Episode Transcript
[00:00:05] Speaker A: 3, 2, red 1.
[00:00:11] Speaker B: Broadcast and live production teams are being asked to do more than ever, often with smaller crews, tighter deadlines, and increasingly complex technology.
So how could automation help production teams work faster, reduce mistakes, and run more consistent shows without adding more complexity? So today we're looking at broadcast automation. From the simple solutions that let operators control multiple systems with the push of a button to automated cameras, production switching, and enterprise level orchestration that can coordinate an entire newscast.
Joining me today are Eden McSpadden, Chief Technology Officer at Arizona Public Media, who recently helped lead the technology transformation of a major new public media facility. And. And our very own Ed Locke, senior solutions architect at Keycode Media, who brings extensive engineering experience designing and deploying broadcast systems for facilities of all sizes. Ian. Ed, thanks for joining us.
Okay, let's kick things off with some clips that have made the rounds over social media. So this news anchor is presenting the news while also operating the video switcher, managing graphics and controlling audio.
[00:01:23] Speaker C: Night two of Punching my own show, be my own director, my own producer, my own host, my own audio director, and managing all of this. Welcome to what the new form of local news looks like. Ifb, an hour and a half of this. Okay, so now we're in an interview, as you can see it playing out right here. And I'm gearing up to go to break. Then we're gonna come back out. I'm going to intro a sports betting story. Cue that up. Is how I queue up everything that the audience can hear and my own mic. So I have to remember to do that. It really is a lot. Plus trying to focus on accurate information anywhere. Houston live and local. We'll be right back.
[00:02:02] Speaker B: So, Ian, I'm curious. What's your reaction to seeing someone responsible for both delivering the news and operating technology behind it? Right. Does it, does this make sense,
[00:02:13] Speaker D: you know, that particular instance? Not really. Because one of the huge challenges in deploying that level of responsibility is how many people can actually do it. Well, you know, you may have one person that can strap the drum to their back, play the harmonica and use the accordion at the same time, but I can guarantee the majority of their staff probably can't.
[00:02:35] Speaker B: So, Ed, you also work with broadcast facilities of all sizes, and is this an example of where automation can meaningfully reduce the burden on an operator?
Or how does it kind of maybe highlight a larger problem with low production workflows? With how low production workflows are being designed?
[00:02:53] Speaker E: Yeah. Well, first I just want to say watching those clips caused me physical pain. I really felt for them. And what they were having to do. I wouldn't say it's a production workflow problem, it's just that that particular workflow, I don't think scales down to a single person.
I think automation could help. I was thinking, you know, maybe we could pick like the top three tasks that they're doing and see if we could script them down into, you know, simple button push.
I was actually thinking, I know this wouldn't be typical, but I was thinking of like a QSC core because you can build a lot of audio workflows that can automatically do levels and mixing and just maybe take some of the hassle, take the whole audio workflow away from the host. But honestly, I don't think even automation is the real solution here. I think that there's just this conflict between the role of the host in that situation to engage with people and being tasked with also engaging with technology. I think it's just, you're just always going to run into a problem there. I don't know that there's one great solution for what, you know, these people are being tasked with doing.
[00:03:59] Speaker B: So is the suggestion that there kind of is a balance and this isn't. This isn't the balance.
[00:04:03] Speaker E: Yeah, it's not. Not the balance. I mean, you know, you know, a junior TD just, I don't know, just someone to take a little bit away. You know, I can see, you know, having, you know, putting some control in the host's hands, but, you know, being responsible for the whole thing just feels like a contradiction in goals.
[00:04:21] Speaker B: So when we are talking about automation, then the clients that you're working with, what does that mean for them? We talked earlier about automation means a lot of different things to a lot of different people. And people categorize things in automation, maybe they shouldn't. But what types of automation are kind of the most popular, that is kind of the most requested that you're getting.
