Guide to the World of 360 Panoramic Virtual Reality – VR360
This post isn’t reviewing cameras, brands etc.It tries to get you understand better the resolutions and the way of image projection of VR360 videos / images.There is no deep tech language used so that it can reach wider audiences who are interested in this subject.
The way of VR360 images are projected is more complicated a regular camera and because there are more and more cameras out there shooting VR360, which is causing the customers to be lost in this World, this guide can help you to understand VR360 image projection.
Video/Film Resolutions used by these cameras
Full High Definition (HD) – 1920 pixel width x 960 pixel height
Ultra High Definition (UHD) 4K – 3840×1920
Cinema 4K – 4096×2048
VR 360 Image Projection Types
Equirectangular – this is the commonly used final image projection type – VR headsets and video playerslikeYoutube use this format
Cube Map – great for post-production works such as object removal,compositing etc. before it’s converted to Equirectangular
Sphere Map – it’s mostly used in 3D softwares where VR360 render is not option and by smartly using a shiny sphere you can get the same result as a VR360 image capture
Single Capture (one camera unit)
Multi-Camera Capturing Units, such as the GoPro setup
How does it work?
To properly project a VR 360 equirectangular image, the device you use have to shoot everything around you.It requires full coverage of the scene and to optimally achieve this you need 6 cameras.It doesn’t have to be 6 cameras to fully cover the scene but this is the least number that can generate the most optimal result.Can be 8 or even 10 but it requires a lot more processing unit and more money and also can be 4 but in this case there is significant distortion in the final image.
Using total of 6 cameras, each camera :
– has a 90 degree Field Of View (FOV)
– has 1:1 aspect ratio, for example 1024×1024 pixels
4 cameras setup horizontally and 2 cameras vertically, top and bottom cams, (vertical setup is more complicated than this but basically top and bottom reflect the way it works) and this way you cover the whole scene.Fully cover the scene means 360 degrees FOV on the horizontal and 180 FOV degrees on the vertical axis.That is 4×90 degrees = 360 on the horizontal and 2×90 degrees = 180 on the vertical axis.
In Single Capture Cameras there isn’t 6 different cameras (although can be 6 sensors) but the build of the lens and the processor of the camera projects the image.Multi Camera Solutions using separate physical cameras.
As a result from the above, a 1920×960 equirectangular image is captured by 6pcs of 480×480 (cc.0,23MP) cameras.4 horizontal 480×4=1920 and 2 vertical 480×2=960
Altough it’s VR360 Full HD it’s a low-res image if you want to use the final result with a VR headset or simply moving around in the scene by mouse in a videoplayer, it can easily be disappointing.This resolution can be better used for apps with low-res videos such as phones and tablets or hobby purposes but not really more than that.
It means you need cc.12000×6000 resolution equirectangular projected image so that you can achieve a real 4K image quality in the final VR 360 scene watched by the VR headset or move around in the scene by your mouse.
Compression and Adding Metadata
Use .mp4 container/extension with h264 codec for the equirectangular video.For Youtube the metadata can only be added to this compression.A major technical limitation is that you cannot go higher than 4K (4096×2304) with this compression.
Single Capture Camera (SCC) Pros and Cons
– smaller and lighter than multi camera solutions
– also much cheaper
– a lot easier scene shooting setup
– somewhat easier post work setup
– lower capture resolution mostly in equirectangular FullHD 1920×960 or 1920×1080
– not all cameras SCC can fully cover the whole scene due to lens and/or processing power limitations
Altough SCCs looks like have more Pro’s than Con’s it doesn’t mean they are better.They can be better but it depends on what you use them for.If you don’t need higher resolution then SCC is the way to go but if you need more details i.e. for films, broadcasting etc. you should pick the Multi Camera System.
Basically the higher resolution you can do the better.Rendering in higher resolution is only the question of the processing power you have (and money to buy)…or a lot of time as you can render for example 12K on a slower computer but by the time it finishes 2 generations of CPUs might have just passed.
This is article isn’t sponsored by any individual or any company.There might be missing information here but overall hopefully you can find this post helpful and you’ve managed to understand VR360 better.Thanks for reading it.