opengl - why does 'far' contribute to near plain in frustum? -
in opengl, made viewing frustum using inverse matrix of
pxv p made glperspect , v made glulookat
at first, think 'far' factor of p doesn't affect vertices in near plain of frustum, affect.
then assume floating-point precision error, , there no 'far' factor in real formula. calculated prove assumption, failed. cannot remove terms contain (near + far) / (2 * near * far) form.
i think near plain of frustum must independent 'far' factor. cannot understand situation. did make mistake?
Comments
Post a Comment