GRASS VALLEY, Calif. — The Grass Valley Group had just rolled out the Kadenza digital production switcher, and there wasn’t a new, big digital video product on the horizon a little over a quarter of century ago, recalls John Abt, founder and CEO of AJA Video Systems.
At the time, Abt — an electrical engineer with a bachelor and master’s degree from California State University — was an engineer at the Grass Valley Group where he had worked on three of its pioneering digital video products during his tenure with the company.
Today, as Abt celebrates the 25-year anniversary of AJA Video since its founding in May 1993, he reflects on those early days and the products that put AJA on the map and shaped its direction while looking to the future, especially as it relates to IP video transport, 4K and 8K video and where the company goes next.
Despite being the company’s CEO, Abt spends most of his time on video product design, pursuing a passion in communications technology that took hold of him as an elementary school student when he became a ham radio operator. Abt spoke with TVTechnology from his engineering lab.
(The following is an edited transcript.)
TVTechnology: Tell me how AJA Video Systems came to be 25 years ago?
John Abt: Well, I had been working at Grass Valley Group for seven or eight years, and was fortunate to work on three pretty big and exciting projects. Really, their first digital projects — the Kaleidoscope DVE, the Model 3000 switcher and the Kadenza digital switcher.
After that, there wasn’t another big one on the horizon. Grass Valley was starting to struggle a little bit.
Having been involved in those digital products and having pretty good knowledge about what was going on with the conversion to digital at the time, I knew I could make some small little converter products that I would probably be able to sell.
So I just decided to break out on my own and go after that. Fortunately it worked out.
TVT: I think a lot of people wonder about the company’s name — AJA Video Systems. Where does AJA come from?
JA: Yeah. It was my wife’s idea, and it’s our son’s initials — Abraham John Abt. He works for the company in sales and has been here for eight or nine years. When we started, he was eight or nine years old.
TVT: You mentioned making some of these smaller glue products. Have you found that doing so gives you a better perspective on what’s going on with customers who are putting systems together and what they need.
JA: I think so. Essentially, the initial two products we did were parallel-to-serial and serial-to-parallel converters.
Digital was just coming on, and there were a number of parallel products out there.
Sony invented SDI and introduced it at NAB in Dallas, I think that was ’86. I remember standing there checking that out thinking “This is going to be the way of the future.”
Parallel digital was a nightmare. You had those 25-pin D-Connectors. The cables were expensive and unreliable. They wouldn’t go very far.
I had literally worked on a lot of parallel digital equipment. The first Grass Valley products were all parallel digital.
It was pretty obvious being at the street level, seeing what was being built and what was sold that all of this parallel was going to go to serial. Hence the need for converters.
If you remember back then, there was D2 composite digital video. There was PAL composite digital and NTSC composite digital, and there was D1.
Our innovation at the time was that our initial product was multistandard automatically. They worked with both flavors of D2 and D1.
All of the post houses in the L.A. area, which was our initial primary market, had to deal with all three of those constantly. They were making masters that went all over the world.
So, our initial product was pretty successful, and a good thing too because that was right about the time I was running out of money.
TVT: What other products proved to be breakout or turning point products?
JA: Well, the first number of years, we were just making those little converter boxes. Then we had D-to-A and A-to-D converters; all types of mini converters.
That went on up until 1997 or ’98 when we started getting into PCI cards. We made PCI cards that were essentially still stores. Some of them you just plugged into a computer, and the computer could download a still to the card and it would come out SDI.
We started selling those to still store vendors, and that was the beginning of our OEM business. So, that was a big change.
Eventually, those PCI cards became fast enough to work with live video as opposed to just stills. And at SIGGRAPH in 2001, we introduced the first KONA card — an SD card and an HD card.
We really had the first HD PCI card that worked well with Final Cut. That was also the beginning of our relationship with Apple.
Back in those days, if you were selling a video I/O device that worked with Final Cut, it had to be certified by Apple. Maybe it still does. We’ve worked with them for so long now that we are kind of grandfathered in there, but we still have to get approval from them every time we make a product that has ProRes.
In the early 2000s, we started making the rackframe cards compatible with the Leitch frame and our own frame.
In the mid-2000s, we started making frame synchronizers. We may be the No. 1 vendor of frame syncs. We sell a lot of frame syncs. We have five or six of them now — all the way up to 4K and multichannel HD. They are considered a “Swiss Army knife” because they do all kinds of conversions.
