if (TexC.z smaller 0) {col=0;} didn’t do the trick. not sure, if it can, since the walls backside is a frontside in 3d-means (guess the tex coords are all within normal range).
if you use culling-counterclockwise on the fx, the texture is only seen one sided (in this case inside the wall).
but maybe it’s enough for your needs to just use a 3d-model without volumina.
i’m sorry to say that the volumina-culling-double projection thing is not what bothers me.
sorry for describing my problem mistakable.
how may that be solved?
solution of this would be a quantum leap in simulation of vvvv-dmx-controled lights.
i think there are some more users beyond me having interest in this.
unfortunately i don’t think that i am able to find the solution on my own within the next 6 months…
The reverse projection occurs when the {q} coordinate becomes negative. It is not easy to act on the {q} coordinate at the fixed functions level, but in a pixel shader it is really simple as shown by the following shader code
yes, i did it. and it provides somehow the possibilty to place a slide independent from a light which shines through the slide.
however for most gobo or beamer projections you wouldn’t need this flexibility. to say it in antoher way: the parameterization of this patch didn’t work for me anymore. i needed a good way to visualize a projector (done with the now very good projector module) and an easy way to use this virtual projector in a simulation.
so i skipped this slide projector patch…
by the way: there is a new version of that projector patch…
It would be nice to update the girlpower with the most current version of the Projector.fx - I just ran into this same old problem with the example from 40beta17 and luckily remembered this thread …
Also modified the .fx to kill all color outside of the [0…1 texture coordinates to fix ugly borders for blendmodes other than ‘add’: