Directional Shadows

[Spectre01]

This is more of an artistic question regarding sector shadows interacting with elevation.

 

Let's say I have two cylindrical towers 256 units tall. For the sake of simplicity, the light source is coming from the west and the length of the shadow is the height of the object. So a 256 high cylinder casts a 256 unit long shadow.

Now, let's say I have a second cylinder with a pit behind it, 64 units in depth. Given the above scenario, would the shadow lengthen to 320 units? If the pit is bisected into sectors of different elevation, would the same shadow's length increase/decrease depending on the height of that sector in relation to the cylinder?

 

Edited February 25, 2020 by Spectre01

[Nine Inch Heels]

1 hour ago, Spectre01 said:

But following the same rule set, what would the accurate shadow for the pillar on the left even look like?

Kinda like so wrt floor shadows:pillar.zip

 

The problem here is then that you need to get a different ceiling brightness by way of boom/ZDoom actions if you wanna be accurate. And if we assume you want to be, then welcome to the spaghetti-zone.

[Nine Inch Heels]

Everything depends on where you want the light-source to be. If we assume the light comes in at a 45° angle (Z-axis), then yes, a 256mu pillar casts a 256 long shadow (ignoring differences of the size of mapunits wrt height), and a height difference of Xmu therefore means the shadow's length increases by Xmu as well, except when it doesn't (in case pillars are at different heights, which changes the angle relative to the light source). Things get difficult when you have light sources close by, and need to account for differences on the XY-axis as well the Z-axis, though.

[Barry Burton]

I'd say if you have knowledge with a 3D modeling program like Blender, maybe you could export the pillar sectors as a .obj file. Import that into Blender, add a lamp object and set the pillar object to cast shadows. I say maybe because I haven't done this myself, but it sounds like it could work.

[Kappes Buur]

This may be of interest.

 

 

Edited February 25, 2020 by Kappes Buur

[Spectre01]

Ok, sounds good. I assume for arches and whatnot, you would use the difference between the ceiling and floor for the shape of the shadow.

 

But following the same rule set, what would the accurate shadow for the pillar on the left even look like?