[00:04:40] Speaker E: So, I mean, again, automation is just a really broad, broad topic. So first you have to understand exactly what they're thinking when they're automation. Are they thinking kind of like traditional, traditional production automation solutions, you know, camera switchers, graphics and things like that, or are they thinking workflow, automation and those types of things? And I would say it's honestly, I mean, it's just a little bit of this, a little bit of that. A big one that, you know, is always being introduced is just kind of, you know, media file automation. Right. How do we manage getting all of the content into the FAC facility and getting all the content out of the facility. So that's, you know, actually and you know, it was like not really considered for a very long time as that's in their department.
[00:05:21] Speaker B: Yeah, that's in their
[00:05:22] Speaker E: department. But it frees up so much time to actually focus on the creative aspects of what you want to do. And then, you know, you know, automation is always, I think, something that is applied very specifically to address a business need.
And that's basically you're just evaluating the workflow, identifying pain points and saying, hey, here's a strategic place where we can, you know, apply some automation. So I mean, I, sorry I can't give you like a more of a precise answer, but I mean it is just a broad topic. It's a conversation. I highly recommend people engage companies like Keycode Media because we have these conversations all the time and we can just at least make that path towards identifying the right automation a little bit easier and a little bit faster.
[00:06:04] Speaker B: And Ian, from the perspective of running a public media organization, how do you identify which parts of a workflow are actually worth automating?
Are there specific tasks or operational bottlenecks that immediately stand out for you?
[00:06:17] Speaker D: Well, I think, you know, with our unique facility in that we do everything from student assisted productions to having clients come in and use our facility, that there are cases where we need both the ability to do semi or fully automated production tasks, but also in the case of where if we're doing a full production and a company's coming here and renting our spaces, they're going to expect discrete access to a switcher, to an audio board, to graphics, to playback and all those separate elements.
So, you know, part of what we were trying to do was make sure in the design that we could try to scale those. So in scenarios where we're using our control room and a student was doing, our student led crew was doing a very simple production, that there was enough automation in there. It didn't tax their skill sets. Where when professional folks come in to use that same space needed all the bells and whistles accessible. So in some cases there's some automation using stream decks and other devices to simplify calling up certain settings and configurations, which is great. But then you have full access to all the regular broadcast interfaces via kvm. So when a full expression operator or VDIO operator comes in, they have access to what they're.
[00:07:35] Speaker B: So that sounds like a great example of what automation can be for each individual user or position. Right, so you have the simplified interface and you have the more tech centric or operator interface. But I'd love to hear about maybe other automations that you've deployed. And I'll ask you as well, Ed, automations that you've deployed that made a noticeable tangible difference to the production team, is there anything that kind of stands out in addition to what you've already mentioned?
[00:08:01] Speaker D: I think like Ed mentioned too, to make the workflows simpler but also more robust too. Because one of the challenges is with any kind of automation system is you need to make sure that it's escapable.
Because I've certainly seen folks who rely overly on automation and if something changes or deviates, trying to go out of automation into manual mode back to automation could be a huge challenge.
So folks need to be very careful when they choose what they want to automate and to what level. Because if you're overly reliant on automation, if something goes wrong, it's hard to get out of. I remember watching a mood broadcast where they were doing some storm coverage and they were running sort of ELC and going through the usual templates, but they wanted to go live to somebody and they obviously punched manually to a live shot. Well, they punched the wrong on up, it was somebody who was chasing their umbrella down the street. Then they tried to punch back to the set, there was no audio.
You know, you could tell that it was that automation director who hit the wrong button, got out of sequence and it just kind of, it cascaded, you know, the, the viewer or listener at home doesn't know what mistakes you might be making unless they see those.
So that's one thing where Ed mentioned earlier, where having someone who's sitting off to the side manning the automation, where your actual talent or presenter is presenting, they can keep talking, they can ad lib, but if someone's behind the scenes needing to make adjustments, fix things before they go live, there's less likelihood that the viewer at home knows that something went wrong.