The next big thing was in 2009 when we introduced our Ki Pro recorder. We literally had the very first ProRes hardware license Apple signed off on.
The first Ki Pro recorder, which we call Ki Pro Classic now, was built to mimic a VTR. Anybody who knew a VTR was comfortable with this, but instead of recording on tape it recorded files on hard-drive modules. Those files were then immediately editable on a Final Cut machine. That was our innovation, by the way. It had never been done before. Now everybody does it.
TVT: You mentioned how you thought at the time that SDI was going to change everything. What are your thoughts about IP transport as an alternative to SDI for media production?
JA: Number one, I think it is inevitable because IP just takes over everything eventually. It’s even taking over asynchronous mode on backhaul. It’s starting to replace SONET on backhaul.
Just the ubiquity of IP means that eventually it’s going to take over everything because of cost. It will be the cheapest way to do it. Not now, but it will be.
The big question is timeframe. There are a lot of people out there who are whole hog transitioning to IP as we speak. And there is another camp that says, “Let’s just see how this plays out.”
If I were building a truck today, I don’t think I would go with IP. I think I would stick with SDI, myself, just because it is a known quantity, it’s safer, and there is not a real cost advantage to IP yet.
We have been involved with IP from the beginning. We are a member of ASPEN and AIMS and all of the plugfests. We are selling IP products.
Hopefully there is eventually a ubiquitous standard like SDI that everybody uses. That’s not guaranteed. You look in the computer world, you’ve got HDMI and DisplayPort constantly battling it out. Of course, DisplayPort is more computer oriented and HDMI is more video oriented.
But SDI is going to be around for a good long while.
TVT: Interesting. I thought the industry was beginning to coalesce around SMPTE ST 2110 as an IP standard for uncompressed video.
JA: Yeah, 2110 looks like it has a shot at becoming the standard. However, there are people who don’t like it. There’s even a website called Stop2110.com.
You have NDI out there. Obviously at this point in time there is a lot more NDI out there than 2110, and customers love it. It’s more of a lower-end thing. It uses JPEG compression to get HD on GigE. It’s kind of plug and play. Maybe it will creep up into high-end broadcast video and post. Maybe not.
One of the issues that’s arisen for 2110 is the software — people say it’s what you get when a hardware engineer designs an IP protocol.
Not everyone is happy with it. However, it does seem things are headed that way.
TVT: One of the things that IP enables is virtualization of video processes done by standalone boxes. What are your thoughts about virtualization and where will it leave AJA and other companies that have made a business out of offering standalone boxes?
JA: I guess what you are alluding to is that specialty products will be less and less [important], and for a hardware company that’s definitely a concern of ours. However, the way I look at is there will always be specialized hardware at the edges for I/O and connectivity and transmission.
Hopefully, we will always be able to make products to fit that need.
TVT: What about broadcast 4K and 4K in general? And does 8K ever enter conversations?
JA: Well, we actually sell an 8K product. It’s not specifically 8K. We have a PCI card we sell to other manufacturers. The product is called the Corvid 88, so it’s a low-profile PCI card that has eight HD-SDI connections that can be either inputs or outputs.
So, to do 8K over 3Gig SDI, you need basically 16 spigots. So, two of our Corvid 88s are used in some 8K product out there.
That is all 3G; 8K on 12G is just four spigots. We can do that on one PCI card, and we will do that. NHK is very serious about 8K as are the digital signage guys.
When it comes to 4K, we introduced our first 4K product at NAB 2010. We had a KONA card, and we were playing back 4K at NAB.
We have all kinds of mini converters that will do 4K. We even introduced mini converters for 4K with 12Gig I/O on them at InfoComm.
The interesting thing about it — I think your average TV viewer doesn’t see a huge improvement over HD. Going from HD to 4K, you really have to point out where the differences are.
So it’s not experiencing a huge uptick, but it will because it is here and everything is going to be made in 4K.
TVT: Where do you think AJA will be in the next 25 years?
JA: It’s hard to predict. Technology changes so fast now that it’s really hard.
Our complex products are on a two to three year development cycle, and the problem with that is nobody really knows what will be going on three years from now. So, you have to take an educated guess and hope you are right.
Otherwise, you’ve just spent a bunch of resources creating a product you aren’t going to be able to sell. It is getting a little difficult to manage a hardware business now.
On top of that, you’ve got the competition coming on strong from Asia. We never had competitors before in China, but we do now.
We are just trying to hang in there and figure out what to make and keep doing the same thing.