Getting to Grips with the Nokia OZO Camera
DESIGN
The first thing worth noting about the camera is it’s brilliant design, it really looks stunning, but it’s not just aesthetic, it’s built like it is for a reason. Firstly, it’s aluminium body acts as a giant heatsink, dissipating heat away from the camera and allowing it to keep running. A frequent problem of 360 cameras is that they overheat due to the cluster of processors that are typically back to back, this is a refreshing approach. How well it manages heat in hot climates though is yet to be seen. The second reason for it’s shape is to mimic the human head, this is a great approach to take – in 360 video you have to think of placing the camera as actually placing someone’s head anyway – but there are technical reasons too in the case of the OZO. The distance between the lenses is the same as the distance between human eyes, this gives a great representation of stereo when stitched.
ADVANTAGES
One of the main advantages of the OZO is the ability to live preview the camera’s feed while on set, this means directors, DOPs and clients can actually see a good representation of what they will get. This is something we have been either hacking together using live streaming software or using video villages with one of the cameras feeds in the past, the Nokia method is much slicker. As with everything in VR there is always a compromise and with this feature we’re tied to the old Oculus ‘DK2’ dev kit as the Nokia software is only currently for the Mac.
SOFTWARE
The software for the camera is very intuitive, if lacking more advanced features, but these are early days. Right now you can adjust the exposure by tweaking shutter speeds but not much else. ISO is fixed, as is aperture. The first piece of software, OZO Review, let’s you monitor live capture, adjust the camera etc, OZO Creator let’s you review, transcode and even stitch footage. The stitch quality we were unable to see as the renders are quite slow but we’ll be running tests on this in the coming weeks. It may be an OK solution for simple previews of footage but for us we will be feeding the footage in to the Foundry’s brilliant NUKE software to produce our stitches with greater control.
There’s a few achilles heels on the camera though, it’s no silver bullet, one of the main ones is the fact that the 3D effect only works in a certain portion of the image. Look behind you and it will be flat, also, looking directly behind you will be the worst quality part of the images as the area is only covered by the far edge of the lenses on the side of the camera – there is no camera looking directly back. I can see what Nokia were thinking here, they needed to place their battery and memory storage somewhere and for the vast majority of 360 video there is a forward view that is the most important.
BENEFITS
The 3D effect of the camera is good, however, there are significant complexities to getting true 3D across the whole front and side 240 or so degrees – this is due to the method Nokia have chosen of lens placements. Trying to replicate the human eye rather than computational 3D from multiple cameras like Google’s jump or Facebook’s 360 camera has it’s plusses and minuses. On the plus side, where the 3D works, it works brilliantly. However, where you move from one pair of lenses to another as you turn your head there can be areas where you see the convergence change – you lose presence. Interestingly, while with Nokia we saw a playback with perfect stereo that actually was not stitched. Still under wraps we understand but coming soon, it looks like Nokia have a way of playing back footage without it having to be stitched and dynamically showing you the best lenses for the view. We were pretty blown away by this – you could do your editing on individual lenses and then play it back in their player for a seamless result. Watch this space.
WHAT NEXT?
Nokia know this camera is not perfect, but then nothing is in VR, it’s the first step on the road to a robust professional camera and workflow and they have made a lot of headway. Nokia is fast updating the firmware and software of the camera and this set of hardware has a lot of space to develop. I know we will be adding this to our arsenal and relish the opportunity to work closely with Nokia in furthering the development of this brilliant piece of tech.
Comments
Responses
Derek White says
I'm really surprised by this review/article. Not once did you mention the elephant in the room which is the appalling picture quality. I'm guessing you didn't get a chance to take the camera for a spin, because if you did, I imagine this would be a very different story.
We bought a Nokia OZO and took delivery in March 2016. After taking it out 3 times, we thought there was something wrong with the camera because of severe noise in the black levels and a terrible picture quality. The material we shot was totally unusable and we actually got a better result from our 4-camera GoPro rigs (obviously not 3D). We contacted Nokia to find out what we were doing wrong and after numerous Skype calls with their marketing and technical teams, we discovered that this was in fact normal!
We simply couldn't believe it. A company with Nokia's brand value releasing a $60,000 camera which was unusable? Surely not. We then started searching for video online of projects shot on OZO and much to our surprise, there weren't any! I know there are a few now, and I urge anyone reading this to take a close look at them. Also, go on to Reddit and see the OZO groups, this is not an isolated case but rather the general consensus regarding the current version of the OZO.
After many hours on Skype with Nokia, we eventually sent the camera back and they refunded our money. Interestingly enough, they offered to refund us during our very first Skype call. In hindsight, that was a big red flag and evidence that they are well aware of the camera's shortcomings, but simply won't admit it. Even now. It's a shame if you ask me. I would happily have helped be an Alpha or Beta tester of their kit if they were just honest about it. It is (in their own words) the equivalent of the RED One when it first launched (except I believe it's more like a RED One Alpha).
Their workflow also leaves much to be desired. The Creator software is totally closed and although they promise to provide export functionality for various platforms, as it stands today, you can only export a DPX sequence from each of the 8 cameras if you want to stitch in 3rd party software. You also can't export to YouTube or Facebook and the 'live' 3 camera demo they had running at CES, was hacked specifically for CES and had a mountain of hardware doing the processing in the background (Nokia's own admission).
So you're pretty much locked into a closed system right now. I have no doubt this will change in time, but I can't help wonder how they are getting away with selling a low quality, noisy, slow and closed solution for $60k. I would love to hear from a production company anywhere in the world using an OZO for commercial projects and making money from it,
The OZO is indeed a 'groundbreaking' camera - or at least it would be if it actually worked (and we really, desperately wanted it to work), we simply couldn't wait and have instead ordered the parts to build a Facebook Surround 360 camera. It will end up costing roughly the same (even though Facebook claims $30k - it's way more), it also clumsy, heavy and has a ton of equipment that needs to follow it, but it shoots a hell of a 360 3D picture and it has a total open software and hardware platform.
If they're able to weather their own storm, I'm sure we'll buy the Nokia OZO v3.0, right now though it's just a great idea, but sadly, not a product.
Jackie says
couldnt agree with you more.
Henry Stuart says
We've actually had more time to play with the Ozo and I think it's a great bit of kit. As with other VR cameras though, it's no silver bullet, I think we have 5 different cameras now and they all have their time and place. The Ozo is particularly good for moving 3D 360 shots in tight spaces - the parallax is better than the Jump (with modified top and bottom cameras on the jump). It's also probably the best current solution for live streaming.
However, you have to really heavily light the scene to play to the cameras strengths. I recently saw a music video shot in Mexico by Here Be Dragons that used the Ozo and it was beautiful, the Ozo was on a zip wire, moving between rooms on a shoot.