So it's, it's, it's making sure when you design these type of things that whether it's your, your post producing it, you want to simplify workflow, not have to do stop pounds. But if you're live, you automation systems in place that don't have so much control or something doesn't work properly, you still have the ability to exit it and then gracefully reenter.
[00:09:56] Speaker B: And is that a judgment call by the designer? I mean, it almost feels like there's a gut feeling of when is over engineering, when is under engineering and when is automating for the sake of automation. Right. Isn't the joke if it, you know, if anything can be automated, you'll spend time automating it. And the time you spend automating it will drastically eclipse however much time you're going to save doing that automation. So where, where is there any kind of rule of thumb as to where to draw the line, when to automate, when not to?
[00:10:25] Speaker D: You know, I think it's important to have a discussion, you know, one with the folks and, and assess the skills of your operators. Because, you know, not every. Some facilities may have really highly skilled operators, some may have lower level skilled operators. So it's understanding the capacity of the operational staff who will be running the equipment and their trainability, and then it's having that conversation with your design partner. Whether you're designing in house or using a consultant like E Code, they need to know what you're capable of doing internally, capable of supporting. Because one of the other challenges is if you put a really sophisticated system in and it's easy for people to use, but the engineers behind the scenes can't keep it running or can't make changes easily.
There's a lot of different angles that I think an outside designer is helpful in coming in as a mutual third party, helping the organization assess what their team skill sets are, what they're trying to do, what they can maintain, and then what their options are. Because, you know, you can certainly say, here's the fully automated version, pros and cons, and here is maybe a hybrid version, which one reaches, you know, your goals best and, or sometimes the budget. Because sometimes having automation and simpler systems actually fits a budget better than having higher systems if they're not needed. Because if you don't need expertise to run them or maintain them and the cost of actually buying them and something simple will work, why spend more money than you need?
[00:12:00] Speaker B: So, Ed, I posed kind of the same question to you, is how much technical expertise do you find that end users need to have? Is that they need to be able to design the automation or is it just they need to troubleshoot the automation or build it from the ground up? What is kind of the technical expertise that you're normally seeing folks need?
[00:12:16] Speaker E: I mean, again, it's levels of automation, right? So if we just start with, hey, you know, I want to do this simple, you know, scripted automation type things. You don't need a lot of expertise. You need to understand device control, networking, you know, as any good engineer should be able to do, troubleshoot to work the kinks out as you're trying to build the automation.
But at the End of the day, the key is, and I think Ian hit on this a little bit, it's an engineer working with a really talented operator identifying what to automate and how. Because you can load a routing scheme and set this parameter and arm a record channel and do all that, but why are you doing it? And that's where you sit down and you have meaningful conversations about what makes the most sense to actually spend the time building. So that's kind of at the top end. But that story changes dramatically if you talk about enterprise automation systems. Okay. You need to have a commitment as an organization to have the technical and engineering resources necessary to both participate in the design, deployment and commissioning. But then of course, the ongoing maintenance of those types of systems. Because those types of systems, you know, they're not, they're not tinker toys, right? You're doing user management, you're monitoring the health of the system. And just by the nature of these enterprise automation systems, you're constantly either tweaking workflows or identifying new workflows you want to build. It's kind of a living beast.
And it's not so much as you go to kind of the top line scripting down, you kind of identify these workflows that you want to do. And some time might go by, you're like, hey, I want to either change this or maybe I want to do this. But it's not a super involved process the way, you know, kind of foundational automation systems are.
[00:13:59] Speaker B: So I want to shift the story a little bit because I think we've been spending a lot of time talking about creating the automation and interfaces for experienced operators, right, Whose work they're doing is repetitive. Right. They've got it down to a science. But when we start getting into the novice levels or even some of the other places Keycode Media will build in automation, whether it be a podcast studio, whether it be a fly pack, and that's much more operator centric, right. More storytelling and less technical. So are there any kind of rules of thumb that you follow with operators in terms of designing automation for them? Is it sometimes like no more than four functions per button press or, you know, make sure that there's some kind of tactile feedback or a certain color flashes? Are there any good design or operator practices for things like that?
