them and do things that we've done upscaling, we've done upcoloring, we've done a whole bunch of workflows. We went from 5.1 to Atmos and DTSX. So we're doing a lot that not only maintains equality but enhances equality. And you're right, you can't compromise equality. So you got to find better places to be efficient, and a lot of what we do is unwinding complexity.
Joe Waltzer:There's a lot of, I think, when most people approach these problems, especially at the studios and places where there's just lots of people involved. There's usually five or six departments that are all upstream and downstream, and each one has their own way of doing things, yeah, and then there's no one who can get them all to talk together and it creates these sort of weird bottlenecks where things are not moving in a way they should, and so we help bridge that gap by building some tech that can sort of solve this. Hey, this department has it in this format, in this system, because there's always some other system out there that we have to access to get the stuff and it needs to go into this department's new CMS system for this reason. So we're getting really good at sort of bridging those gaps and helping departments talk to each other, and that's where we get a lot of efficiency, especially at the studios, because it can be complicated. I saw a chart back in the day at the studio and I won't name the name, but there were, I believe, 17 different steps where metadata started here. By the time it got to the users, there had been 17 people or places that had touched it on the way, and it was so complex that we were building apps and technologies. So at the end we could validate that this was valid tech and valid metadata. So we were just basically parsing what came out and said is this blank? Is this not? I mean just basic stuff. But it was so complicated for them to just source metadata properly and get it out to users. That's what we had to do and that's, I think, part of the system to users at that. That's what we had to do and that's, I think, sort of part of the system.
Joe Waltzer:Like you said, the studios have been around forever and they've just been building systems on top of systems. So a lot of us is just understanding the systems and that can be complex in itself and then tying those together and we get a lot of efficiencies when you start speaking holistically instead of just departmentally. So that's helped us quite a bit. And then there's just efficiency of understanding all the options right. So do you want to be in the cloud? Do you want to be on-prem?
Joe Waltzer:What technology are you using to do this thing? Are you doing manual labor when you shouldn't? There's a lot of manual processes that we can automate. We do a lot of automation and finding new technologies that make things very simple, very efficient. So those are probably the ones. It's just understanding how to communicate with the people involved and solving the problems in a way that works for everyone. That's a skill just in and of itself, and it has been. A big corporation knows that those internal communications can be some of the hardest things. And then understanding the technology and talking to partners partners like yourself and the VPUs you offer. We love having that information because we can go to our partners and say, hey, there's this new tech out there and it saves you a bundle, it's fast, it's efficient. You'll be shocked. And there's technologies out there that shock people and they say wait that's for real.
Joe Waltzer:And you're like, yeah, I can transcode your stuff. Yeah, super fast, super easy. Yeah, you probably have one on-prem and I'll just install it and I'll build a system on top of it that won't cost you that much and it'll run the whole studio on there if you want. That's right, they're on. They're sort of shocked, but that's that's where we get. Most of our efficiencies is just having the knowledge and the ability to sort of understand the problem and find the solutions.
Mark Donnigan:You know it really is. It's. It's fascinating how so much in business in some ways I'd say so much in life is really just comes down to problem solving. Right, Absolutely. You know it's like, hey, I'm trying to, I'm trying to better connect with my spouse, you know, or my kids, or my neighbor, or or you know business, you know it's like it's problem solving. You know there's a disconnect. Can I figure out where that is? How can I bridge that?
Mark Donnigan:And when you kind of just break it down to like its most basic, you know, like in this case, for um, solving these issues for efficiency for the studios, like you say, some of it is just starting with geez, let's remove some complexity. Do we need 17 steps? Well, we need eight. You know it'd be great if we could have 10, but no, we don, we don't need 17,. You know like, wow, right there, you know.
Mark Donnigan:And on the encoding side, um, lots of room there for efficiency because over the years, as codex standards have gotten added, um, um, you know different encoders. In some cases there's three, four, five, six different encoders. I also will not name the studio, but you know them very well and I can recall just being just, I was speechless. When I was in a senior director's office and there was a very old Mac computer on a cart I don't even remember it was a Mac 2, CX or what it was, but it was old on a cart and it clearly was being used because the keyboard you could tell it was being used, but it was unplugged and I'm a Mac fan. And so we started a conversation and this person casually says yeah, I don't use it that much, I don't even remember, but once a month, twice every couple of months, I fire it up because we have to create an encode of one format that only ran on that machine.
