Hello,

I would like to know how to render particles in Mantra the same as it's done in opengl. I would like my points to keep the same size on screen independently of their distance to the camera.
Of course, I can render with an opengl ROP but there is some polygonal geometry added to the particles I would like to shade. And I don't want to dive in glsl shading dev if possible …

Thanks for your help !

Vincent

Would rendering them with an orthographic camera work?

you can them in into a vopsop along with the origin of the camera, and subtract their position from the origin of the camera, and then scale their pscale accordingly

I am using perspective camera.

I will adopt zachary suggestion I guess.
I have now to study how is organised the fbx camera subnet …

Hey !
It makes me thinks to one question :
if the camera is parented to nulls. Let say one null for position and one for rotation. How can get, in sop context, the camera position ( world space) ?

thanks

vbk!!!
if the camera is parented to nulls. Let say one null for position and one for rotation. How can get, in sop context, the camera position ( world space) ?

Easiest way is with an Object Merge SOP. Merge in the camOrigin SOP from inside the camera and make sure to set Transform to “Into This Object”. This will fetch the point and crack all the transforms to ensure the point is in true world space position.

If you want orientation, you can always append a point SOP after the camOrigin null inside the camera. Set the Point SOP's Normals to 0,0,1 and set UP to 0,1,0. The Object merge will fetch the point and properly transform those vectors set in the camera so you can now copy stuff to that point and get it's orientation.



If you want to use expressions, look at the vtorigin() expressions but that would be slower than the Object Merge SOP.