Archived Old Bugs

[Micha_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

... could be a Vray engine bug: invisible light is visible at the reflection and produce noise.

left side show the GI prepass - right side show final pass


http://www.simulacrum.de/download/No...sibleLight.3dm

[Micha_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

... why is the refraction so different? Mistake or Bug?

www.simulacrum.de/download/StrongRefraction.3dm

[joe_b]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

I seem to remember invisible lights causing us problems before...As far as the refraction I haven't gotten too in-depth with it, but perhaps its a geometry issue?

[dalomar_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

Micha. No bug. The glass on the right and left are modeled as solids, and the two in the middle are two planes that are back to back each with the normals facing out. Close the middle two and it will be fine.

[Micha_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

Should back-to-back and solid not cause the same look? What is the difference? The normal dirs are the same.

[Micha_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

Hmm, I render a notebook screen with glossieness 0.9 and subdivs 44 but I don't get a noisefree reflection of the arealight (adaptive amount 0.85 aand noise 0.005). Could it be that we have a general noisy reflection of area lights problem? My only chance to get a smooth reflection is to use the universal setup.

[dalomar_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

Should back-to-back and solid not cause the same look? What is the difference? The normal dirs are the same.

No it won't produce the same result. Each of those planes are considered as an individual material/object. Since refraction needs that second surface to actually produce the second bending of light in doesn't get that with those two surfaces being separate. Once the volume is enclosed (or at least joined to the point where those two surfaces are connected) it will read the backface of that second surface as the exit surface and bend the light per the refraction IOR. With out them being joined the backface of the second surface is not regarded as the exit surface and is disregarded.

[Micha_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

So we can not render a hollow glas sphere? And for architecture renderings I need 6 surfaces instead 2 for each window? :-\

[dalomar_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

No it has nothing to do with something being hollow. All objects are hollow in Rhino because all Rhino does is make the outside surfaces of a solid, not the solid inside. The issue is that the entrance and exit surface must be joined, or in other words they must be part of the same object. As for your question of 6 surfaces instead of 2 the answer is yes and no. Using 6 surfaces joined together to make a box is the best way (I use either offsetsrf then solid or extrudesrf), but as long as the front and back surface are joined together by a surface on one of the sides (3 surfaces total) then it will be considered one object and Vray will read it as an entrance and exit surface. That being said the second option is likely to cause a few issues when applying fog to the material. You may only see it 1 in 50 times, but it will cause an issue. Also having a closed solid will automatically flip all of your faces outward. If you don't have it closed, then it is possible to have faces flip in the wrong direction, even though it is joined to other surfaces.

[Micha_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

By hollow glas sphere I mean a sphere with a second sphere inside. The seconds sphere is a mesh sphere with inverted normals. Seems to be inpossible per standard use.

But ... I found a trick for my two-surface-window-glass and hollow-glas-sphere: create meshes from NURBS objects and join it. Not conected meshes can be joined and rendered like expected.

With out them being joined the backface of the second surface is not regarded as the exit surface and is disregarded.

What are the rules for this seconds surface? Is it ignored? For what is it good to use?

[PhiD]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

All of the high GI multiplier values are because the physical camera is enabled by default. . .

Thanks dalomar, I haven't yet played around with the physical camera, I guess I didn't expect it to be on by default. It all makes a lot more sense now.

Have briefly browsed through the manual - looks good - but haven't had the time to fully take it all in.

Thanks again

Paul

[dalomar_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

By hollow glas sphere I mean a sphere with a second sphere inside. The seconds sphere is a mesh sphere with inverted normals. Seems to be inpossible per standard use.

Nurbs wise it is impossible, but if you build it via meshes then it should work. Thats one of the reasons why the material test scene is a mesh file. The bubbles that are in the glass need to be considered all one object. A basic rule of thumb is that if you select the object and it applies the material to all of the things you want it to (bubbles, holes, etc) then Vray will read it as one object.

What are the rules for this seconds surface? Is it ignored? For what is it good to use?

With typical materials the backfaces of an object are considered the same as front faces, but with the refraction materials the backfaces need to be read as the exit surface in order to have the correct refraction. The thing is that it will only be read as an exit surface if the rays are traced from the same object. With out that I think it just disregards the backfaces. That would explain why the structure behind those center pieces of glass look like they are in the exact place they would be if the glass was not there.

[Micha_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

By hollow glas sphere I mean a sphere with a second sphere inside. The seconds sphere is a mesh sphere with inverted normals. Seems to be inpossible per standard use.

Nurbs wise it is impossible, but if you build it via meshes then it should work. Thats one of the reasons why the material test scene is a mesh file. The bubbles that are in the glass need to be considered all one object. A basic rule of thumb is that if you select the object and it applies the material to all of the things you want it to (bubbles, holes, etc) then Vray will read it as one object.

Hmm, but if I assign the glass material to the two glass door surfaces, than I get the wrong refraction look. So, the same material seems to be not enough to get .... . I don't understand, how the bubbles in glass works without join the meshes. Isn't it not the same like my independent glass door surfaces? Whats the difference?

[dalomar_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

Its all in how you join and build the geometry. Nurbs won't work for separate surfaces because of how nurbs are. Even with meshes they need to be joined and have the normals facing the correct direction.

Here's an example...You wanted to know if you could do a hollow sphere and here's how. In the screen shot you will see 3 hollow spheres. The first is two nurbs spheres, which because they are closed surfaces must have their normals facing outwards. The second is a two mesh spheres, but as you see the inner sphere still has its normals facing outwards. The thrid sphere has the normals of the inner sphere pointing inwards. The last two examples have each of the two mesh spheres joined.



Here is the rendered example



If you want you can download the file here
http://www.box.net/shared/xez307mtr9

[Micha_cg]

Re: **** PLEASE POST ALL VFR4 PROBLEMS HERE *****

Thank you Damien for all this interesting infos. ;D