How to Scale SnapStream (Video)
SnapStream User Conference (NY 2018)
Watch SnapStream customers The Daily Show with Trevor Noah, CBS, SportsNet NY and ABC discuss how their organizations integrate TV into their workflow at scale.
David Hadley from CBS discusses how they are building a large SnapStream cluster so various shows, including The Late Show with Stephen Colbert and The Rundown with Robin Thede, can access the same TV content, but still keep their show’s accounts separate. Adam Cooper explains how their ABC producers pick and choose clips from their SnapStream recordings to import into their Avid system. Gil Santana with SportsNet NY describes how SnapStream allows every staffer at the organization easy access to their 24/7 SNY recordings to quality control all the content they create. Dennis McMahon from the Daily Show with Trevor Noah discusses how they use SnapStream to record hours of TV so writers can insert TV clips in the show for various comedic segments.
Also watch our Using SnapStream to Influence Culture video.
Panelists
Dennis McMahon | Director of IT
“Our goal is to entertain people, while satirizing the daily events of the country, with a lean towards politics.” His career didn’t actually start in IT. It started in TV production on “Who Wants to Be a Millionaire” where he rose to the title of Producer.
David Hadley | Technical Director
David oversees all technical operations of the internal media department at CBS Corporate Headquarters in NY. Includes CBS VTC network infrastructure (globally), CBS Content Portal (webcasting), SnapStream server administration, media SAN, transcoding servers, edit suites, broadcast studio and control room, internal cable head end, corporate a/v operations, executive media support.
Guildry Santana | Director of Information Technology
Gil was the key management lead in SNY’s 2 year Infrastructure relocation from Time Life Building to 4 World Trade Center in 2017. He oversees the day-to-day operations in accordance with best practices for high availability, high performance and usability. Regularly available to the team to provide managerial direction and support during off-hours related issues.
Adam Cooper | Technology & Workflow Strategy
Adam Cooper joined Disney|ABC Television in 2002 as a Production Assistant for ABC News Graphics. In 2008, he became Manager, Technology, for ABC News focusing on graphics and off-air recording technologies. Adam has spearheaded technology projects for special broadcasts including multiple elections, inaugurations, space shuttle launches, the Royal Wedding, and others.
SnapStream Admin Panel Transcript
0:00:00:>>MATT QUERING: All right, guys. Well, thank you very much for coming this afternoon. And this is our last session of the day. It’s a good one. It’s one I’m really looking forward to. We have a lot of, you know, great panelists up here — big time SnapStream users. And this is going to be really exciting. So for those of you who don’t know, I’m — my name is Matt Quering. I’m our East Coast account executive working — I live here in New York, and I handle most of our, you know, East Coast business. I know a lot of you, but for those who don’t, you know, that’s who I am. Our panel — you know, as I mentioned, you know, we have a lot of longtime, long, long time SnapStream customers. And so they’re going to provide a lot of great insight into SnapStream and technology in general. That’s what, you know, I’m really looking forward to. The title of our panel is very stable, geniuses present how to scale SnapStream. So, you know…
0:00:50:>>UNIDENTIFIED MAN: (Unintelligible).
0:00:50:>>MATT QUERING: …(laughter) these, you know, these gentlemen have, you know, started off, you know, with small SnapStream instances. They’ve grown them. And we’re going to get into that. I’ll start by introducing all the panelists, and then we’ll, you know, get into a chat, you know, later on. We’ll begin with David Hadley here at CBS. David is the technical director at CBS BlackRock, the internal media department at CBS corporate headquarters. He oversees all technical operations, which includes video, editing, corporate meetings, maintaining the RF head and distribution, running a studio, managing videoconferencing for all of CBS globally as well as SnapStream administration. Interesting note about David’s career was that he started his career in showbiz at Studio 54, the infamous and legendary Studio 54. So I’ll let David talk a little bit about his SnapStream set up.
0:01:40:>>DAVID HADLEY: Hi, everyone. So we’ve got a bunch of different instances of SnapStream across multiple buildings that are being used individually by different groups of users right now. We’re in the midst of building a one large cluster that’s going to have all the users in separate rooms sharing certain content and keeping their individual content separate. We have — I think there’s a slide up there. Yeah, if can see. So I don’t run all these SnapStreams by “The Late Show,” The Run Down” and our platform cluster and “Samantha Bee,” I administer. So right now we have all these different users who are kind of competitors in a way, right? Smith — he certainly doesn’t want to share their content, their secret content with anyone else. You know, records. Even then there are certain things you don’t want another team to know what you’re recording and why you’re recording it. So most of these places are recording the same things over and over and over again. Now we’re trying to get to a point where all that duplicate media can be kept in some sort of a community silo but everybody can have their separate assets with themselves and keep them secure from the rest of the folks. And while that seems like it might be a really easy thing to do, it’s….
0:03:15:>>MATT QUERING: Absolutely.
0:03:16:>>DAVID HADLEY: We’re going to figure out how to do it.
0:03:18:>>MATT QUERING: Thank you, David. Next up is Adam Cooper from ABC. For those of you who did not know Adam prior to today, I’m sure you do now because he’s been shouting out from the audience all afternoon, but, no, Adam — he joined Disney ABC Television in 2002 as a production assistant for ABC News graphics. In 2008, he became manager technology for ABC News focusing on graphics and unfair reporting technologies. He spearheaded technology projects for special broadcasts including multiple elections, inaugurations, space shuttle launches, the royal wedding and others. A fun fact is that Adam is a true Disney fanatic. He’s actually been there more than 50 times, and he’s never come back sick. So, I guess…
0:03:59:(LAUGHTER).
0:03:59:>>MATT QUERING: Yeah. And Adam, if you can just walk us through your set.
