This is rather a tip than a complete tutorial. It also contains a pre rendered VR Testcard at the bottom.
There are some great possibilities of using your VR Headset to view 3d surround content even movies.
One wonderful 3d surround video player is vrPlayer ( https://play.google.com/store/apps/details?id=com.stephanelx.vrplayer , http://www.vimersiv.com/ ).
However, if you render two spheres equirectangular with blender and combine them in the compositor to view them in VR Player, that works well when you look forward, but you loose 3D information when you turn to the side or get it wrong when turning backwards. My complete post describes that in further detail regarding the result and how to use the VR testcard:
Now to the point how you can do it in blender to produce even perfect 3D animations for your VR headset:
You have to render just a small region of let’s say 1 to 3 pixel columns for each eye looking forward. Then you turn the cameras and render the next part. When you stick everything together you really get a perfect VR image that works also when you turn your head to the side or look back.
But how to do it?
you can define a render region with shift+b in the viewport and check “border” and"crop" in the dimensions section of the render properties.
You can do this programmtically with some python code snippets (using a full HD render size of 1920x1080):
use border and crop
bpy.context.scene.render.use_border = True bpy.context.scene.render.use_crop_to_border = True # set border for this tile bpy.context.scene.render.border_min_x = .500 bpy.context.scene.render.border_max_x = .5016 bpy.context.scene.render.border_min_y = 0 bpy.context.scene.render.border_max_y = 1
I just managed to get results with 3 pixels, not with one pixel because when I reduced it further it just rendered out transparently without contents.
If you have an idea how to create a one-pixel-wide column rendering I would be happy to read about it.
But this 3-pixel-result was enough for the VR test card which I released to public domain. So feel free to use it as you wish.
My creation process was quite creative and a bit quick and dirty. That’s why I didn’t just post the blend file. To get it ready for production rendering I would have to invest some more time also rethinking how to render all the parts with animations (I just used an animation to create the rotating camera which is not possible if you have an animation in the movie). But when I think a bit further this should be also combinable with real movie footage recorded with e.g. Panocam 3D (link included in my post). When combining it with Sebastian Koenig’s Track Match Blend methods ( https://cloud.blender.org/training/track-match-blend-2/ ), we can produce real-real-3d VR movies combined with perfect animated content in Blender. This is a really exciting era we live in
Oh, the file ending .jpeg is intentionally because VR player needs it to use the image.