Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Computing screen coordinates
1 view
Skip to first unread message
listechtony
Dec 15, 2009, 5:08:01 PM12/15/09
to
I have a vertex buffer of world coordinates which when I apply the correct
transformation matrices render correctly to the screen, as expected. Now I
wish to compute screen coordinates for each vertex to use to populate another
vertex buffer of transformed coordinates because I have some symbols I wish
to render in screen space, depending on when the vertices lie. I use the
transformation matrices to do this manually and normally this also works OK.
Yet, when I get "extreme" transformations, say when I zoom in very close and
some of the matrices have very small numbers, e.g. below 1 e-5, then the
symbols do not match exactly with the vertices. I am confused by this because
I would have thought that the rendering pipeline is using exactly the same
matices as I am using manually, yet there does seem to be some rounding
issues here. How does the rendering pipeline use the matices to compute
screen coordinates and why would I be getting slight differences at extreme
cases, where normally I seem to be getting the same coords by applying these
transformations manually?
transformation matrices render correctly to the screen, as expected. Now I
wish to compute screen coordinates for each vertex to use to populate another
vertex buffer of transformed coordinates because I have some symbols I wish
to render in screen space, depending on when the vertices lie. I use the
transformation matrices to do this manually and normally this also works OK.
Yet, when I get "extreme" transformations, say when I zoom in very close and
some of the matrices have very small numbers, e.g. below 1 e-5, then the
symbols do not match exactly with the vertices. I am confused by this because
I would have thought that the rendering pipeline is using exactly the same
matices as I am using manually, yet there does seem to be some rounding
issues here. How does the rendering pipeline use the matices to compute
screen coordinates and why would I be getting slight differences at extreme
cases, where normally I seem to be getting the same coords by applying these
transformations manually?
listechtony
Jan 11, 2010, 4:36:01 PM1/11/10
to
Can anyone help here?
0 new messages