0:04:04:>>ADAM COOPER: Sure. Thank you, all. I will try to be brief as difficult as that is. So our system is a little bit different than some of you’ve talked about. It’s probably similar to David’s here is. Our original spec was much more of an unfair reporting system. We are in that house. That is not any broadcast secret here. And the original design of this system was to try and take some of those daily types of records that the shows require for the broadcasts for the editing purposes or rolling pieces and take that out of our Avot system, be able to instead of recording, you know, thousands and thousands of hours of ISIS be able to report on the side and let the producers pick, you know, the two minutes that they need from that moment in time and send that into our Avot. So we have a pretty deep integration with our ibid system. The original design was very much that producers can come in clip and send it to avid and then all the craft editing happens in those environments. So it does end up in any of the incoming folders. And over the years, our SnapStream system has expanded and expanded and expanded as we have figured out that it does some pretty good Swiss Army things. And as the technology has evolved, we have grown the system. So that’s kind of the sharper. The interesting part of us is that we use not just the standard fees going in. We have — we’re spit about half of each GFCI these days. So on the dashboard where aired NDA and 2110. And what was missing was our spin for us that’s actually a very interesting conversation. Because as every other broadcaster in this room, everyone is frantically running to IP. And the more and the easier you can record IP, the better you are.
0:05:47:>>MATT QUERING: Thanks, Adam.
0:05:48:>>ADAM COOPER: And yeah, it’s actually 60 times, but….
0:05:53:>>MATT QUERING: (Laughter) Next up, is Gil Santana, director of information technology at SportsNet, New York. SportsNet New York is a regional sports network in the linear TV network for the New York Mets. Do we have any Mets fans? All right. All right.
0:06:05:(LAUGHTER)
0:06:06:>>UNIDENTIFIED MAN: Big fan. Big fan.
0:06:13:>>MATT QUERING: Gil’s been with SNY since 2005, and he was the key management lead on their move from the Time Life building down the World Trade Center. If none of you have been there, beautiful office space. Maybe Gil will would take you for a tour. He’s passionate about sports, early adopter of all things tech, an avid research for the equity markets and loves New York City real estate. He currently lives on the Upper West Side of Manhattan with his wife, 8-year-old son and 2-year-old old daughter. And Gil’s going to take us through his setup.
0:06:41:>>GIL SANTANA: So thank you, again, SnapStream for having me on this panel. At SNYR, integration with SnapStream is a lot smaller. We’re a regional sports network. So our use case for technology really is, has been to basically have every single sort of like staffer at SNY to be able to see the content that we are creating. So we’re primarily recording SNY 24 by 7 on our SnapStream server. And basically you have honored talent, our executives. You have even the engineers who do routine air checks and even our sales team that are either clipping the content that’s reported and sending it out to agencies. We’ve also — as you can see, we’ve also integrated the set top boxes. And those are going to be put in conference rooms and even in executive offices. So, again, because they run our senior management team every, almost every day weekly get together and sort of, you know, sit together and hash out how they can enhance the product that we are distributing out to our customers.
0:07:53:>>MATT QUERING: Thanks, Gil. And now last but certainly not least, Dennis McMahon Director of Information Technology at “The Daily Show.” Dennis manages all I.T. infrastructure at “The Daily Show.” And he’s been at the show since 2010 going back to the days with Jon Stewart. He’s spearheaded SnapStream’s growth of the show, which now oversees hundreds of terabytes of archival television footage. An interesting note about Dennis’s career he actually didn’t start in I.T. He starred in TV production on “Who Wants To Be A Millionaire” where he was a producer at the show. And Dennis…
0:08:25:>>DENNIS MCMAHON: Thanks, man. So we have a similar setup as the rest of these guys. You know, obviously it’s “The Daily Show” so we’re recording stuff constantly. Anyone that’s familiar with the show knows how important the video that we roll into support our stories jokes is. You know, it’s an integral part of the show and what started in 2010 has morphed into this beast of a thing I manage now in 2018 because the producers at my show want to keep everything. Which if anyone works in technology knows the limitations of physical space, it’s quite difficult. So it’s been quite an experience learning how to deal with that. You know, we started in MPEG 2. And we still have MPEG 2 going. And we’re transitioning, you know, over the past couple of years to H264. But I have this backlog of video that just sits and it’s been a challenge. But, you know, the guys at SnapStream have been great. And the product is very impressive with the way it can go through eight years of 24/7 recordings instantaneously for the producers. And it’s really changed the workflow at the show and allowed us to do some really great things with what we have.
0:09:54:>>MATT QUERING: Thanks, Dennis. So let’s get into the discussion. And where I’d like to start is, you know, the subtitle of this panel is how to scale SnapStream. But I’d like to talk about technology in general for a second. And when you gentlemen are reviewing technologies that you’re anticipating you’re going to scale throughout your organization, what are some of those qualities that you’re looking for in those technologies that you say, all right, you know, it might be used by this group today but next year it might be used by, you know, five groups? Who would like to start with that one?
0:10:27:>>ADAM COOPER: Don’t everybody jump. I’ll take it just because I talk. I mean, we have a very and I think probably that in the larger positions is a fairly formal process to evaluate new technology. I mean, clearly everyone’s always looking and you bring it to the table. And in the larger and larger organizations you go through these processes — every three to five years they show looking at quartet analogy and saying OK here’s is our man, here is our Pam, here’s our, you know, whatever media or any of those places. Generally, there’s a formalized process that says, you know, we want to come look at this. There’s a project to replace this demand growth supplement. We have our sourcing division, technology team, I.T. teams — what are the teams financial guys? And there’s a whole group of us that come together and we look at what are we trying to do to, spec out the requirements generally put out an RFI of some sort, doing evaluation, generally a proof of concept and ultimately figure out from a financial point can we afford it. Things we look at — the production requirements. You know, does it fulfill the need? Is it growth future looking? 24/7 redundancy, 365 support from the company. All other things you think of in an organization. Like, all the organizations sitting up here. I mean, we’re all going all the time. And downtime is very hard when you have a system like ours. You know, upgrades are challenging. To bring the whole system down for an hour is a small act of having to find out one hour that you can bring it down and do things.
0:12:01:>>DENNIS MCMAHON: I think an important thing to think about is in today’s world when technology jumps out at us so often is making sure that your users are the people you support think it through. You know, it’s a common theme that we want today simply because we’ve seen. We all act like children at older age because that’s who it’s being marketed to. So often if new technology crosses my desk and it’s coming from a source whoever it may be that they’ve seen it in their friends’ feed or something like that, I have to pose the question like, OK. It’s great, but what are you going to use it for? You know, because that’s a valid answer to me because you just want it in the building and then it’s my job. You know, so you have to prove to me what it’s used for. And most of the time if it’s not anything that’s — if it’s anything that’s going to stay around, that reason will be there. If it’s anything that’s going to fall by the wayside, that reason generally is not there. And it’s a good way for me to just parse through the requests knowing that, hey, I want to support you, but, you know, this is my job, too. So it’s important that you know we’re on the same page with that type of stuff. And second, what Adam said support is probably number one for me. I’ve dealt with companies that have a less personalized support aspect. And you feel the difference. You know, you feel the difference. Your company happens to be very good at support. And I’ve dealt with some other ones that are and some that are not. And that usually shows up after you’ve made the purchase unfortunately. But it’s a good thing. It’s a good thing for most companies to focus on. So I would say that’s number one when I’m looking into that.
0:13:49:>>MATT QUERING: What you said before about there being a lot of technologies. You know, how has things changed over the last five, 10, 15 years? Is it you think it’s a good thing that there are more technologies or more almost the same time, you know, now it’s creating additional layers of, you know, fluff essentially?
0:14:07:>>DENNIS MCMAHON: I was just having a discussion with Jeff. We — you’re familiar back then and how some users at our company are cutting the legs off of some of the newer technology and applying it in an older way simply because they’re not familiar with the original architecture that is doing exactly what they’re using this to do. And I think that that shows up to a greater extent only because the instinct isn’t to learn what’s going on necessarily, which I understand not everyone has a technical mind or has the desire to read into this stuff but just simply apply what they see. And we have a lot of duplication in technology out there simply because they’re trying to be first in market or the glamorous new way to do things. And I think it does muddy the waters. But in general, it’s everything’s going in the right direction because the people behind it I think are are sending in the right way.
0:15:07:>>MATT QUERING: Thank you.
0:15:11:>>GIL SANTANA: I want to speak to, you know, some of the things that Dennis meant on, you know, spoke about, like, the user experience. I mean, one of the things I try to look for every year and actually enjoy is going to each individual business unit within SNY to sort gage what’s their pain point, what kind of legacy workflow are they currently doing that, you know, we might have a solution in place right now that they just didn’t even know exists.
0:15:39:>>MATT QUERING: Right. Right.
0:15:39:>>GIL SANTANA: Because, again, I mean, you just go through every year and you implement solutions for a particular team. And, you know, although we’re a small organization, sometimes teams can operate in and silo. They’re kind of worried about what their daily tasks are. So, you know, part of one of the things, again, I enjoy is go to those teams I sit down with them and sort of like analyze what are areas where, you know, you have an individual that’s doing a particular workflow that the turnover in that employee to get someone else to kind of come in is a big ramp up on them. And that’s sort of like the areas I like to, like, target when I want to introduce the technology. Because if the business unit has a large pain point and we know at some point it’s affecting the business, I can then bring in a piece of a platform or solution that senior management will be behind. And it’s all about — I try to focus on the user and that’s part of, like, my criteria.
0:16:39:>>MATT QUERING: Got it. Thank you.
0:16:42:>>DAVID HADLEY: Well, different than Adam, even though we’re a big shop, our department is a very small department. So we don’t vet out things the same way you do at all in our department. We have a use case that comes up or a user that needs to solve a problem, and we try and come up with a solution. A couple of those things over time have turned into these enormous machines, right? So like SnapStream for instance as one of the things. We were doing off air records for executives in the sales department on VHS back in the day. This is like 2000 — going back to the ’80s really, right? But finally SnapStream shows up. And my boss founded in some magazine or something. Look there’s this little thing. Nobody ever heard of it before. And we’re, like, well this is going to solve a lot of problems. So we’ve got this little instance of SnapStream. I think we started with a 10 tuner SD system that with very little amount of storage. Over time, more and more people started using this box. And finally some other users at CBS wanted to use SnapStream. So we were already the people that have been using it for so long. It’s like, oh, let us take care of that for you because we know what we’re doing already. So instead of you guys reinventing the wheel and having to have some duplication in administration or whatever, we’ll just take. In over in our little department, we’ve always been kind of boutique because we, you know, we started out with executives and regular non broadcast people users. So we’re — we can turn very quickly. Now some of the things have grown exponentially, but we can do that with a lot of different technologies. And sometimes it works and sometimes it doesn’t. But we haven’t spent a billion dollars going into it. So different than you guys, we don’t put — not that we don’t put any thought into a purchase or whatever but we can make a small, build a small test rig up whatever it is and see if it works and see how people respond to it. We don’t typically go looking for technology and say, hey, you know, does somebody want this, we wait until there’s a user that has a problem that needs to be solved. And then we try and help others them solve or push them away from — you don’t want to do that. No, that’s not going to work or there’s no way to deal with it or management. So we’ve kind of come at most of our technology kind of in a completely different direction.
0:19:35:>>ADAM COOPER: So completely great. And I’ll tell the little inside baseball story. Our first usage of SnapStream was in my office on a personal machine where I used to be on TV. The project was at the time — we’re going back a lot of years to 2010, 2009 roughly. I was asked to do an analysis of pure proper HD content across the different networks. So we were looking at as HD was coming in, how much of each of the broadcasts were really doing pure HD content versus wings versus black panels. And so I had to look at all of the networks and say jeez they’re doing — of their two hour broadcast they’re doing 20 minutes of it of their the two hour broadcast they’re doing 13 minutes of or it or. It was a long time ago. It was three months of my life I’ll never get back. And to your point of how some of these things start, it was literally on an old HP — God, what was it at the time? An XW8400, I think. And with a silicon dust encoder and I figured out how to get my piece streams in there, and off we went. The interesting question, though, where we ended up was we started with those and then it grew organically as you’re describing. And now all of a sudden you’ve got a system that the entire department division works. And when we go to move forward off of that, that’s where you get into larger processes. Things do tend to start kind of organically. You know, hey this is a cool piece of tech. Wow. Look at this. And then a couple of people did this. And all of a sudden people are like, well, how did you get the 4 HD records of the network. We can’t even do that here today. And it’s like well I figured out if you do this widget, plus this widget and out, off you go. And that has grown into, you know, our system is really I would call pretty big. From a storage perspective, we’re definitely smaller than you guys. From a transcode record raw input power, I think we’re pretty darn big.
0:21:27:>>MATT QUERING: No, that’s great. And David, I’d like to, you know, go back to you because you were talking about, you know, those small different snapchat instances. And now you’re moving more toward a centralized system. And can you just kind of walk us through when you conceived of this notion of kind of moving everyone from these small siloed systems to a centralized system? What was your thought process like there? And how were you trying to, you know, balance effectively redundancy versus efficiency?
0:21:56:>>DAVID HADLEY: Well, we’ve had a couple of small disasters over the years. The building losing power for four days kind of hurt and made us realize that redundancy is a very important thing. Luckily, all our SnapStream are not in one place. So we were able to then move content over it and recover. But, you know, we have actual clients, not just our own users that are relying on this. So there was kind of a scramble there. And this is not the broadcast center. We don’t have redundant power. You know, we have UPSs, but they’re good for — I mean, I didn’t even have enough time to power down all the units. You know, there was a fire out in the street. And they said we’re going to cut power. And I had to run to another building to try and get on the way so that I could power down service. And I got half the power down before nothing was responding. Anyway, that’s a little bit off subject. But that showed us how important redundancy is for that part. So anyway, we have all these different silos that just organically grew. We started with our own then another user came in. And the fastest way to get them up and running was to let them borrow ours while we spun up another full instance. So that at that point we’re like OK. Once we get this working, we’ll just hand it off to them. And we’ll make sure stuff keeps working and let them be, you know, their own captains of the box. We did that a couple of different times and then we got to a point — and through all this process, almost all of our clients don’t want to throw anything out, right? You know, nobody wants to get rid of anything. So we have all this redundancy going on. And our storage is increasing and increasing and increasing. And now another show wants to come in? And there needs to be another silo bill. And the deciders at that point are like really how many times are we going to do this, are we just going to keep doing this and doing it. How many records are we making of the same thing? And what I had already wanted to do for a number of years was can’t we consolidate all this. But once you’ve got somebody up and running, it’s not that easy to say hey we’re going to change your whole workflows, right? So when we got to that point work here was the tipping point, here was the new instance that had to happen and the money people were like why would we buy more of this, you know, we were able to say look now’s the time we can start with this as our cornerstone for a whole new way of doing things, right? We can move all the other users in to one conglomerated cluster with the right user permissions and get rid of all these duplicate records, get rid of all the redundancy. So that’s what we’ve been moving towards. And it’s — there’s so many little details. You know, like on the surface it seems pretty easy, right? Oh, well stand this up. Everybody gets access to this. And everybody has their own little buckets of stuff that they can deal with on top of it. Well, there’s a lot of things that SnapStream has to do to make that happen in a way where users still feel like they’re captains of their own ship at the same time. So we’re working those out little by little. And we’ve got a pretty good plan in place, but there’s still still some work that SnapStream team has to do to help us with that, that they’re willing to do. They’ve been very — you guys have been great with your willingness to make all this work.
0:25:40:>>MATT QUERING: Thank our our development team, yeah.
0:25:43:>>DAVID HADLEY: Yes, so we’re about a quarter of the way. We’re almost done with phase 1 and then — which is just standing up the first cluster in a new environment. And then we’re going to start dropping a master node on top and bring another cluster in and having two completely different groups of users whose streams cannot cross. Like, that’s a big no-no. They can’t see what each other is doing at all. Even by mistake, you don’t want that to happen. So, you know, that’s when it’s going to be scary. When we get to the next step…
0:26:22:>>MATT QUERING: Sure.
0:26:23:>>DAVID HADLEY: …And now you’ve got a yes. This is going to work. And nobody’s going to see anybody else’s stuff. But you’re all going to be able to see that stuff that everybody gets to see, which is just the raw records of — in these cases, all the news channels, right. There’s no reason that everybody can’t see those together. And then we can start getting rid of some of this redundancy of the duplication that we have over and over and over again. I can’t get people right now to Shakira, I know you’re working on this up to to delete the overnight anderson Cooper from, you know, 2 in the morning that’s the same thing he did it at 9:00 p.m.. Well, what if something’s different on it? Like, they just — you know.
