Procedural atmospheric scattering #16314
Conversation
and more things as well
mostly
and more things as well
mostly
|
You added a new feature but didn't update the readme. Please run |
Procedural sky integration corrections
There was a problem hiding this comment.
I think this is supposed to be alphabetical
| pub struct Atmosphere { | ||
| /// Radius of the planet | ||
| /// | ||
| /// units: km |
There was a problem hiding this comment.
While it might be beneficial for the computations to be done in km, the public API should use meters to stay consistent with the rest of bevys units. This conversion precision loss will be a one time thing when computing the uniform, and will be completely negligible.
| /// Radius at which we consider the atmosphere to 'end' for our | ||
| /// calculations (from center of planet) | ||
| /// | ||
| /// units: km |
| /// The scattering optical density of rayleigh particulare, or how | ||
| /// much light it scatters per kilometer | ||
| /// | ||
| /// units: km^-1 |
There was a problem hiding this comment.
I'd use SI unless this is what measurements are usually given in.
| /// it scatters per kilometer. | ||
| /// | ||
| /// units: km^-1 | ||
| pub mie_scattering: f32, |
There was a problem hiding this comment.
I won't bore you by repeating myself any more. 😅
|
|
||
| /// A conversion factor between scene units and kilometers, used to | ||
| /// combat floating-point precision issues. | ||
| pub scene_units_to_km: f32, |
There was a problem hiding this comment.
Same here obviously, but I think it becomes even clearer when this defaults to 1.0.
There was a problem hiding this comment.
A comment at the top of the file would be nice, describing when to use this?
examples/3d/atmosphere.rs
Outdated
| // Sun | ||
| commands.spawn(( | ||
| DirectionalLight { | ||
| color: Color::srgb(0.98, 0.95, 0.82), |
There was a problem hiding this comment.
Setting the directional light's color to srgb white (1,1,1) here results in more physically correct result since the transmittance though the atmosphere of the directional light is already accounted for with the throughput variable and the analytical integration method used.
| let sphere_mesh = meshes.add(Mesh::from(Sphere { radius: 1.0 })); | ||
|
|
||
| // light probe spheres | ||
| commands.spawn(( |
There was a problem hiding this comment.
The environment does not show up on the light probes, but this will be used as test subject in a future iteration where we generate the environment cube maps from sampling the sky-view LUT.
| fn setup_camera_fog(mut commands: Commands) { | ||
| commands.spawn(( | ||
| Camera3d::default(), | ||
| Camera { |
There was a problem hiding this comment.
Setting the exposure lower at noon and higher at sunset would make the showcase look more balanced, as well as adding a secondary directional light (moon) at night time to showcase a day-night cycle.
|
FYI turning on |
existing color approximated atmosphere transmittance (yellowish), so it was skewing the resulting atmosphere look
this is a bug then, msaa should be compatible. will fix 👍 |
Implement procedural atmospheric scattering from Sebastien Hillaire's 2020 paper. This approach should scale well even down to mobile hardware, and is physically accurate.
Co-author: @mate-h
He helped massively with getting this over the finish line, ensuring everything was physically correct, correcting several places where I had misunderstood or misapplied the paper, and improving the performance in several places as well. Thanks!
Showcase:
For followup
Credits
Built off of @mtsr's original branch, which handled the transmittance lut (arguably the most important part, and I still don't understand the parametrization)