[00:14:46] Speaker E: I don't know. I try to avoid operators. Okay? You know, I build the technology, they need to use it. No, I'm kidding. You know, Ian mentioned something like this too. There is no right answer to that question. And you're probably going to hate this because you always want a direct. You want a direct answer to it. But you. Again, I think it goes back to something. Also, I said earlier an engineer working with a talented operator deciding what it is. Now, if I zoom out and put it in a vacuum, yes, there are some rules of the thumb that you can apply, and that is that pay attention to the subjective evaluation of what you want to build. It is so easy to say, hey, I need a user interface and it does this and it does that and it does the other thing, right? And then you can, you know, build that and an operator sits in front of it and it's completely unintuitive, right? They're like, wait, no, every time I think I want to do this, my hand wants to go to this part of the screen instead. It's over on that part of the screen. So you just need to kind of understand that there's a real subjective component to all of this and keep that in mind when you're designing. And I'll be honest, I would say it is more common than not that concepts that we develop during design, especially when there's long project cycles and it's going to be months until things are actually commissioned. When the boots are on the ground, actually deploying these types of automations, they're tweaked, they're tweaked in real time. So do your best to at least consider the human component of it. Engage operators where they're available. But you don't always know who's going to be operating these systems early on in design. So just I just do that recursive loop, right? The, the objective, what do I need to accomplish? And then the subjective how do I want to accomplish that and where do I draw the lines? So, you know, kind of a broad answer. I'm sorry, but, you know, I mean, we're talking about, you know, workflows here, and workflows vary and they're very dependent on environments and human beings.
[00:16:45] Speaker B: And Ian, for your specific deployment, what you and your teams are doing, how are you kind of approaching that design for the end operator, who may not be technical?
[00:16:56] Speaker D: You know, it's a combination of, you know, making sure that the design team is aware of the operating environment tools required. You know, placing too many keyboards or things out of reach or sight lines not in the right place can make it difficult, you know, so that's where that, that partnership with your design team is so important because they don't just need to know what you're supposed to be doing. But then understand the environment and where it's going to be done.
I'm big on putting together SOPs, you know, and being able to have a one sheet where someone who comes in can basically look at a basic folder or one sheet and say okay, this is how I need to turn things on. How I can do the basic setups. If it takes more than a sheet or two of paper with some basic instructions to begin to fire up, it's no longer simple.
You know, I worked at the Pentagon. The press briefing room There has a 3 camera PTZ system and they had a full production truck size video switcher and a 64 bus audio board to run that room. It's never more than one camera and one Microsoft for the majority, you know, and the, because they have, they, they cycle folks through like every six months. As far as their, their military operators, you know, they had this binder of things for them to try to figure out and how to run basically all the one camera, one mic room, you know, so making sure that you design the room and systems, you know, that are purposeful and, and right sized. Because again you don't need a, a calypso switcher to run three PTZ cameras and be able to, I said if you can't distill it down to simple enough steps to operate that in a short amount of time, you can't train somebody or you can't put you know, a binder there with a page or two that somebody sat down and had the basic skill sets they could flip through and know how to use that particular equipment. And that's very much what we did here. You know, we were trying to look at, okay, if a student had to do a basic three camera production, we have PTZs in our studio, what would they need to do to fire up the system to be able to do that?
Now when the production crew comes in, we have our five big studio cameras that need to be in use. You've got a camera shader, you've got your tv, you've got a lot more people involved. So it's not a one sheet, it's trained people who know how to utilize those particular facilities. But the automated parts are usually the people that require that guide. Because again like the young lady who was doing the there, if she called in sick or was on vacation, who fills those shoes?