0:27:08:(LAUGHTER)
0:27:09:>>DAVID HADLEY: So — things you know are just exact duplicates. And nobody wants to get rid of anything. So it will be really nice to say look this is everything. This is what you all get. And then moving forward once we’ve built this new workflow for users, when the next group comes in, that — which is what we’re trying to do with this — the group were used that it’s going to use this first stack is this is just be normal for them instead of, oh, but we always did it this way and now you’ve got to do it this way. No, this is just how it works. You know, so it’s as much about changing people’s workflows and having them be happy with. Nobody wants to change, right? This works. Why would I want to change it? As it is about building a box that will do things in a better way.
0:27:55:>>MATT QUERING: Absolutely. Yeah. Adam, you have a centralized system, you know, right now. And, you know, how have — you know, how do you balance between redundancy and efficiency? And then can you also talk about kind of just building out those workflows for all those different user?
0:28:12:>>ADAM COOPER: Sure. The balancers — I guess a challenge. We went into our system because it was designed to be a feeder system for other things. We actually started with our system with a very hard role, which what you guys are going to laugh about was we went in wanting to hold two weeks worth of content. We wanted to — everyone knows when UPS where it kind of eats its tail. We actually wanted our system to do that. So any of the content that isn’t — we kind of did an analysis of what the producers were using and roughly how long the content had to be held. And short of a very large long breaking news story in which case we have an entire archive system as you can imagine, we wanted our SnapStream to hold a minimum of 10 days, maximum 14 days worth of content because we figured that was two weekends worth of spanning producer time generally stories lasted about that long. If they lasted, they could go in clip and take it. So we started centralizing. We did have a couple systems we centralized pretty quickly. The scaling we found — we had no very quickly. And I think we might had one of the first ones because we kept getting it and saying we need this all to be a single cluster. You know, as David was saying, the guy is just asking the really good for doing development. There are some pieces that have been less than moving forward the way I want them to. Redundancy and fell over as one of the big questions for us because as you if its in clusters. We do have those moment in times that if you have to upgrade or reboot another other things, it’s like OK. So how’s Thursday at 11:02 p.m. for everybody? And all of us and you’re frantically breaking down clusters and shows squeezes and turns coders and da-ta-da-ta-ta. But it’s actually worked pretty well. The other interesting thing is that as we’ve had SnapStream, we have a large trenchcoats system — everyone knows vantage? Telestream for factories, all those fun things. We have found a tremendous amount of uses for our systems. The biggest one that these guys were great building for us was XD can. So very quickly we need XD camp 50 as our base record format. They built us XD camp stuff that we could record MPEG 2, transcode it internally or handed off to our ranch’s system and send it over. You know, I mean you’re going to hear a recurring theme, I think, from most of us. Generally, (unintelligible) so please jump in. That if you bring into the dev team and there is a business case for it, they will find a way to do it. XD camp’s an obvious one. And a lot of broadcasters said this is a base added format. DNXHD 145, 220, whatever you were doing. Those types of things — like, bring it forward to the team because they’re very good at saying, oh, geez I’ve heard that from seven people now. OK. We’ll do it. So we ended up doing a lot of workflows based on that. Our system feeds what we call our screening room, which is our publicly available. I put that in quotes. This system by which our producers and everybody out on the Web can hit our clips and the different retention rate. It is a much lower bit rate that can go back and screen stuff. As we’ve gotten more and more workflows, everything from, you know, sending cups back and forth to the West Coast, screening — the digital team is now coming on board. We just opened up to a whole bunch of our producers. On the digital side, who do a lot of news lauding for the desks, we have a big, you know, our entire floor is desks, you know, national desk, international desk. The digital desk really wanted it so we just it to a whole bunch of new people because they wanted to be able to go back and look and say, oh, geez we are watching the CNN feed and I missed what he said. So we want to go back two seconds and see, oh, geez did he really say, you know, Mr. Gorbachev tear down the wall. Sorry, showing my age. What did I miss?
0:31:47:>>MATT QUERING: I think that’s great.
0:31:48:>>ADAM COOPER: OK. I mean, I’m intrigued about these guys because you’ve got kind of broadcast usage. And I’m intrigued about sports usage. It’s a whole different world.
0:31:57:>>MATT QUERING: Yeah, Gil, if you could — I remember the first meeting we had. It was you, me, Eric. And, you know, we were kind of discussing different use cases. And can you tell us about, you know, how the system has grown just across different departments?
0:32:08:>>GIL SANTANA: Sure. You know, I think what’s been successful at us and why with SnapStream it’s kind of a testament to your developers has been the ease of use of the platform itself. That’s why it has been so successful sort of like tool for the entire company. Almost every single user at SNY has the capability to utilize SnapStream. And, you know, that came about because the moment we started strategizing which groups were going to have like access to it, it just became inherent that every sort of department had a need to either see the content that we were seeing for QC reasons or you had our sales team that had no way of really taking in content, sharing it with the agencies to see the elements because a lot of times they would have four. And they want to show the agency. This is this great sort of like campaign that we — you know, this is what your ad dollars are being spent on. And it’s also an upscale to Chevy, to all those — so like sales didn’t have a workflow in place where they could easily go and share that content off. They’d have to go into our media manager. And he was being constantly inundated with pulling clips to burn a DVD. And it wasn’t just — the media manager wasn’t just doing that for sales. Our promos team was being inundated with these sort of like similar requests. Then it came down where you had it on your talent who would regularly go to the media manager and say I need a clip of such and such because they would build reels for themselves — right? — to promote themselves. You had communications, the communications department if you had someone who was, you know, coming into the studio. So all these sort of like areas had no easy to use platform to access this content, this great content that we had. And when we brought SnapStream on board, it was ubiquitous across the company. I mean and it spread like wildfire because of how easily was to clip a piece of content and then share it to anyone and make it open it from any device. You know, we built it, made — you know, I made sure that when it was being implemented, it was single sign on our corporate network so we could easily access the cloud. So I had all of the, you know, all of the cloud capabilities. The storage front wasn’t too much of an issue for us because, again, we were really utilizing our SnapStream for is, again, to go out to those agencies for our senior management team to go in every day kind of QC the content that you’re creating and how we can enhance it for the customers. Now they just regularly go in to conference rooms and in their office and literally just sit down on the set top box and bring up last night’s Geico sports night. You know, on top of producing the Mets and Jets, we do live content daily at night and it’s just kind of refining every piece of content that we do every single day. And it really is just that ease of use. That was why it’s been so successful. And, again, it’s testament to your development team. You guys did a fabulous job.
