You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Objects placed as children of the camera should look static on screen relative to camera motion (or at least that is my interpretation from reading #121). Instead, when a movement is sharp (e.g. tracking error jumps) the children objects can visually jump some amount, this being more noticeable the closest the object is to the camera. It also causes a small visual lag in general, even with smooth movements and no tracking errors. I'm using monado
My application is a quite unusual setup in where I implement a VR headset within the scene (using viewports very close to the cameras), so I can play with parallax, culling, brightness, etc, for behavioral optometry research. The visual jumps and lag are very amplified in my setup so I tried to work around it by updating the positions of the objects that render the viewports, in relation to the hmd global transform, but I haven't found a way to do it early enough and I get tearing.
Is there a way to ensure object positions are current relative to the arvrcamera, before rendering?
The text was updated successfully, but these errors were encountered:
Objects placed as children of the camera should look static on screen relative to camera motion (or at least that is my interpretation from reading #121). Instead, when a movement is sharp (e.g. tracking error jumps) the children objects can visually jump some amount, this being more noticeable the closest the object is to the camera. It also causes a small visual lag in general, even with smooth movements and no tracking errors. I'm using monado
My application is a quite unusual setup in where I implement a VR headset within the scene (using viewports very close to the cameras), so I can play with parallax, culling, brightness, etc, for behavioral optometry research. The visual jumps and lag are very amplified in my setup so I tried to work around it by updating the positions of the objects that render the viewports, in relation to the hmd global transform, but I haven't found a way to do it early enough and I get tearing.
Is there a way to ensure object positions are current relative to the arvrcamera, before rendering?
The text was updated successfully, but these errors were encountered: