aalavandhaann
2010-11-11 20:33:55 UTC
Dear Pv3D gurus,
All this while i have been trying to make a 2d editor which will help me to
draw maps by just drawing lines around in the 2d screen for the walls and
other library items which has an equivalent 3d representation for a FPS
based game. Then i take those line starting and end point values and
construct them as 3D walls using rectangles/planes by deciphering the x,y,z
and rotational values mathematically( just by applying 2d y value as z value
in 3d. The height of the wall is arbitrarily chosen. The rotation value is
calculated using Math.atan2(lineend.y - linestart.y,lineend.x -
linestart.x)) based on the x,y values from the map created using 2d. All i
see is that the 2d points are successfully converted and i see the same
orientation and shape correctly in 3D which is reproduced as rooms using
Rectangle and Plane class of papervision3D.
But somehow i am not happy with the results coz the perspective(focus and
zoom) seems to change based on the arbitrary height values(i guess so) i
give while reconstructing the walls from the 2d line values. Where as if i
use a 3D modeling program like blender or sketchup to draw the 3D directly
then the models are rendered perfectly as visualized and created without any
problems.
My question here is if we manually add Rectangles and Planes (the primitive
objects) to construct a room does not give the feel but where as the model
of a room designed in a 3d program is beautifully rendered. Is there any
thing that i am missing here? are there any other special properties that
comes with a collada that fixes the way our pv3d camera renders the display
object?
Please help and your suggestions and advices would be appreciated.
Regards,
Ashok Srinivasan.
All this while i have been trying to make a 2d editor which will help me to
draw maps by just drawing lines around in the 2d screen for the walls and
other library items which has an equivalent 3d representation for a FPS
based game. Then i take those line starting and end point values and
construct them as 3D walls using rectangles/planes by deciphering the x,y,z
and rotational values mathematically( just by applying 2d y value as z value
in 3d. The height of the wall is arbitrarily chosen. The rotation value is
calculated using Math.atan2(lineend.y - linestart.y,lineend.x -
linestart.x)) based on the x,y values from the map created using 2d. All i
see is that the 2d points are successfully converted and i see the same
orientation and shape correctly in 3D which is reproduced as rooms using
Rectangle and Plane class of papervision3D.
But somehow i am not happy with the results coz the perspective(focus and
zoom) seems to change based on the arbitrary height values(i guess so) i
give while reconstructing the walls from the 2d line values. Where as if i
use a 3D modeling program like blender or sketchup to draw the 3D directly
then the models are rendered perfectly as visualized and created without any
problems.
My question here is if we manually add Rectangles and Planes (the primitive
objects) to construct a room does not give the feel but where as the model
of a room designed in a 3d program is beautifully rendered. Is there any
thing that i am missing here? are there any other special properties that
comes with a collada that fixes the way our pv3d camera renders the display
object?
Please help and your suggestions and advices would be appreciated.
Regards,
Ashok Srinivasan.
--
View this message in context: http://papervision3d.758870.n4.nabble.com/Advice-needed-thanks-in-advance-tp3038513p3038513.html
Sent from the Papervision3D mailing list archive at Nabble.com.
View this message in context: http://papervision3d.758870.n4.nabble.com/Advice-needed-thanks-in-advance-tp3038513p3038513.html
Sent from the Papervision3D mailing list archive at Nabble.com.
