Hehe…
If I set the shadow map to be say, 8192, I barely get any shadow at all. If I set it to 64, I get thick shadows.
I know why this is happening, but I don’t know what to do about it.
So… What should we do? I need sharp shadows, but don’t know how to get them.
Ok, let me put this in the wishlist pile… Shadow density. Just a multiplier based on the shadow density spinner that lets me make shadows more or less dense. Tint would be nice too.
What are the particles in this scene like? The idea is that higher resolution maps should make the result sharper without changing the average attenuation.
Box o’ particles. 10 million. With a 3D map for the color and opacity.
I’ve ended up with another situation where this is important. Shadow map size of 12 gives me deep rich shadows, but no crispness. 4096 makes no shadows visible. This is with nearest neighbor or bicubic.
But the biggest issue comes where the renders look VERY much different when the camera is far, and when near.
Attached is some samples with the shadow map set to 512.
I’ve ended up with another situation where this is important. Shadow map size of 12 gives me deep rich shadows, but no crispness. 4096 makes no shadows visible. This is with nearest neighbor or bicubic.
But the biggest issue comes where the renders look VERY much different when the camera is far, and when near.
Attached are some samples with the shadow map set to 512.
Let me run this by the developers for suggestions, but I'm not sure if there is an easy solution.
As you zoom in, the number of particles per rendered pixel is actually decreasing. The renderings appear to correctly reflect the resulting lower density. For smoke or dust clouds, this makes perfect sense, but when simulating a surface, it obviously isn't what you want. My first impulse would be to crank up the volume of particles.
Oh, right. I need to try this on a non-black background so I see the difference between the shadows and the background. Or look at the alpha.
No, it’s definately the shadows.
When far, the pixels are well lit and opaque the alpha. When close, the pixels are dark and opaque the alpha.
EDIT: Oh, and it would be cool if the background color swatch could reference the environment background color. Not a huge deal, but it confused me for a second.
When far, the pixels are well lit and opaque the alpha. When close, the
pixels are dark and opaque the alpha. <<
Got it. I’ll look into it here and get a simple case logged. (I’ve seen a
similar issue where moire patterns appear relative to the map size and
particle density.)
Oh, and it would be cool if the background color swatch could reference
the environment background color. Not a huge deal, but it confused me for a
second. <<
Agreed. I’ll log this.
I can think of a couple of different ways this could be implemented. Do you
have any preference on how this should work?
A checkbox to use environment background would be fine. But hey, it’s maxscript, so I suppose I could do that myself once the release came out.
A checkbox to use environment
background would be fine. But
hey, it’s maxscript, so I
suppose I could do that myself
once the release came out.
Turns out it is not as straight-forward as assumed. We will add the option to use the Max environment background in the actual plugin code.
Note that the swatch in the Scripted GUI allows for non-zero Alpha values to be specified. In the case of the Env. Background Color, we will always assume an Alpha of 0.0.
Cheers,
Borislav “Bobo” Petrov
Technical Director 3D VFX
Frantic Films Winnipeg
A checkbox to use environment
background would be fine. But
hey, it’s maxscript, so I
suppose I could do that myself
once the release came out.
Turns out it is not as
straight-forward as assumed.
We will add the option to use
the Max environment background
in the actual plugin code.
Just a note that support for Env. Background color is available in build 0.9.4 and is ON by default.
Cheers,
Borislav “Bobo” Petrov
Technical Director 3D VFX
Frantic Films Winnipeg
Oh, right. I need to try this on a non-black background so I see the
difference between the shadows and the background. Or look at the alpha. <<
If they aren’t too big, zip up and post EXR files here. Otherwise, feel
free to email them to me.
I did a simple test with a “box o’ particles” to verify that nothing really odd was happening on my end. Spotlight, pflow box emitter, 250,000 particles, default krakatoa settings. As I move the camera in and out, the value of the lit faces of the box and the shadowed faces of the box change quite a bit. Also, changing the size of the shadow map drastically changes the shading, which goes back to my previous wish.