[BUG] Particle Shadows Offset limitation.

Hello everyone.

I found a bug, on particle with Translucent Shadows.

Basically, the particle system has a limitation to be casted. In some tests, the size of particles, the fixed bounds of it, and the distance to ground (on my case are clouds with shadows). All this itens, compose some math, that limit the distance for a particle cast the shadows.

I cant cast shadows with any distance greater than 10000 units (z).
On 4.5, I can cast shadows without fixed bounds. But on 4.6 To cast a shadow I need to set the bounds, but this comes with this limitation.

Anyone can help?

Thank you. For all help.

But, now try to spin “Sun” position like 17 PM. The angle will produce another type of bounds and the shadows will go off. The depth will be more than 17k.
Also, put your Emitter at some 150000 (units z height), scale like 45000 (x and y) just like real clouds are. There is no bounds capable to handle it. Since any particle greater than 2000 units, at any distance from ground, simple loses your shadows.

Hey -

You should be able to go into your Particle Emitter’s General Details in Cascade and manual set the Bounds to a higher max Z value and get as much shadowing as your particle emitter will generate.

Thank You

Eric Ketchum

Hey -

I will track them down for you and upload them as soon as I find (or remake) them.

Eric Ketchum

Can you share this emitter with me? I want look about normals. (material/emitter). I see this answer, and im curious.

Here is the Project File with the Materials and Particle Systems. Remember these are GPU systems but the Volumetric Translucent Shadowing should remain the same for both CPU and GPU Particles.

In regards to the issues of cloud cover generating shadows, you really should be looking into using light functions instead of trying to generate actual shadows from your clouds as it will be more performance friendly and honestly easier to control in game though the setup can be a bit hectic to match in the beginning

Eric Ketchum

link text

Hi Eric. Thanks for help.

Here, i have examples, no matter what you setup, after ~15k distance. All Shadows gone. Especially, if light is angled.
But look, there is a relation Between shadow distance and particle size, the CPU particles (2 Right ones), has no fixed bounds.

[link text][2]

Hey -

I think there might be an issue here and I am digging into it a little further I will let you know as I continue to investigate. Your project was very helpful.

Thank You

Eric Ketchum

Any news Eric ?

Hey -

I just tested this with 4.7 Preview 8 and it looks to be fixed in 4.7. Remember though if you are using GPU particles the light will have to be at least Stationary in order to see the shadow and as always the bounds may not be an issue but the dynamic Cascading distance for shadows will have to be adjusted to met your needs.

Thank You

Eric Ketchum

Any Problem with Movable + Raycasted ?

Not that I can see, but as I said with great distances you will need to do some work with your cascaded shadow distances to ensure that the shadows are visible as you want them to be.