How Nokia broke into virtual reality with its Ozo camera

Above: Ozo at sunset

Image Credit: Nokia

VB: How much does each one cost?

Voltolina: It’s $45,000. The development of components that didn’t already exist is what pushes up the price. Cameras, since day one, have been thought of as one lens and one sensor. All the components related to the electronics around it assume you’ll have very high resolution, but only one sensor. When you combine eight together, the SOC that has to coordinate all the sensors — that component didn’t exist. We were forced to create an FPGA to do something that hadn’t been done before. You’re synchronizing all those 2K by 2K sensors at 30 frames per second. The data rate is significant. There were no components that can encode, in real time, all eight streams in one SOC that’s produced at affordable volumes.

Also, the sensor itself — we use a square sensor, which is the best geometry for capturing the fisheye lens. Most sensors are rectangular. That leaves a lot of sensor cells that won’t be used. We also needed all the images to be fully synchronized. I don’t know if you’re familiar with a rolling shutter versus a global shutter, but if you have multiple rolling shutters, the exposure to light is never 100 percent synchronized for all of them. It creates eight images that all have slight differences, and when you stitch them together they don’t match. We had to introduce a global shutter, which has a smaller market and costs more.

The lenses are custom made, because the geometry of the camera didn’t exist before. All the components are pretty much designed for this new purpose. The fact that we don’t already have millions of cameras on the market using them makes those components more expensive.

VB: If it comes out at that price, what kind of users are attracted to it at this point?

Voltolina: We started selling the Ozo in February of this year, in North America. At this point we’ve reached most of the rest of the world, including Europe and China. Our main customers are studios, big and small, that are already producing VR content. People who’ve already tried to make professional VR content or 360 video. When they see the Ozo, they understand the benefit immediately. It’s more expensive, but the savings in time during production — particularly in the stitching and post-production — is so significant that it pays off rapidly.

You can imagine that when you set up with actors on location, you’re spending the most money per hour that you ever will during a video production. If you have one camera where you can see what’s happening and coordinate the actors accordingly — you move here, you move there — and then you have to shoot with another camera, you can’t review what’s happening until later. If you have to go back and shoot again that’s a huge amount of money.

Currently VR experiences are mainly additional marketing for movies, things like that. They’re not movies in their own right. It’s the VR experience of Pete’s Dragon or Jungle Book, features like that. Or a lot of commercials, of course. Live events are big. We’re seeing people experiment with live events almost every week. We just finished up in China for the Strawberry Festival, a music festival, streaming that to several countries.

Above: Ozo

Image Credit: Nokia

VB: What’s proving the most popular so far? Is it the live streams, or produced recorded content?

Voltolina: It always depends on who’s in the movie. [laughs] It’s not so much about whether it’s live or not. It’s about who’s the star. There was a movie opening for Disney that we streamed, Alice Through the Looking Glass. Pink, the singer, did a concert live for the premiere. Of course that drove a lot of viewers. But we also did a music video with OneRepublic for a new single called “Kids.” They released a 2D music video and then a VR experience. That wasn’t live, but on social media all the fans were really intrigued by the fact that they could watch the video, and then see it from a different direction in VR. Fans would watch it again and again as they discovered new things in the experience. The flow, the story of the whole thing is much more than just the band gathered around the camera.

Another one that was very popular was Pete’s Dragon. In the VR experience you really get to fly on a dragon. You can look around at the wings and the tail. The video itself is like an airplane ride. You’re flying over New Zealand. But the fact that you’re on top of a dragon — for a lot of fans that was a big attraction. It’s always a combination of subject and story. And of course, it helps if you have a major star.

VB: What’s the next step for you? Do you have a road map going forward?

Voltolina: The next steps go in two directions. One is toward a more complete solution. If we’re capturing more and more data, we need to efficiently carry that data to viewing devices, which are getting better and better. Last year was the year of Cardboard. We started seeing the first Oculus and HTC headsets this year. Now we have PlayStation VR, Daydream. Many more devices will be released with higher resolution and better performance. The highest level of immersion keeps going up.

Also, the number of people who are at least familiar with 2D 360 video is going up. That gives more of an incentive to go for immersive VR. The technology to enable Ozo Live or Ozo Player for better immersive playback, that’s currently our next step.

VB: So that’s increasing resolution?

Voltolina: Definitely resolution, but it’s quality in general. Resolution is always mentioned because it’s an easy way to describe better quality, but at some point resolution gets to where you can’t even distinguish. Better visual quality in general is coming, including the quality of stitching. That’s already improved tremendously. With the camera, you get the Ozo Creator software, which does 3D stitching. We’ve released three new versions with significant improvements each time.

Another area is production with multiple Ozos for live streaming. We’ll support a way of producing a VR experience that doesn’t just use one camera. It’ll incorporate commentary, different locations, and so on.

VB: Will you be able to bring it down in price?

Voltolina: We started at $60,000 when we first announced it. By summertime we adjusted to $45,000. The reason is, the first few months we were producing the earliest units and we weren’t sure we could scale manufacturing up to serve the whole world. We started with one region, North America, to see if the product would be well-received, and then if we could scale manufacturing. That happened in August. That’s how we were able to bring the price down. Since it’s professional equipment, a lot of rental houses are carrying it now, too, just like high-end cameras from Sony or Panasonic.