ozo-vr-camera

We get asked for the Nokia Ozo alot. Its an expensive machine with a large number of shortcomings so its time we made our opinion public. Note this is an opinion, not necessarily shared by the whole team here but this should be used as a guide to potential new users to the 360° filming industry.


Firstly, just because of the price, this system is not a silver bullet to high quality 360° video. Quite the opposite in fact. If you want to use this system, despite its many shortcomings, you will need to have a post team in place, or access to one with licensed software built by Nokia directly for Ozo users.

Due to its high cost set by Nokia, people are mostly assuming it is the go to rig – it is not and should not be. Freedom360 rigs are the standard systems for a reason right now, the main shortcoming that F360 doesn’t offer stereo. The combine GoPro quality is higher than what a Nokia Ozo outputs. The Ozo uses 8 2k x 2k sensors which appear to be optimised for mobile applications. The colour science is bland and compression is extremely heavy – you’re going to end up with super soft shots with blocky compression artifacts everywhere.

Noise is another problem, consistent with most cameras for low light applications in 360°, but the Ozo is next to useless in the dark, and shadows and even highlights are often blown out. There is basically hardly any lattitude.

Here follows an example of some Ozo output – look for yourself at how the background rocks look all blocky and next to unwatchable. This is Nokia shared footage, you would think they know how to get something good out of their own camera? Apparently not.

 

Data is absolutely massive. You are looking at 10hours of post rendering for a 1minute clip (10-30 seconds per frame). That is just not practical for any kind of shoot these days. In 2 years this will come down, but it 2 years we will have systems far exceeding what the Ozo does from other manufacturers. They suggest an offline edit using footage from two lenses which don’t stitch together due to their parallax distance – you’re going to be editing with incomplete material. You’ll want to have lots of media standing by on set as you’ll be recording 1.5GB per second.

What does it do right? Well the stereo is reasonable in the front 240+ degrees horizontally. Its flat going towards the rear and there is an overtly obvious autostitch line sitting there.

The spatial audio microphone setup is market leading, simply by virtue no other cameras exist which have spatial audio built into them. Spatial audio can be achieved with a separate mic like the TetraMic or Ambeo. The live stream is a reasonable experiment in simplification which is going to be available in Q3 2016, but the starting images are so poor one questions if it is worth the effort over 6-cam gopro streaming.

And to cost… You are looking at for a full setup at least £55,000 (£42,000 to early adopters) or £3,000+per day hire for the kit alone, plus technicians dedicated on set and in further cost in post. For around 1/10th the price, you can get higher quality elsewhere.

On Nokia’s own Ozo forum, you will find many current customers/owners struggling with the substandard image qualities.

https://support.ozo.nokia.com/s/question/0D558000004uGTwCAM

And here follows an independent review of another Nokia Ozo user with similar feelings.

https://www.reddit.com/r/oculus/comments/4kwjz4/just_wrapped_a_shoot_with_the_nokia_ozo_noisy/


The truth is this system is a prototype. They got the design right, but the lenses and sensors are simply under par and they need to start work on a tethered system which will reduce the rendering time. 1/3rd of the price would have been a better starting point, but the Ozo is going to lose value extremely quickly. The price for an incomplete beta system is high these days.

Ozo reference manual for post workflow:

https://dl.dropboxusercontent.com/u/6364348/OzoPostWorkflow-2016-05-02.pdf