Joe Waltzer:In this guy's office.
Mark Donnigan:Now this was more than 10 years ago and so I'm sure by now that Mac has been retired and you know. But it was just like I remember walking out slapping my forehead, going this is a. You know, this is the one of the biggest. You know, we all watch and love their content and there's something they deliver on a cart in somebody's office and that's the machine, you know, that's used to encode. So yeah, it's fascinating.
Joe Waltzer:Why don't we?
Mark Donnigan:talk about, because a lot of our audience are involved in the encoding side of the workflow and so you know they're kind of waiting for okay. You know so how the studios make decisions on encoders, and you know. So I am curious for most of your clients are they coming to you with a prescribed hey, we've chosen, you know, this software solution, or we're running this hardware, or you know? In other words, they're sort of dictating to you. You know what they've chosen, or are they sort of past that? And a lot of it is. Look, we just need to validate the quality and however you get there, it's up to you, or what does that look like?
Joe Waltzer:That's a great question. It's a mix of both, although the studios the bigger studios have more of a. This is how things work and we can't change things because you know this was decided by sort of a department or another division.
Mark Donnigan:Well, there's a there's a although the the the days of the golden eyes, I think, have or have sort of gone away, but you know, I remember very keenly the days where literally they were called golden eyes and literally they were the ones who could say pass or fail, you know so so those people still exist and they still have influence and they're actually great people.
Joe Waltzer:I've met a bunch of people in studios who are in charge of sort of approving new workflows. And hey, this SDR to HDR conversion is approved through the studio and things like that, yeah that's right.
Joe Waltzer:So we do have lots of conversations like that. There's lots of workflows that are sort of pre-approved, and then there's there's some considerations that come through, like legal considerations. Some of the studios own it's like parts of the h264 codec. So since they're a licensor, they're not going to use any of the, the google, the vp, you know eight or nine. That's not allowed to, because that's that's against, even if it's a better choice for them. Yeah, technically. So yeah, it's like everything else. You sort of have to understand what their, what their specific, specific situation is and why they are where they are and what they're trying to do, and then work around that. And mostly on the edges you can just recommend and they're always very amenable to the conversation like, oh, why would you choose this software instead of this software? We've used this forever and you could just do it quick. Well, here's why. And once you show them that the other way is more efficient and they save money, then that's oh, they're like oh, well, that's okay, makes sense.
Joe Waltzer:Yeah, so it's usually high level decisions around some fairly big financial or legal or just high level quality stuff, but I've never felt like anyone was blocking us from having a good solution for silly reasons. Yeah, I think the days of them saying this is how we do it and there's no other way are over there. Yeah, they're learning that. They've got to be flexible. They need to. Yeah, they need to be flexible and they need to.
Mark Donnigan:Yeah, yeah, I agree, we certainly see that as well, you know. Of course, we're very proud of the fact that our VPUs, the Quadra, our second generation, produces really stunning quality. And, you know, can you beat it with a highly tuned, highly optimized software encoder? Absolutely Day and night, you know. But that's a given, because I have access to all of the knobs and I can, I can throw a ton of compute at it and, you know, when you're not fixed function, you can, you can do more, right, um, but I think there's a very, very healthy balance right now, front on the premium, we're seeing it, and you know, I I assume you're seeing it based on your comments too where the premium suppliers so the vertically integrated media companies, those who own content that's highly, highly valuable, franchises that could be worth billions, where they're reaching this more, shall we say, balanced perspective between how much they're going to invest in delivering a certain amount of quality and then what's really required by the consumer, are watching.
Mark Donnigan:And I won't say any names, but it doesn't matter. You know any of your favorite streaming services and almost all of them have the top paid plan, which is where you get 4k, yep, um, as soon as you go. Sometimes you just go the next level down, not to the free tier, but even the sort of subsidized tier, and your highest profile is 1080p. How many consumers are perfectly happy, do not feel they're getting a bad experience? In fact, tell their friends how great their new 4K TV looks while they're streaming 1080p. You know happens a lot Absolutely.