0:35:24:>>MATT QUERING: Yeah, I think the changeover from MPEG 2 to Hdot264 was was huge for that being able to scale it because I think there were, you know, a couple of pain points, you know, with just adding new users with that web player plugin.
0:35:35:>>GIL SANTANA: I agree. That was — and again, that made it — you know, in addition to users accessing it in-house — we usually put it on the corporate network. It’s also available to them remotely. So we have a lot of users who will sometimes — you know, when Wi-Fi is good VPN into the network. And either on their iOS device or on their laptop, we’ll be cue seeing content as they’re traveling in. Because, again, you’ll have a director that sort of — maybe he hired a freelance director that night and he wants to see how “Geico Sports Night” ran and or how “Daily News Live” ran. And now we have that capability whereas before that content was kind of siloed off. He’d have to wait until it came into the office going to a they, have an AP or PA pull up that content for them.
0:36:18:>>MATT QUERING: Great. great. Yeah, I would — Dennis, I want to, you know, talk about “The Daily Show.” You know, Dennis you’ve been there. You were there with Jon. You know, “The Daily Show’s” really, you know, revolutionized late night TV, essentially created a genre of television. And Dennis, over the years, you know, you’ve been building this long-term, you know, archive of recordings. Any time I go in to speak with prospective customers they oftentimes say, well, how does “The Daily Show” do it. And it’s a very valuable asset that you have. So my first question is, you know, would you consider selling it today to the highest bidder?
0:36:51:>>DENNIS MCMAHON: Jeff, what do you think? Yeah, Jeff says yes.
0:36:54:(LAUGHTER)
0:36:54:>>DENNIS MCMAHON: I…
0:36:55:>>UNIDENTIFIED MAN: (Unintelligible).
0:37:01:>>ADAM COOPER: You don’t need it. You got Bitcoin. Don’t worry about it.
0:37:04:>>MATT QUERING: But, no, you know, it’s a very valuable lesson. How do you, you know, protect that asset? You know that’s…
0:37:09:>>DENNIS MCMAHON: …The protection of it is difficult. You know, I brought Jeff into that because he’s who I’ve worked most closely. He actually inherited that responsibility from someone that went on to work for Oliver and Sampi and all that stuff. But, David, to your question we have an agreement where Jeff and some other people maintain those overnights because the repeat viewing showings are like cancer to this SnapStream system. I mean, the channels need to broadcast everything. And you’re caught in this middle point where you need to report 24/7 for breaking news. But breaking news isn’t all the time yet. I mean, I’m sure we’re moving towards nonstop breaking news, which is going to be a lot of fun. But protecting the asset difficult because hardware. I’m relying on hardware.
0:38:05:>>MATT QUERING: Right.
0:38:06:>>DENNIS MCMAHON: And hardware is inherently faulty. I mean, that’s its job. Its job is to work until it can’t work anymore. Like, any other piece of hardware, you’re familiar with a car, phone, an oven. It all works the same way. It’s all made in mechanical parts it will fail at some point. So it’s it’s sort of like racing the clock. And the only thing to do with that is to not let costs grow exponentially. You know, although hardware storage costs are coming down in size is going up and all that’s working in a good inverse rate is to sit down with the people who are reliant on it and explain this concept to them. And, you know, early on in my time at “The Daily Show,” you know, I was brought on almost immediately after the important decisions were made, which is always fun for when you get to a job and you’re faced with that. So then my job is then to, you know, to manage these decisions. So it’s after the fact of making people realize that the desire to keep everything is very natural but the ability to do it is very difficult. So now as you well know, we’re trying to transcode off of this stuff. And I think maybe in late 2013 we — it was posed. You know, my inclination and my training is, you know, what about redundancy. You know, it’s very important in the I.T. world. And I kept looking at that redundancy and just seeing duplication as my only possible solution. And the price tag of duplication is exactly what you paid for the first system, which is there’s a lot of money invested in this system and now you’re going to investigate it twice for the potential of it all going away in one fell swoop for footage that may or may not ever be used. And when you come to my company, which is we produce one show. We are not a conglomeration of anything. The financial responsibility is very difficult, you know, to do that. So then you get out the word that yo, if there’s a fire in this room or this building blows up this is gone. And they say no that can’t be. And then you say to not do it that it’s going to cost this. And then it’s — let’s hope there’s never a fire in this room.
