I’m currently building out ofxCardboard and need to know how I should structure the addon so it works on iOS and Android. I’ve been working on the Android side of things and found an clean iOS port someone else made. The iOS port needs to get into OF land but that should be easy.
I was thinking the other day that it would be nice to align all of the VR addons so they work very similarly to each other.
I think the thing that would unify all of these is to have a general model of how to set up the VR camera in the scene. It’d be like an ofxVR abstract class that ofxCardboard, ofxOculus, ofxVRWhatever would extend.
The unifying things to me are:
- attach to a base OF camera to set the basic view port (ofxOculusRift does this by the ofCamera* baseCamera ref)
- camera projection abstraction (what are you looking at, activating things through gaze)
- basic abstraction for field of view, frame buffer resolution, etc
If we had two or three good VR addons it would make sense to start making an abstraction.
I will say this sounds a lot like the idea we had for ofx3DCamera to abstract depth cameras, and so far the hardware moved so quickly and was diverse enough that the abstraction has had trouble catching on.
word. Yeah I it is an ambitious goal to have some kind of unifying base with how quickly the devices change. I don’t see Cardboard changing much haha considering how low tech the device is compared to everything else. I do see Oculus and GearVR being moving targets. They’ve already released a new version of the GearVR
Lets work/hack on something to try and unify everything.
(side note) what’s up with the slide in video ads on this page? cc: @arturo
im just looking to get this running on iOS.
have looked at your ofxCardboard addon and looks like you’ve done quite a lot of work for Android.
seems like porting it to iOS should be pretty straight forward?
just wanted to hear your opinion on it first before getting stuck in…
Hey @obviousjim, was there any work toward ofx3DCamera I could take a look at? I’m starting to work on something for KinectV2 that packages up things like live vs playback, recording, directly connected vs streaming a remote image (via websockets or zmq), etc. I was thinking it’d be nice if this was uncoupled from the camera hardware as well, so I’m curious what you were thinking for the cam abstraction! Cheers