How vr camera transform gets applied

void UCameraComponent::GetCameraView(float DeltaTime, FMinimalViewInfo& DesiredView)

[5:52]

in there

[5:53]

force the scale to 1, or the inverse of your actor's scale

[5:53]

usually that case is handle through world to meters scale

[5:53]

are you modifying that at all?

[5:53]

because that would be the other path

[5:54]

modifying w2m with the scale of the actor

[5:54]

which is easier

[5:54]

that would scale the IPD

[5:54]

and the camera's motion

[5:54]

so if you're scaling up your actor 2x, scale the world to meters scale 2x as well

so is GetCameraView() where the actual viewmatrix gets calculated?

don't have the code open at the moment, but it grabs the view matrices twice

[6:07]

the head pose is used to update the cameracomponent in the function i pasted above, that gets the gross movement

[6:07]

and then CalculateStereoViewOffset gets the offsets per eye

[6:08]

and the view matrices are updated again in PreRenderViewFamily_RenderThread

[6:08]

and updated per eye in PreRenderView_RenderThread