0:40:40:(LAUGHTER)
0:40:41:>>DENNIS MCMAHON: You know, it’s just the way of the business. So the best thing that we can do is — thankfully I have a guy like Jeff to help me manage this and make some hard decisions. A lot of it is people don’t want to make the decisions because then the responsibility falls on them. I have no problem with that, which I considered asset for me in my position. But if you explain it enough to people. They do understand. You know, so going to H-264 is huge. We’re now raising everything over to H264 for moving off of some archival systems to different archival systems. And the hope is, you know, now I just stay on top of it. And every so odd years the technology will increase. But now once I get the file size down H264, it’s at least manageable and going ahead telling my users there has to be some limitations and you have to work with me here on this is important. And I think I’ve gotten to a spot where the people that make the decisions are willing to support those decisions that they’ve made. And once you get a couple of people on board like that, the explanation becomes that much easier. And you pray that there’s not another fire in that room, which we had once already. So.
0:41:57:>>MATT QUERING: Thank you. You know, I just got the two minute warning. Let’s talk a little bit about the future. You know, one thing that the future holds obviously is the push to the cloud. That’s coming on a lot. And then also, you know, I’d be interested in what you guys have seen. You know, like what’s the biggest new tech that you’ve seen in the last 12 months in new technology.
0:42:19:>>ADAM COOPER: Count backwards by 12. That’s a good question. I mean, look Every company is looking at cloud. Every broadcast company thinks the cloud is both its greatest brand asset and the biggest devil spawn. I mean, you know, you can’t look at it and say, oh, my God it doesn’t have the complete double-edge sword. From a SnapStream perspective, one of the things that I think is interesting is some concept of hybridization where you could have replication, federation, you know, a cord system in one place and, you know, subsets in other places maybe for media tagged and moved off. You know, virtualization is obviously a big issue. And the big question for a lot of us. And it’s a very different conversation to virtual a desktop a trading floor where you’re pushing numbers verses virtualizing media space. You know, 60 frames a second is really hard to do and virtualization, technically not impossible at the moment. So, you know, you get a lot of challenges. And then you get a lot of format challenges. We’re 720 P. You guys are H80.
0:43:23:>>DENNIS MCMAHON: H80.
0:43:23:>>ADAM COOPER: You’re 80. So their 30 frames a second. We’re 60. You’re 60.
0:43:29:>>DAVID HADLEY: We’re doing almost all our SnapStream 720.
0:43:33:>>ADAM COOPER: So you’re sometimes 60. So even you got a great division. Here this table is 720p. This is JDA. You know, so there’s a lot of questions. I mean, there are always directional desires. And you know, I don’t sit there and laugh and say half is road map is things I say we’ve got to go this way. I think hybridization is the next big thing for the statute systems. I think the ability to make that open more single sign on all those types of things. Coolest tech. That’s a hard one. You know, I — look there’s a lot of really cool tech. I’ve got to tell you. And this is kind of the joke that, you know, anyone who broadcasts go back a couple of years. Remember when we had 100,000 cameras. Remember the lenses were $30,00, $50,000, $80,000. You had to carry around huge packs and you just send out a team to shoot. Anyone realize that this little thing has more power to record and transmit than those packs ever did and could singlehandedly in a producer’s hand overwhelm any sense to me? So now every producer has this in their hand as they walk out the door. And these formats as we were all jokingly say are the pain of our existence. So I think quite honestly it’d be — I think probably one of the biggest technologies probably come into its own in the last 12 months. I would say bonded cellular, personally. I think the fact that we went from having to require satellites and big gowns and Kayibanda and sedans. It’s not like we don’t use it anymore. And we’re breaking news. So we’re a little bit different than this. But the fact that a producer can go and Optus to Periscope for podcasts back on anything. And we have acceleration now to get things back. Personally, I think fundamentally on the ability to get that latency down from a broadcast perspective was very interesting.
0:45:11:>>MATT QUERING: Right. No, that’s great. Anyone else want to add anything?
0:45:14:>>GIL SANTANA: Well, I think what maybe be exciting in our realm of live sports is just seeing sort of like lift off of video delivery of live sports events, like getting to the future where in that game we can start delivering that over IP. Obviously, because of costs.
0:45:31:>>MATT QUERING: Right.
0:45:31:>>GIL SANTANA: You know, having truck and satellite uplinks, I mean, those are huge costs. And in an area where we’re at. We’re in RSN. And you see the level of disruption that sort of like regional sports networks and just how television networks as a whole getting to the point where you can deliver the video live sports event in scale reliably around the world or just, you know, within your territory. I think it’s just an exciting thing for me because, again, like I’m excited to sort of, like, see all the areas within broadcasting that — yeah, you sort of look around and there are many single points of failures devices that just have no automation, lot of like barriers of entry where — it’s kind of like specialized. And it’s not technically specialized, like, engineering wise. But you just look at it, it’s got one power supply hooked up to it. And you go and you work with the engineers and you’re like that’s how you’re doing the workflow. And it’s like you’re just seeing that that’s where things are going. And, you know, companies like Google are already doing it for their live events and just now needs to go into the area live sports.
0:46:37:>>MATT QUERING: Right.
0:46:37:>>GIL SANTANA: We’re dipping our toes slowly. But, you know, my BP knows it. That’s where we have to go because that’s an area that we know we can cut cost. And if we can deliver it reliably to the masses, it’s going to be good for the business.
0:46:52:>>MATT QUERING: Yeah, that’s great. You know, I was just given to stop signal, but I would like to take a couple of questions if anyone has any questions for our panelists. Yes, Bob. Go head.
0:47:03:>>UNIDENTIFIED MAN: So this is not necessarily just for this panel, maybe the previous panel as well. What I’m hearing is that you folks are have trouble to get a hold of stuff and operate SnapStream and it would really help out to find people to get the train running. Is there anything that SnapStream can do to kind of help you out. I’m not doing necessarily a dating agency. But I know in a few future journalism schools and in some universities. Is there anything we can do to get you people able to use SnapStream right out the box?
0:47:40:>>ADAM COOPER: I would actually throw the question around to say from people who work in broadcast is it really that hard because any producer worth their salt works in Avot from here or anything else. SnapStream is, I mean, a 10 seconds thing, 30. Part of what we loved about it was that the user interface was really cool. I would argue that your buttons might try to be a little clearer and, you know, we can change the iconography. But, I mean, please anyone setting up a panel, like any of our producers if I showed to them for all of two minutes, that’s about as long as I have to teach them usually.
0:48:13:>>DENNIS MCMAHON: Yeah, I don’t think the difficulty is in the teaching of of the application itself. But I don’t think you’re ever going to be able to teach the understanding of the application on the environment. You know, at least from our standpoint, that’s where we come from. But if you give a producer the ability to do anything that they want to do, they’ll see that and that’s it. You know, and that’s just is unteachable.
0:48:40:>>ADAM COOPER: But I will say there are pieces that really do need some work. And I think the administration portions absolutely do. And I think as we look forward to the more complex systems get and the larger and the more pieces, one of the things that we’ve lost is we’ve gone into the new SnapStream technologies is we used to have a really nice interface to administer all the tuners are in one spot. The question could do everything. Now that you have encoders here, another type of encoder here, tuners here, there is no central UI anymore. You literally have to go to a box, a box, a box, a box a box. And then you get into certain areas that are centralized. Like, OK, I can designate these incoming streams as globals. That’s where the challenges come. Things like managing records and seeing conflicts. As we start having more and more channels and more and more streams and all those things, we need much better tools to be able to visualize that very quickly because it is no longer a limitation of the tuner allocation. It is like, oh, geez my, you know, 26 tuner is going to throw out my 20. You know, there’s no way to visually see very quickly from a record perspective. Here’s a quick way to see what’s going to conflict.
0:49:49:>>MATT QUERING: Sure. Yeah. And, David, I know you’ve, you know, brought this up as well. So if you like to expand on that.
0:49:54:>>DAVID HADLEY: And also just be able to see all my tuners at once where you going to say I want to see what’s on tuner x. Not — you know, and you can’t. There’s no easy way to do that. There really isn’t. And it seems like such an obvious thing. I know it’s not as easy as it sounds. But, you know, you have to go through a whole process to do that when it should just be what’s on tuner 26. You know…
0:50:20:>>MATT QUERING: Sure.
0:50:21:>>DAVID HADLEY: …There it is. It’s working. It’s not.
0:50:24:>>MATT QUERING: Yeah. Sure. Hopefully, you know, we’re hearing you guys loud and clear. And hopefully, you know, 2018 and beyond, you know, we can address some of these.
0:50:30:>>ADAM COOPER: There’s a question in the back.
0:50:33:>>MATT QUERING: Yes, sir.
0:50:33:>>UNIDENTIFIED MAN #1: I’m not picking up on this term tuner. And, one, I’ve used SnapStream in two different environments.
0:50:40:>>MATT QUERING: OK.
0:50:40:>>UNIDENTIFIED MAN #1: It’s amazing listening tool to the world. Like, the world needs linear cable or broadcast to social media to text to Netflix, Hulu, Vimeo, etc.. How does — how do you all in your environment think about a listening post beyond the tuner? And how does the company think about providing its clients with that same power in media that podcasts as a whole and there’s so much media happening and much conversation happening? Are you looking to other forms of media apply the same rubric of listening, slicing, tagging, clipping and sharing?
0:51:22:>>MATT QUERING: Yeah. I mean, I would probably echo…
0:51:24:>>UNIDENTIFIED MAN #2: You mind giving get your background. Like, you mentioned you use SnapStream.
0:51:27:>>UNIDENTIFIED MAN #1: Oh, yeah. I told you. I used to work on “The Daily Show.” Dennis and Jeff — I used to work at the Onion. I’m on my own now. I do writing. I’m a multimedia human. And so I love SnapStream. I’m like, oh, that’s a superpower. But now I want it in other domains. Like, we’ve conquered — you’ve conquered a world but there are more world than these. So I’m curious to how you think about that.
0:51:52:>>UNIDENTIFIED MAN #2: Yeah, we talked earlier about how we’ve take in IP streams now as sources. We’ve only take in the IP sources that are available to us. So we take in UNP, RTMP. We’re adding HLS and MPEG data. It will be release — it’s coming out next month. So somebody that’s doing it over the top live broadcast channel can now, you know, basically send that two place, can send that to their CDN that’s pushing out to the rest of the world. And they can send it to SnapStream where they can record it, live clip And do all the things we can do with live streams and SnapStreams but from this IP stream. But does that mean that we can, you know, take in Youtube videos or something that’s on Netflix — you know, Hulu Live came up in here earlier. I don’t know that’s what they call it. But they’re pushing out live channels through Hulu. We don’t support those right now. We — probably getting support for those would involve going to those like over-the-top providers and the striking deals because those are all, you know, access controlled streams. So not that easy. You know, technically it’s not really a technical problem. It’s more of a licensing problem when it comes to ascertain some of those over-the-top things. But it’s certainly — it’s definitely on the radar. We’ve gotten customers and they’ve said — they’ve asked us, hey, why don’t you record Netflix. I don’t even think they know what they want.
0:53:22: (CROSSTALK)
0:53:23:>>UNIDENTIFIED MAN #2: So thank you. We’re definitely thinking about that.
0:53:35:>>MATT QUERING: One more question. Anyone. Alright.
0:53:38:>>ADAM COOPER: I’m sitting here so you’re challenged.
0:53:41:>>MATT QUERING: All right. I guess, we’ll leave it there. Well, how about a round of applause for our panel?
0:53:43: (APPLAUSE)
