Michael Kammes: Good to hear your voice Larry, thank you very much.
Larry Jordan: You know, this move to BeBop is relatively new; how’s it going?
Michael Kammes: I tell you, some days you feel like a rock star and other days you feel like the guy who’s cleaning up after the rock stars. What’s very interesting is that, over the course of my career, it’s been about getting a faster horse; right, more horsepower, faster storage, just incrementally make things better. Even moving from tape to tapeless was an incremental step; now that I’m working at BeBop and everything is based around Cloud methodologies, it’s no longer trying to get a faster horse, it’s introducing people to Teslas and cars; it’s a complete different way of thinking.
Larry Jordan: That’s what I want to talk about; because, this week, we’re looking about interesting ideas that have caught our attention. What’s an interesting idea that’s occurred to you?
Michael Kammes: In retrospect, it almost seems like, why didn’t I think of this earlier? You know, in post-production especially, we’ve been relegated to the technology that is on our desktop, or on the floor next to our desk, or in a machine room that’s down the hall. It’s very rare that we leverage the power of data centers around the world to do a lot of the heavy lifting that we’re still relying on our local machines for.
Larry Jordan: But, when you say heavy lifting, a data center is essentially a server that’s located somewhere other than our site; but a server’s being shared by multiple users. How does moving to a server help us, because, now we’re time sharing?
Michael Kammes: That’s a great way of looking at it and I’ll address the second part first. When you’re dealing with data centers, yes, it’s brokered time; but the work that you may need to do may not require you to have that investment of that $9,000 Mac Pro. You may be able to get by with paying a few dollars for compute time and then be done with it. There is something to be said for the financials of something like that.
Michael Kammes: The other angle is that, when we look at these data centers, the machines that are in data centers are far faster than just about anything any of us are going to have on our desktop, or again, even in our machine room. By leveraging the power of these machines, whether they’re shared or not, to render faster, to export faster, to playback in real time what our local machine can’t and to use the bandwidth that these data centers already have to the internet at large. There’s no reason we shouldn’t offload all of those responsibilities to data centers, instead of relying on what we have at home, or at work.
Larry Jordan: I want to bypass the first mile, last mile speed challenge; instead, I want to look at the workflow concept. Is our workflow the same if we’re editing using locally attached storage, versus editing versus the Cloud?
Michael Kammes: That’s a great question and I think, when folks try and look at using the Cloud, the first immediate knee jerk reaction is to use one of the great sites like Frame IO and even YouTube, to some extent; where it’s the last mile for the review and approve process. That’s a fantastic way of using the Cloud, but that only augments what you’re doing on Prem. You’re still following the workflow you’ve been using for years.
Michael Kammes: When we start introducing the Cloud into post-production workflows, we have to look at, well, do we want to push high resolution media up to the Cloud and edit with that? Sure it will take longer, but the machines up in the Cloud can handle it faster, can render it faster and then you can deliver from the Cloud, as opposed to bringing the media back down to wherever you’re working and push it out from there.
Larry Jordan: It sounds to me that, a good alternative would be a proxy workflow; store the high resolution media locally, feed proxies to the Cloud and use the power of the Cloud for proxy editing.
Michael Kammes: You’re completely right. If you want to cut down on the time it takes to upload, you can certainly utilize proxies and then you can localize the project file and reconnect to the high resolution. In that case, then it falls somewhat into our more traditional offline and online process. However, if you’re uploading proxies, you now are stuck with what proxies are in the Cloud; which means, you always have to localize.
Michael Kammes: We also look at things like disaster recovery and just simple backups. If you’re working on Prem and your non rated drive goes down, or your SAN, a drive goes dead, unless you’ve got back-up somewhere, you’re in a really, really tough spot. If you’re working in the Cloud, traditionally, there’s snapshotting, there’s back-ups, there’s what they call five nines uptime, or 99.999% uptime. While there is a longer upload time with media, you have all these fringe benefits, including disaster recovery already in the Cloud.
Larry Jordan: Should I mention last week’s Facebook’s outage of multiple hours, in terms of 99999 reliability?
Michael Kammes: I believe that, for Facebook, it wasn’t for where media is usually stored, internetable fashion, like Amazon, or Google, or Microsoft.
Larry Jordan: Alright, well I won’t mention it then. Where do you see the best use of the Cloud? Is it really the editorial process? Is it more in review and approval? Is it more in distribution? I mean, do we really have to commit our entire workflow to the Cloud, or can we pick and choose?
Michael Kammes: You can certainly pick and choose and you can certainly have hybrid workflows; because the Cloud, as you pointed out, isn’t perfect in every aspect. For example, if you want to grade something on a server in a data center, if you’re doing high end grading, you’re going to need something that can transmit that from a data center in real time and, right now, that’s only in theoretical deployments. It’s not something that we can just turn on and it just works. There has to be some concessions somewhere.
Larry Jordan: What’s the best option? If we’re nervous about the Cloud, where should we start dipping our toe in the water?
Michael Kammes: I think a lot of folks will dip their toes in the Cloud water by saying, okay, maybe I’ll use TeamViewer, or maybe I’ll use Remote Desktop, or maybe I’ll use any of these free, or almost free apps for my media workflow. I think what most people have to realize is that, many of these tools are not meant for the niche media and entertainment industry; so it’s important to find Cloud tools that are optimized for media and entertainment and then exploit those, as opposed to trying to fit a square peg in a round whole.
Larry Jordan: What are some tasks that the Cloud is eminently suited for?
Michael Kammes: That’s a good question Larry. Workflows are a lot like dominoes, or houses of cards. If you move one, then you may move all of them and that can be disastrous. If you’re looking for something that the Cloud can do well, that isn’t a big change or burden to what you’re doing now, then review and approve. I think a lot of services have already done that. I’ve already mentioned Frame IO, among a handful of others that are already doing this to facilitate that review and approve process.
Michael Kammes: However, if you look at data centers, they have more horsepower than you could ever have; so their best job is for the rendering and is for the exporting of media. The problem with that is, for you to utilize rendering in the Cloud, you have to have a good chunk of media up there as well, or some kind of proprietary application, which can render in the Cloud and then localize those results, without too much impact to your workflow. Unfortunately, there are not a lot of services out there that can do that for creative editorial.
Larry Jordan: You’ve been involved in workflow since you and I first met many, many years ago and now you’re really working for a company which is focused on Cloud services. What is it about the Cloud that most excites you?
Michael Kammes: The fact that it’s brand new and when I say brand new, I mean, brand new for the media and entertainment industry. If we look, at some of the leaders in our industry, in terms of creative software, we look at things like Avid, right? Their Cloud offering is just beginning. If we look at companies like Apple, in terms of creative editorial, they’re not even doing a lot that’s public, I should say. Then we look at Adobe, who has been dipping their toe in the Cloud for years, but still doesn’t have a complete Cloud solution.
Michael Kammes: It’s very exciting to see these tools that I’ve worked with for decades, that drive our creative industry in terms of technology and for them not to have a solution, that means it’s new technology and I’m uber excited to see where it goes.
Larry Jordan: Are you really that hopeful about media and the Cloud?
Michael Kammes: I am. My biggest concern is that, we’re too early. My concern is that, the bandwidth that we’re afforded here in the US will not increase significantly enough to make this viable. But the tech is there, the tech works, I really feel it’s going to be the future; just whether we’re too early.
Larry Jordan: Michael, for people that want to keep track of what you’re thinking and doing, where can they go on the web?
Michael Kammes: Thank you for yours Larry, take care.
Digital Production BUZZ: Michael Kammes – Successful Storytelling With New Technology, The Cloud & Distributed Workflows
On this episode of The Buzz, we talk with experts about the current state of technology and how...
BeBop Launches New Virtualized Editing Platform on IBM Cloud
BeBop Technology has built a virtualized editing platform that uses IBM Cloud to offer a secure,...
How to Set Up a 100% Cloud-Native Editing Suite with iconik and BASE Media Cloud
This webinar demonstrates how iconik and BASE Media Cloud can be used to create an editing suite...