Can you imagine the binder of notes you would need to know all the buttons you need to hit and what sequence to actually pull out.
Yeah, you have to scale it in a way that just makes Sense for the team you have.
[00:19:46] Speaker B: So Ian, let's move around. Kind of the individual buttons, right. We're seeing a lot of automated PTZ tracking, specialty presets, auto advancing teleprompters and automated production switching, plus AI voice assistants. Right. So which of these technologies are you seeing adopted most often or most requested? And what kind of productions can kind of benefit from these? I don't want to say automation tools, but these enhanced automation tools,
[00:20:14] Speaker D: you know, one of the things that we did, we have a fully integrated facility. So our AV rooms have camera systems or zoom rooms. You know, in a lot of those spaces in our community studio we have Q system with tracking. So the cameras track the voice.
So again, in locations where we're doing productions that will either have low skilled operators or sometimes no skilled operators, you know, having those levels of automation that do automatic gain control cameras do their automatic tracking, I think are great conference rooms.
Our community studio uses those, you know, places of worship and other places where you might want to deploy a system that's very simple and requires minimal operators. I think the technology has really gotten much more reliable and smoother so you don't have to worry about, you know, really odd things happening.
You know, we have AI transcription for our captions, which I think is really useful. Having folks dial in and have to utilize those has always been a challenge, especially since so many of the older systems were phone line.
Having something internally that does it for you and the accuracy of the systems nowadays just helps speed up things, the process.
And one less thing that we have to worry about, you know, engaging and setting up locally.
[00:21:37] Speaker B: And are those kind of the, the, are those also the tech things that you're being asked about Ed, or that we're putting in or is there anything else?
[00:21:45] Speaker E: Yeah, I mean, you know, animated production. Animated, yeah. Automated production, switching, camera presets and things like that.
You know, I see a lot of potential in the like auto tracking, PTZ type stuff, you know, certainly in conference rooms and things like that. House of worship. I can see that in kind of the core broadcast space.
You know, I just haven't. I personally haven't had an opportunity yet to work on where that was like an integral component of it. And you know, part of the question was, you know, what productions benefit the most from it. And you know, I mean, this might not be a great answer. I think this whole interview so far is me saying, you're probably not going to like the answer. But you know, I think that these tools solve business issues, business concerns. Right? It is, it is you know, you're evaluating what you're trying to accomplish and you're saying, hey, automated, you know, switching or just having the PTZ do auto tracking, that's going to help me. I don't have an operator here, or I need this to run autonomously or however it works.
So the productions that are going to benefit the most from these technologies are the ones that need these technologies. I wouldn't consider it like my first go to design choice, hey, let's design an entire facility that just there's no humans anywhere and cameras are automatically doing what they're doing and you know, everything's just running on autopilot. I wouldn't do that. I still think having humans in the loop is good and valuable and reasonable. And it's just applying automation where it makes sense for your business model so that you can accomplish your goal. Right? That's the key. How do we get you to accomplish your goal? That business model doesn't always support a human being in every chair. So we apply automation where we need to in order to facilitate, you know, getting the product out that you need to get out.
[00:23:35] Speaker D: And folks who are deciding on whether they want to choose automation, it's important too to know what their product may look like when using automation. You know, do they need the exact crisp camera shots? Do they need organic movement of cameras throughout a show? Because the one thing that will happen when you engage automation is you use templates or you use AI or other kind of pre programmed things, shows or content can begin to look very similar. I mean, I can watch a newscast that uses automation and within a couple days I can call their next shots because they're using templates.
Saves them time, saves them functionality within a control room. But it also makes your newscast very repetitive. It looks very similar where on a, on a show that requires that you're doing a talk show and you really want to be able to have organic camera movement as conversations happen. Automation is not a great choice for that. That's where you really want. Whether it's a PTZ camera operator or somebody in the room with an actual camera doing those push ins, push outs and pans and stuff that automation doesn't do. It just not. It's not what it's designed for at the moment.