Mark Donnigan:So yeah.
Joe Waltzer:It's sort of a curse of being in the industry is that, as you know, as you do this more and more, you can see the problems. Yeah, when I watch 4K TV, I can clearly see the difference between all sorts of different things and bit rates and you know I'm too into the weeds to not see these things.
Joe Waltzer:But yeah, when you have conversations with people and this is a lot, we have a lot of these conversations when we talk to clients, because they all tend to get more excited about some of the cool stuff they could do and let's do this and this and this.
Joe Waltzer:We can up-res to 4K and lots of conversations like that. And we always want to say, hey, this is the consumer utopias we're going for. Do they care? Is this going to be material impactful to them? Because this is expensive. What you're asking us to do is a big workflow.
Joe Waltzer:So we're always just trying to make sure and stopping and, hey, let's have a check-in, because it's easy to get carried away with cool new tech. It's easy, as the industry experts, to be like this is a improvement and wow, look at that new visual fidelity and look at that. You know like, hey, we're, we're saving all this money on this, this and this, and you know that we went up to 24 megabits from from 18. And look, you can tell if you look here, here and here, and so there's a lot of like that's the way.
Joe Waltzer:Let's just step back, because we're talking to the consumer, not to you. You know, yeah, you won't, that's right. You're not the only person. So we, we do have all those conversations and there's just conversations to have, and it's not because we think we're experts on what consumer taste is. That's not where we're at. But we always want to have the conversation because we want to understand everyone's intentions because it's tough. It's tough to step back and say how do the consumers see this and do we really need to spend this money on something that for the most part people 8K was a big one? I'll talk about that because people were very high in 8K for a long time a year and a half two years ago.
Mark Donnigan:I would love to hear your experience there because, yeah, boy, I can't even remember the last time I was in a session at an industry event and there was discussion of 8K, but boy, it was hot for a short period.
Joe Waltzer:Yeah, and we kept having that conversation because we'd say if we do 8K, that's basically four times more expensive than 4K. Yeah, I mean, you get some efficiencies with the new codecs. But you're talking to H.265, h.266, just to make it reasonable to stream it.
Joe Waltzer:And now you're talking less devices like H.265 on device hardware. Decoding isn't a thing in a lot of places, so you can't like it's just like. 8k is cool, yeah. And when you see an 8K TV and an 8K image and it's flying, and you're standing there in front of 120 inches of beauty, you're like this is amazing, yeah. But that's when you have the conversation. This is going to cost four to five times as much as your 4k stream which how many people are even watching that.
Joe Waltzer:This is going to cost 16, 20 times your 1080p stream. Like, yeah, like this, let's all step back and just make sure we're doing this for the right reason and we're happy to help you if that's what you want. But you know, just we do a lot of no-transcript. Then the coverage is like oh okay, yeah, you're right. That's probably not worth the effort.
Mark Donnigan:Yeah, absolutely my observation of everybody, and I'm not trying to single out. You know Disney does a great job at Disney plus. So my reference to Netflix is not to you know, say you know they're the only ones doing a great job. But you know, let's face it, I think universally Netflix is, is, is a, is a standard, um, that, um, you know, certainly, as I travel around the world, you literally hear the comments from incumbents in that region, in that country, who say we need to be at Netflix quality. You know, like that's what we're shooting for because their customers are consuming and that becomes their reference.
Mark Donnigan:You know no-transcript. You know you look at their 4k and you look at the bit rates and obviously those of us who you know are really into the weeds or into the pixels that's a better way to put it and into the pixels, you know can make cases. You know where it's like ah, I can see where they traded off. You know here and there and, uh, I wish they would have. You know, like you. But but when you step back and you look at their choices on the whole you have to say it's brilliant because it looks amazing when you look at the aggregate of the whole experience. You know, over the course of a, you know our episode of a TV show, or you know a two hour movie or whatever. It looks brilliant. The trade-offs were usually absolutely the right ones, like if you're going to trade off, like they trade off in the right places. And and that's in contrast to kind of the approach of a few years ago you know before some of the advancements in the newer codex standards or things like content adaptive encoding. You know CAE, which you know it's also funny. It's like even that's not talked about as much anymore because that's just like just built in to pretty much every encoder. You know everybody's rate control has some sort of content adaptive. You know whether someone calls it that or you know they might have some other marketing. You know uh name for it, um, but um, uh, you know it.
Mark Donnigan:I remember the days where you know there were some uh, I remember some of the first 4K. I think it was the BBC streamed the World Cup and look, this is not an indictment on the BBC, because they're a very advanced organization, but this must have been 2017, maybe somewhere around there and they wrote a blog post about it even so, I mean, even they pointed out and they were streaming at 40 megabits, you know, for 4k, and that's because they, you know, they were not going to compromise on quality bar. It was 4k, p 60 and and HEVC obviously, and if you think about it back then, like, yeah, that was kind of the state of where, of where things were, but they had a lot of difficulties with buffering, you know, rebuffering events, et cetera. You know, um, my point is that it's, it's fascinating how we've all gotten more fit, you know, in terms of our budgets.
Mark Donnigan:How much money we're spending on our operations is kind of like, okay, yes, we can easily spin that up over here, we can go, you know, use this, you know commercial service or whatever, but you know, like, is there a more efficient way, meaning, is there a less expensive way? So we've gotten more fit, you know, in in our, in our budgets. We've gotten more fit in our bit rates.
Mark Donnigan:You know, we've gotten more fit in. You know just sort of balancing. You know where we're compromising and you know where we say hey, you know that's not good enough. You know so we do need to. You know we need to do a better job there. So and I think that helps everybody, you know help the consumer yeah, and it's.
Joe Waltzer:I mean, you make a great point about the quality and so the desire to make it as the best possible, but at the same time, you're, you know you're, you're making your backend struggle. So is it even better? Quality? Like if you can't deliver the 4K to the client seamlessly, you've made a bad decision. And the picture quality, the pixels, that would be amazing, but people are going to struggle to even see those pixels Exactly.
Mark Donnigan:And who cares if it's amazing quality, but every 60, 75 seconds there's a rebuffering event, you know. Or every five minutes, like at some point, it's like even as a video file. Myself I would, especially if it's like a sporting event, or I would gladly trade off reducing quality for just a smooth experience. You know, absolutely.
Joe Waltzer:And then how much more?
Mark Donnigan:would our customers, our viewers, our subscribers do that?
Joe Waltzer:Yeah, I mean, there's just so many little things to go into that you have to look at holistically because there's risks too.
Joe Waltzer:It might not crash your production environment, but if you have a really complicated ABR tree and a really complicated delivery system and you're forcing yourself to go multi CDN, gobbling up all this bandwidth and you know that's, that's a great cool technical solution we see a lot of that too Like this is a cool new way to do this tech. We're like that's great too, but now there's so much complexity involved that it's going to be hard to fix. You want to keep things simple and you know we everyone appreciates having really high bandwidth streams and really brilliant pictures, but everyone appreciates having really high bandwidth streams and really brilliant pictures. But when it crashes and it doesn't work and you're getting written up in all the websites because your broadcast failed, because the solution was just a little more fragile than you realized, yeah, that's not great. So lots of conversations around that, not just quality, but quality of delivery, quality of your infrastructure, just quality overall and how do you make it all work?
Mark Donnigan:You know, I like to think of it as UX, you know, like user experience, you know, and user experience in this context is the whole it's, it's, it's what the?
Mark Donnigan:user. You know experiences and then what they value so well, joe, let's, let's bring the plane in for a landing, so to speak. This has been, you know, great I, I really I feel just partially nostalgic because you know I'm hearkening back as we're having conversations you're sharing, you know, your experiences. I'm hearkening back to really a very fond period in my career. You know where I got to spend a lot of time in LA and on the studio lots and you know, with these people doing amazing things and you know it really is amazing when you're in a screening room on a major studio lot and you're standing next to one or two or three of the golden eyes and they're sitting there. And you're standing next to one or two or three of the golden eyes and they're sitting there. You're standing three or four feet off this massive screen and they're showing you what they're looking at. I can remember multiple times walking back to the rental car and just being like, wow, this is the coolest job ever.
Joe Waltzer:I love it. I love all of this.
Mark Donnigan:I love the problem solving.
Joe Waltzer:I the the fun tech. I love the conversations. I love talking to you about this stuff. This is so it gets me up in the morning it's an amazing industry. Yeah, combined technology and content and people from different walks of life. It's, it's cool. Yeah, it's cool, it's unreal well, let's, let's.
Mark Donnigan:Let's wrap up with this, though, because, um, you know you, you have an interesting perspective that is slightly different than someone who's you know, maybe you know building a video platform and a large social network, or working in a streaming platform. You know what? What trends do you see coming, what? And let's not? You know I'm. You know we could discuss around maybe business models and things, but I'm not thinking that. But, like, technically, you know, is there, are you seeing people having more conversations around maybe certain codecs, or you know technology directions or the way they're building their workflows or the way they're thinking of building their workflows? Anything stand out that would be interesting.
Joe Waltzer:So the big one that stands out and this has probably been happening forever, but I see it more and more is the repatriation from the cloud. We get a lot of people who are reevaluating their what they're doing in the cloud and why they're doing it there and, you know, is it really the cost effective way to do it? So we have a lot of conversations about people like this is my infrastructure, it's all built in the cloud. Here's the bill I get every month. It's enormous. And the bosses are like, like you said, we're everyone's sort of cost conscious now and they're like hey, how do we reduce? Okay, you could do this, this and this. Move this to on-prem. There's lots of ways to fix this. If you invest a little capital here, it gets you this gains here. So that's probably one of our biggest just general conversations we see as a technology trend.
Joe Waltzer:And then the next one is the obvious one. It's AI. Everyone's asking about AI. I get probably more questions about AI than anything, but I don't have as many conversations because they don't go quite as far as everyone wants them to go. There's a lot of people sort of proposing AI solutions and we've dug into quite a few of them and we're we're always hopeful to find some sort of diamond in the rough that, hey, this really solves the problem, but none of them are quite there yet. So everyone's interested. We've got our our you know eyes and ears out there looking for that. We've talked to a lot of partners that's what we do a lot of conventions, to see what's out there and get a feel for where AI is taking us, because it's going to go somewhere. I mean, I think everyone agrees that the technology is in its infancy and it's going to head somewhere interesting. Now, five years from now, 10 years from now, and we've seen VO get launched from Google and amazing things are coming.
Joe Waltzer:So we have a lot of people asking if AI is the solution, and the answer right now is I don't think it's what you think it is. Yeah, I think it still is going to take a little more effort than you hope and it's not really the turnkey solution you're looking for. But keep an eye out because maybe one day it will be, or maybe one day it'll be close enough that it's going to solve your problem.
Joe Waltzer:And then the other exciting thing about AI I'd like to say is that it starts to unlock the creators in all of us. So these AI tools are allowing people who aren't movie studios to make compelling content. There's some amazing stuff out there. I was watching a creator that made ads that got on I think it was the Super Bowl or something that was just all AI generated and it looks amazing. And this is just one person in their house with maybe another person on their team, and they're making really high-quality content. So I'm excited to think more people are making more cool content. I mean more content out there for specific audiences. You don't have to make billion-dollar Marvel movies that are trying to cater to everyone. You can make some really good content that's just for a niche audience, but since it doesn't cost $200 million to make, people can enjoy it and you can make money by making specialized content.
Joe Waltzer:So AI is probably the big one that there's a lot of talk, as you know, and we're in all of that. It's exciting times.
Mark Donnigan:It is.
Joe Waltzer:We're going to talk about this in 10, 20 years, about how we were there for the AI revolution, I think.
