Hello,
I am trying to use OpenVidu with ARCore, to achieve a video call whilst one of the phones would still be able to use the Augmented Reality portion while communicating in the video call.
I think there’s a general problem with cameras in which there can’t be two different components trying to use the camera at the same time. ARCore is relatively closed in what it does: an update() function that you need to call once per frame that accesses the camera and recalculates the position of the phone and the virtual objects, I was looking for the possibility of providing the frame from the camera manually instead of using the classes like VideoTrack, VideoCapturer and VideoSource. Is there any way to do this? Would it be somewhat straightforward? Would it involve just implementing / extending any interfaces or classes that already exist in webrtc / openvidu?
Thank you in advace.
(Notice: I accidentally deleted the post I had made before so I am again reposting it, sorry)