[00:24:41] Speaker B: So Ed, at the enterprise level, automation can extend well beyond individual devices.
Newsroom and production orchestration systems can coordinate rundowns, camera shots, graphics, video playback and other actions throughout the entire show. So how does that differ from the simpler automation we've been discussing and what kind of facilities does that benefit?
[00:25:01] Speaker E: Yeah, so, you know, just, you know, touch again, I was talking about the top down kind of versus bottom up approach. You know, the simpler automation kind of reaches down and touches individual systems and you can script, you want an order of events to happen. At the enterprise level, you're really kind of building the environment from the foundation up with automation in mind. In other words, you're deploying systems that are inherently interoperable and usually an orchestration layer that can then, you know, have full access to the full range of capabilities of all of the systems with an abstraction layer to where you're not just saying, I want this task, this task, this task, I want this operation to proceed, that operation followed by this next operation.
And you know, that can extend all the way out to, you know, out of the studio. Right beyond the actual production, maybe that's also, you know, automating, getting clips out to Instagram or, you know, you know, YouTube shorts or Facebook or doing all of this. So really at the enterprise level, it's a holistic approach to the entire environment and it's not restricted to any one kind of step in the overall workflow. At the enterprise level, you're orchestrating everything from someone bringing a camera card into the facility to as I said, someone publishing to a video on demand server as well as all of the social media sites and things like that and literally everything in between. And you're just, you're picking how you want things to work, you're picking where your human checks and balances are.
And you know, it's just a completely different model. Right. And that's one of the things that I really like about these simpler automation solutions is it used to be you had to at least deploy some of the foundation to get any of the automation. And now you really can just layer in some quick and easy tools that are both easy to deploy and easy to remove should you need to or want to for some reason.
So yeah, they're very, very different dynamics. Overall, there's a place not just for either or I think there's a place for both. Just because I'm saying I really like enterprise asset management and I think it has a tremendous amount of value. And if the cost risk benefit analysis works for you, I think it should be deployed.
I think I'd still see a stream deck, I think I'd still see companion, I think I'd still want to see all of these other tools that are there because everything kind of works, you know, builds on each other and also works in conjunction with Another. So, yeah, there's, you know, very significant differences in, you know, application engineering effort, but not necessarily, you know, a huge distinction in what they're doing for an operator sitting in the chair.
[00:27:53] Speaker B: So, Ian, you recently completed a major facility investment intended to serve AZPM well into the future. When you're making decisions at that scale, how do you evaluate which automation capabilities should be in the facility from day one?
[00:28:09] Speaker D: You know, we began the process by assessing what we are currently doing. Okay, holistically, look at your organization again. This is from conference rooms to radio control rooms to television studios to post spaces to digital distribution.
And then it's work with your staff and go, okay, what would we like to be able to do that we either can't do now or can't do well now? And then the third part is looking in the future, what might we like to be able to do down the road?
So when it came to automation, I mean, everything from simple stuff like, wow, I'd love to be able to walk into a conference room, hit one button and start a zoom meeting. You know, we have that now. In our old building, it was log in here, plug a laptop in there, fire up the projector over here. There were multiple steps to do something as simple as I want to do a zoom with a group. I, you know, it's the top of the hour, we need to start it up. We have one button access to do that in that space. Now, you know, calling up our VR sets, you know, we basically wanted to be able to have operators go in there and switch between shows because at some point we'll probably be doing shows somewhat back to back. And to be able to exit one VR set and enter a new one, change the setup configurations that are needed is key. Our old system was very complicated, didn't have that middleware that would allow us to do that. And it was hours in between bringing down one set and trying to launch the next one. So it's looking at what we were doing at the time, saying what couldn't we do? What would we like to do or need to do? And then just, you know, choosing in our case a platform of SMPTE 2110, all IP based, everything from IPTV to our conference rooms to make sure that we're structurally ready for what we might want to do in the future. And mentioning structure. Part of that is making sure that you wire your facility to make sure you could do it. Because you never know how much fiber, how much data, how much air conditioning, how much power you might need when you're building out a new space and trying to guess what you'll be doing in the future. We do green screen VR work right now. Down the road, we might want to bring in LED volumes which require more power, more air conditioning, and more data connections.
