Fprime does a difference in rendering....


#1

Look at the difference between LW & Fprime render:

LW:

Fprime:

All setting is constant, one area light and one fill light but the default LW render gots artifacts (and of course it is slow).


#2

What are you talking about?

There is a little bit of darkness at the bottom in the LW render, but if you never saw the Fprime render, would you notice it?

In the hectic enviroment I work in, I wish I had time to be so meticulous. From an artistic standpoint, both of those images are exactly the same. That’s more important than the technical nitpicking.

and when I have a difference in Fprime and LW versions, it much more pronounced than that. Many times Fprime just does simply does not render things the way they are in LW, especially refraction. I can’t use it most of the time.

Mike


#3

I noticed about the black artifacts beside the button area…

Those artifacts appear with the following model structures:


Well, I just think FPrime handle this better…


#4

hey dickma by chance could you show a wire of the shirt?

Thats a wierd one for you as well …

could it be lighting …from my expierience lw and fprime read (render) lighting properties slighting different especially when is comes to radiosity


#5

Agreed :smiley:

Here what I found, with a scene of mines, how strange …

LW render :

FPrime render :

The funniest of this is that this scene was originally designed for a TNT2, where it displays in realtime at 30FPS (screenshot here). I used a lot of transparency planes to mimic volumetric fog, that’s what making FPrime so unhappy imho :stuck_out_tongue:

No matter, FPrime still is an amazing piece of code !

EDIT :
Dick Ma : your small sample shows non-manifold polygons, that should be forbidden in subpatch… maybe that’s what going wrong…
see http://www.cgtalk.com/archive/index.php/t-95063.html


#6

Dick Ma: I see those shadow(?) artifacts on both images, except they are not in the same places.
What did you use as a fill light and what parameters?

Krix


#7

there is nothing special with the wireframe. As I mention before about the modeling structure.

About the fill light it is a distance light only.

Dick Ma : your small sample shows non-manifold polygons, that should be forbidden in subpatch… maybe that’s what going wrong…
see http://www.cgtalk.com/archive/index.php/t-95063.html

About non-manifold polygons. I know Maya don’t accept it to do subpatch “or technically say no subdivision” But as lightwave accept it… well this is not the main problem.

P.S. in the lightwave Layout the improper geometry artifact also shown as similar as the LW render result. May be Lightwave renderer directly reads the data from layout while FPrime has its own manipulation of geometry data.


#8

There’s a worse artifact in the LW render, it’s a fan of black triangles right in the center near the shirt’s buttons.

And yes, artifacts DO matter in production, especially during animation, since I would expect that LW’s errors may flicker or pop as view changes or the shirt is deformed.
For a still, the lower darkening is ignorable, but not in an anim.


#9

I ran into a similar problem with transparency planes for fake volumetrics. Worley Lab tech support helped me out. The difference is that if you don’t have LW raytrace transparency on, you can see through any number of non-refractive planes.

FPrime ALWAYS traces, so it will follow LW’s ray recursion limit. So the LW scene will show all the transparency, and FPrime will STOP at the ray recursion limit. The fix was easy, I just increased ray recursion and my problem was immediately solved.

I bet that’s what you’re seeing too.


#10

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.