Mark Donnigan:Yeah, yeah, totally yeah, I'm excited about all those uh same things you know. So, yeah, the uh cloud, um, repatriation, that you're you know you're talking about, you know kind of um, that that really is a big one and you know it's actually fascinating. Um, I somehow, um, you know there's a, there's a guy who's a bit of an influencer in the sas. Um, you know software as a service business and you know his company. He's built some, some pretty successful apps and and you know he's a real, you know mover, thinker, shaker. You know he's an entrepreneur, ceo, technical guy, not in the video space at all, but he started talking a couple of years ago and doing a lot of writing on X and LinkedIn and blog posts and whatever about their whole journey from moving their workflow off the cloud to a hundred percent on prem and and and look, you know he's running general purpose compute because you know he's a SAS company, right. But as I was reading the documentation of his journey and and beyond just the incredible cost savings and the efficiencies it it, it just became clear that you know there are tools that I think some days tools run almost run their course, and and again, not for everybody and it also doesn't mean that even if you're going to maybe move one of your significant workflows on-prem or you're going to do some sort of hybrid, that doesn't mean you're still going to have a whole bunch of other stuff in your business that should be running on public cloud. There's zero reason to go put it in a data center, you know. So it's not like it's exactly a zero or a one, but this whole trend, I think it's a bit of a challenge because and I've never worked inside a cloud, so you know, I can only peer in from the outside and it's always easy to give advice, you know, to. You know people from the outside because I feel I'm so smart, you know. And you know people from the outside because I feel I'm so smart, you know, and how come these people aren't getting with it? Um, but I can sort of understand where you know they've enjoyed building a really profitable business and um. And it's hard to cut your rates, cut your costs. You know, um, it's hard to do that, but I think they're gonna have to. Um.
Mark Donnigan:Just as a case in point, um, you know one of our partners, um, just recently came to market and in fact when they announced their, their price card. I mean, we actually like followed up twice with them. Be like is, is this, is this correct? Like or like, before we start amplifying this on LinkedIn and everything like, are you sure that number is correct? They're like yeah, yeah, yeah, go to the website. You know, and this is a public company, so you know it's like we go. It's correct 42 cents an hour for a VM. That is a very capable cpu um with a vpu. So 42 cents an hour and and it's now they have a.
Mark Donnigan:This particular company has a one vpu and a two vpu solution, um, one vpu know for those who aren't familiar capacity can do 32, 1080p30, a live encodes, av1, hev CRH264. And look, I'm not saying this to you know. That's why I'm intentionally not naming the company, because I'm not trying to like, but I'm saying like 42 cents an hour, an hour. Now go to your favorite public cloud and kind of look at pricing and then think about what it would take to encode 32, 1080p, 30 streams, like it's going to be a lot more, and I don't mean like 20% more, I mean multiples more. You know, and that's exciting, though for me, that's a challenge for the incumbents, but that's exciting for though for me. That's a challenge for the incumbents, but that's exciting for me because now these new business models like fast can get unlocked, where we can deliver high quality but do it incredibly economically.
Joe Waltzer:You're right, it's out there. People don't know and hopefully, maybe, listening to this, they start to learn that there are those incredible solutions out there, those things that you just like I didn't even know that was possible, didn't even know, yeah, and we've actually.
Joe Waltzer:this is a fun story, I'll just end. My little thing with this is that we, when we first started, that was our goal to be efficient enough that we could help you with that catalog stuff, because we had to be inexpensive. And we have a team in the Philippines and we've done a lot of stuff to just get operational efficiency out.
Joe Waltzer:And we were quoting studios. They were saying, hey, what would be the price to do XYZ to convert this file to this type? And we'd go through and we'd say, okay, it's this number. And they'd come back and say that's so low. We don't believe you. They'd be like that costs so little that we're going to disregard your bid because we think you're lying and we think you're that kind of company who was just under bidding to get the business.
Mark Donnigan:Yeah.
Joe Waltzer:And so we actually had conversations internally like, geez, maybe we should just triple our, our, our costs.
Mark Donnigan:But we're not going to do that, we just had to educate people like Like no, no, it's because we've just we've been doing this enough that we know, yeah, we've optimized how we deliver our service and our you know our staffing and everything's optimized, even in our yeah, funny, joe. Obviously, h.264 is much less compute intensive than HEVC and HEVC is less compute intensive than AV1. So if you're going to benchmark H.264, you know it's roughly about a 7 to 10x improvement VPU to CPU, 7 and 10x. And you know, as always, there can be a case where it's a little bit less, there can be cases where it's higher, but you know, and so that's where we get our 10x, you know, improvement from yeah, um, when you go to hevc, that becomes about 20x roughly. You know again, it can be, you know, 12, 14, 15 to 2022, but you know, somewhere in there, when you go to AV one, it's like 40 X. Well, so you know.
Mark Donnigan:So then the discussion internally was okay, we are accurate to claim a 40 X improvement. Because if you look at AV one, like, like that's and, uh, you know, um, there were certain people who were like, yeah, you know, 40 X, 40 X. And I said guys, like, no one will believe us, number one, you know. And then they're going to, oh, another encoding vendor, you know, make a. I don't even need to talk to those guys. They're. They must be crazy.
Mark Donnigan:And and I said you know what Even a 10X will make people's jaws drop and go. That sounds too good to be true, but let's go check it out. So that's why most of our marketing material you'll see both 10X and 20X, sometimes based on kind of when the message was written and where our state of mind was at the time, and other times there's very specific reasons. But a lot of times you'll see 10 X. You know because, because if I can bring anybody a 10 X improvement, they're going to be well, I need to validate it. But if this is true, sign me up, you know. And then when they find that might actually be getting a 20 or 30 or 40 X, they're just like they want to hug and kiss you. It's like this is amazing.
Joe Waltzer:It's trust I don't know if you've been on the other side, but I have and it's people will tell you what you want to hear. So they get your business and then they get you trapped in contracts. It was the bane of. I think everyone who's done this has been stuck with this contractor who said they would do a thing. Yeah, and they made a promise and it was hard to validate and you trusted them and then they didn't deliver. They didn't deliver. I think people, I've been burned so many times and I get it. I get why you're like that can't be true. Yeah, but it is people, because you're right with the VPUs, we've been looking into them. Our clients are all excited. We're excited to use them. They're a revolution. They really are. They're going to unlock so many things I don't think people see now, because when you get to those efficiencies, all of a sudden HEVC isn't as much of a problem as you thought it was to work with.
Mark Donnigan:That's right. It's faster, that's right.
Joe Waltzer:You don't to spend $4 or $5 in the cloud to transcode that. You can put that in a non-prem PC that doesn't cost much and you can start decoding HEVC at pennies. It went from dollars to pennies, correct? And it adds up when you do 1,000, 2,000, 5,000 transcodes, all at $5 each. That number is so big that people are like that's not worth it. Yeah, and now I'm like, no, that's going to cost you five cents each. I'm like, oh well, that's nothing.
Mark Donnigan:It is not Really that's impossible. You're like, no, no, no, it's totally, trust me.
Joe Waltzer:And then we have to build POCs. We're doing that right now where we have. Let me show you We'll do something for you and we'll kind of prove it to you. Yeah, there's many exciting things happening, and VPUs is one of the ones that we're excited to help out with, because it really is. It's going to unlock a lot of things for a lot of our partners.
Mark Donnigan:That's great. Well, Joe, this has been a great conversation. Thank you for joining us on Voices of video. Really appreciate it.
Joe Waltzer:It was great. It was good to talk to you.
Mark Donnigan:As always, Mark, Appreciate the time yeah absolutely Well, good, well, I know that we're going to be together at IBC. Uh, in fact you're going to be in our booth. So, um, you know, for anybody, uh, this episode will be coming out, um you know, probably um six weeks or so before IBC starts. But make sure you come see Joe and come see Arcadian and come hang out in the VPU ecosystem booth, which is the NetEnt booth.
Joe Waltzer:It's going to be a great time we love to talk, let's just do this. We'll just chat about what everyone's talking about. I love this conversation.
Mark Donnigan:This is great. We'll have a little cocktail party going and it'll be, good Amsterdam's fun. So yeah, share some more stories and have a great time.
Joe Waltzer:I can't wait to see you there. It'll be fun.
Mark Donnigan:Yeah, very cool, okay, awesome. Well, and to all of our listeners, thank you again for listening. We really appreciate all of you and until next time, happy encoding.
