THE POST BELOW IS MORE THAN 5 YEARS OLD. RELATED SUPPORT INFORMATION MIGHT BE OUTDATED OR DEPRECATED
On 30/10/2006 at 04:30, xxxxxxxx wrote:
Cinema 4D Version: 9.6
Platform: Windows ; Mac ; Mac OSX ;
Language(s) : C++ ;
I would like to know the difference between a volumetric shader and a volumetric videopost plugin.
What is the advantage of a videopost implementation and what are the cons?
On 30/10/2006 at 07:38, xxxxxxxx wrote:
I would say the main advantage is that you can actually attach volume shaders to objects, thus easily define where the volumetric calclations happen. I think you can do this with a videopost plugin too, but it seems easier with shaders. With volumentric shaders you get also some additional functions for instance Displace which you would have to recreate in a videopost plugin.
On 30/10/2006 at 08:24, xxxxxxxx wrote:
thanks first of all. What I actually meant was if there are any pros and cons between the two concerning the rendering pipeline. Are there any advantages that a shader is calculated in post instead of in a volume shader?
The reason I am asking is that when I use TraceGeometry in an environment shader, I am checking for transparencies with GetSurfaceData() and if there is full transparency (for example on a plane with an alpha texture) I simply keep on tracing and check the next location. Now if this hit is fully opaque, I set the transparency and color at that point.
Works fine but the problem is that this seems to affect the full transparency of the former hit point (so the alpha textured plane). How can I get rid of this behavior, or rather, what is the reason for this behavior?
Here is some pseudocode:
pStart = pHit;
pStart = pHit;
Check this for an example to see the effect I mean:
(there is a slight white border of the plane visible although I am setting vd->trans = 1.0 when there is full transparency)
Hope this is clearer. Did anybody encounter something like this? It´s really important.
On 08/11/2006 at 13:38, xxxxxxxx wrote:
I don't know if I'm stupid, but I don't understand the attached picture. What should it look like to be fine? And what is it that we're looking at?
On 08/11/2006 at 15:44, xxxxxxxx wrote:
you see a plane with a mapped circular gradient on it (from inner white to outer black. the gradient shader is in the alpha channel).
Usually the inner part (as you can see in the picture) is visible and fades out radially. That works of course.
Now I have a volumetric environment shader that works as described in the pseudocode.
I hit the plane, then I check the surface transparency with GetSurfaceData() and if it´s fully opaque, I do my calculations and return. If it´s transparent, I do the calculations and keep on tracing forward. if it´s fully transparent (sd->trans==0.f) then I don´t do any calculations and keep on tracing forward.
Now, the edges of the plane in the image are fully transparent (sd->trans==0.f) and therefore they should be invisible. But when you look closely, the edges shine a little white (you can identify them in front of the background).
My environment shader is based on distance of the intersection point to the eye. When there is no background, this effect doesn´t come up. Only when I keep on tracing and hit another object this happens.
I first thought, it´s a bug in my code, but when I do the same with the environment object´s fog, the same effect happens.
Find an example scene of the fog problem here:
How can I solve this problem?
On 09/11/2006 at 15:38, xxxxxxxx wrote:
ah, and that´s why I asked the question if there is a difference between VP and Shader implementation. Maybe this effect does not appear in a volumetric VideoPost implementation?
On 15/11/2006 at 12:05, xxxxxxxx wrote:
So, is there a fix now possible or not? I really don´t wanna be impolite, I just think after 2 weeks it should be possible to answer this question or at least give me a "we don´t know" or so. I sore inbetween nothing and less.
On 16/11/2006 at 00:51, xxxxxxxx wrote:
Let me just say at least that I've looked at the fog file and can confirm that there is a visible border. It seems to me that there is less fog is the fog goes in two segments (cam -> plane -> floor) than one (cam -> floor). That makes me suspicious that perhaps something is not linear when adding the transparencies up, so that .1 + .1 < .2. It's even clearer when you have multiple planes in front of each other. I don't have any answer, but I'll have to think about this more tonight!
On 24/11/2006 at 01:34, xxxxxxxx wrote:
So, have you thought about it? I NEED an answer.