So any kind of automation, it's what does today's automation look like and what does the holodeck look like 20 years from now?
[00:30:28] Speaker B: So one of the biggest concerns with automation is what happens when something doesn't go according to plan. A camera doesn't move, a graphic fails to load, a device stops responding. So how do you think about resiliency and reliability when designing a facility that needs to support production, as you mentioned, for years to come when the holodecks arrive?
[00:30:49] Speaker D: Well, you know, one of the things that we didn't have in our other facility was the ability to cut away from automation that does our on air. So we have three channels of broadcast television. It runs through a scripted on air playback system, but we had no way to punch around that. So if that automation went south, it was patches and buttons and a lot of reconfiguration to take something else to air. You know, working with the systems integrators, key to say, hey, here's a problem that we have. We want to augment our current automation, expand what it can do itself, but then also have options for plan B. We have elections coming up and the ability to do something that we couldn't do in the past, which was we can transfer control outboard from our automation system to our television control room during election news, cut ins for them to do the portion they need, and then go back to our automation system, something we couldn't do in the other building.
You know, so looking at how those type of systems work and having that redundancy and backup options are key because going back to that very first video, if that gal had a system crash in front of her or something, glitch, how does she work around that? You know, there isn't what I could see a clear way of doing it. So you don't want to over automate something, but you don't want to under automate as well.
[00:32:07] Speaker B: Ed, for the facilities you've worked with and engineered, what are some of the best practices for making those workflows and automations recoverable and also still leaving hooks in there for an engineer to go in manually and fix things on the fly?
[00:32:22] Speaker E: Well, I mean, if you consider that one of the objectives of deploying automation is to be able to do more right.
So you have to be able to identify what I would, you know, what I've called like a minimally viable product. Right. So I'm adding these layers of automation and these cool tools and these cool features so that I can really increase my production value, for instance, or so that I can, you know, operate with this particular arrangement of staff.
If you're not thinking about what is my minimally viable product and how do I accomplish that, if I automation fails, then you are setting yourself up for, you know, some kind it will happen eventually where something will fail. And if you haven't already kind of connected the dots between, you know, what is my gazins and what am I getting out and how can I accomplish that in a minimally viable way.
Document that as a workflow. Make sure you're testing that as you're deploying the automation, you know, you know, not necessarily turn the automation off, but go manually, touch this, manually tweak that, manually turn that knob and make sure that you can still produce that minimally viable product. And that type of conversation is just really important. Do not design a super high tech everything automated facility and just assume that nothing will ever go wrong. Always be thinking about that path that not disaster recovery so much, but just what are my alternative means of accomplishing my goal? So it's just a conversation that you need to have and don't forget to have it because you'll be reminded at some point and it won't be fun. Yeah.
[00:34:02] Speaker B: So Ed, what's one or maybe two common mistakes that you see quite frequently when you're approached about automation, whether it be enterprise or even something simpler?
[00:34:12] Speaker E: There's one common mistake and it is just prolific. And I have to talk clients down every time and they want to do too much right away. Okay.
You do not want to just assume that you're deploying this new environment and that you know exactly all of the automation that's going to actually add value.
I always advocate that you kind of start in this phased approach, this kind of foundational environment where you know, there are some kind of four things that you know, you want to automate and then there are some workflows that you probably want to automate. But if you try to choose exactly how day one, you're going to end up in a problem. So the automation can absolutely get in the way of the way people want to work.
So, you know, especially when you're embracing on a bunch of new technology, get that technology in there, get your hands on it, use it, spend some time working internally to massage the way you want it. To operate and then say, okay, now I'm going to automate this. And at least you know that for, you know, maybe six months before you get tired of the way that works and you want to try something new. At least you know, there's going to be some longevity for it as opposed to just having to, you know, knock your head against a wall from day one. Because automation was designed not in a vacuum, but outside of the context of, of the reality of sitting in the chairs and trying to do the work. So I mean, that's the biggest thing 100%. It's a conversation I've been having for dating myself decades.
And, you know, it's still a conversation that is top of mind every day. So that's the biggest mistake. Approach automation, you know, as a step by step methodical process, you know, and start as minimally as you can to where it still, you know, adds value to your day one operations but doesn't cause unnecessary problems as you try to grow into a more robust workflow.
[00:36:14] Speaker B: That's a, that's a great one. And Ian, is that at the top of your list as well?
[00:36:19] Speaker D: When I agree in that I approached it in doing it in pieces and we basically have looked from the very beginning of our workflow to the very end, you know, so we edit in DaVinci Resolve and we're using Claude AI to write scripts that when they bring content in from the field, it pulls it into Resolve server, creates skeletons on timelines, sorts media into bins, and then on the way out, we employed a telestream Vantage system that when we have a finished product that we want to distribute across different platforms, we can drop that product in there, caption it, extract captions, do a vertical video version of it, do our YouTube video of it. All the different things that would in the past were being done by an editor, exporting those all individually. So we automated ingest and we've automated final distribution. So those were just two small things, but had a huge impact on time and cost.
And then you kind of find the things in the middle that work as well too. I mean, some may seem like small things, but when you're talking about saving a producer an hour worth of time in loading content in, or saving an editor a couple hours on Transcode and distribution, those add up, those are big wins.
[00:37:32] Speaker B: So for folks who are interested in learning more about automation and some of the possibilities, do you have any resources that you could share that people may be able to learn more about this?
[00:37:42] Speaker D: Sure, I can probably provide a couple links Online to resources that we've had a lot of. It is again, talk with people who are using similar systems, find out what they have, how it's worked, whether they've liked it. Because a lot of what we did going into the project was exactly that. Because there are things that may sound great but deploying them may be difficult or they may not have the return that you want. So when it comes to, you know, exploring AI, you know, obviously it's, it's task your editors with that. What are folks doing in the field? What are people, you know, lauding? What are people complaining about? I mean we found about the Claude plugin to DaVinci resolve because one of our editors had discovered it online, started playing with it at home, brought it in and said, hey, look what we can do. And we began to integrate that into our workflow. You know, I brought the skill sets of transcoding farm management from previous projects and saying, hey, on our back end, I'm watching my editor spend an inordinate amount of time, you know, trying to create and distribute content versus let's get the, the automation system that does that. And you know, Telestream Vantage is one of the bigger ones that's out there. It's an enterprise class system that will do a lot of things for you. So there are manufacturers that do a lot of different things along the way from newsroom automation. You know, I mentioned Sony elc, Ross Overdrive, Grass Valley Ignite. Those are tremendous tools if deployed properly. A lot of stations around the country are currently using them. But I think a lot of the fun stuff for the smaller ones too that are niche, whether it's like I said, the ingest or the output that actually save you the most time. Because just eliminating a few bodies in your control room or in your studio might save you money. But what really saves your existing staff more time, makes them more efficient, makes them able to produce a better product and more of it. Those are the automation tools that I always to want look for.
[00:39:35] Speaker B: Well, I've absolutely loved the Ed and Ian show, so Ed, thank you so much for being with us today. Ian from azpm, thank you so much for being with us today as well. Thanks for watching Broadcast to post. Don't forget to follow Keycode Media on social and contact us about your projects at keycodemedia. Com.
