1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
//! # Rendering
use crate::color::*;
use crate::ray::*;
use crate::tile::*;
use crate::profile::*;
use crate::produce::*;
use crate::math::*;
use crate::acc::*;
use crate::frustrum::depth_linear;
use crate::mask::CompressedMasks;
use crate::cam::{Camera, CameraPerspective};
use crate::{
IndexFlag,
PixelPos,
RayHit,
Rgba,
Triangle,
Vector,
};
/// Stores arguments for shaders.
pub struct ShaderData<Args> {
/// The ray depth and index of graphics primitive.
pub hit: RayHit,
/// The depth in range `0.0` to `1.0` where `0.0` is near clip plane
/// and `1.0` is far clip plane.
pub depth_linear: f32,
/// The internal address of hit graphics primitive in producer.
///
/// For example, when a triangle is hit,
/// the triangle might belong to some voxel in internal address space.
/// In this case, the produces generates triangles from voxels.
/// With ther words, this value tells which voxel gets hit.
pub internal_offset: Option<usize>,
/// Customized arguments to the shader.
pub args: Args,
}
/// The type of shader.
///
/// A shader might modify the default color before accumulation.
pub type Shader<Color, Args> = fn(&mut Color, ShaderData<Args>);
/// Stores data needed during rendering.
pub struct Renderer<'a, Scene, Prod, Img, A, ShaderArgs, P>
where Scene: Sync, Prod: Produce<Triangle> + Sync + ?Sized, A: Acc,
{
/// Scene data.
pub scene: Scene,
/// Gets a default ray color prior to shading.
///
/// Make sure to convert to the same color space as the accumulator.
///
/// For example, if the accumulator uses linear color space,
/// and you use sRGB color space in scene data,
/// then you should convert to linear color space.
pub scene_ray_color: fn(&Scene, f32, usize) -> (A::In, ShaderArgs),
/// A customized shader transform prior to color accumulation.
///
/// This can be used to change the color of the ray.
pub shader: Shader<A::In, ShaderArgs>,
/// Returns `true` if color is transparent, `false` otherwise.
///
/// This is used to filter out colors that do not contribute in accumulator.
pub is_transparent: fn(&A::In) -> bool,
/// Produces the final color from accumulator.
pub acc_to_linear_rgba: fn(A::Out) -> Rgba,
/// A producer.
pub producer: &'a Prod,
/// The target image.
pub img: &'a mut Img,
/// Get the size of the image in pixels.
pub size: fn(&Img) -> PixelPos,
/// Writes pixel to image.
pub pxl: fn(&mut Img, PixelPos, c: Rgba<u8>),
/// Accumulator data.
///
/// This is used to pre-configure the accumulator with some data.
pub acc_data: A::Data,
/// The camera perspective.
pub persp: &'a CameraPerspective,
/// The camera.
pub cam: &'a Camera,
/// Can be used to scale or flip axis.
pub flip_xyz: Vector,
/// Stores compressed masks per render tile.
pub compr_masks: &'a mut [CompressedMasks],
/// Stores compressed masks for pre-pre-processing.
pub pre_compr_masks: &'a mut [CompressedMasks],
/// Stores compressed masks for adaptive sub-tiling.
pub sub_compr_masks: &'a mut Vec<Vec<CompressedMasks>>,
/// A limit on the number of triangles per tile before doing adaptive sub-tiling.
///
/// Adaptive sub-tiling produces sub-tiles such that the average number of triangles
/// is less than this limit, down to 2x2 sub-tiles.
pub sub_tile_triangle_limit: u32,
/// Data used to store performance profiling information.
pub profile: &'a mut P,
/// Reports the amount of seconds taken to render (`None` if profiling is disabled).
pub profile_render: fn(&mut P, Option<f64>),
/// Reports profile compress data and amount of seconds (`None` if profiling is disabled).
pub profile_compress: fn(&mut P, ProfileCompressData, Option<f64>),
/// Whether to use adaptive sub-tiling.
///
/// Adaptive sub-tiling splits tiles that have a large amount
/// of triangles into sub-tiles, where the average amount of triangles
/// is less thatn `sub_tile_triangle_limit`.
pub sub_masks: bool,
/// Whether to use pre-masks.
///
/// This is a pre-pre-processing step where run-length compression
/// on masks uses a lower resolution, to speed up pre-processing.
pub pre_masks: bool,
/// Whether profiling is enabled.
///
/// Profiling gathers some data about the rendering process.
pub profile_enabled: bool,
/// A limit of how many accumulations per tile.
///
/// Is used to prevent inner infinite rendering loops.
///
/// This is usually some multiple of the accumulator buffer size.
pub acc_limit: u32,
/// The scale ratio between pre-tile size and tile size.
pub scale_to_pre_tile_size: u32,
}
impl<Scene, Prod, Img, Accumulator, ShaderArgs, P>
Renderer<'_, Scene, Prod, Img, Accumulator, ShaderArgs, P>
where Scene: Sync,
Prod: Produce<Triangle> + Sync + ?Sized,
Accumulator: Acc,
{
/// Render with some render tile size.
///
/// The tile size should be optimized for adaptive sub-tile rendering.
/// Use `optimal_sub_tile_size`.
pub fn render<const TILE_SIZE: usize>(self) {
let Renderer {
scene, scene_ray_color, producer,
img, size, pxl, acc_data, persp, cam, flip_xyz,
compr_masks, pre_compr_masks, sub_compr_masks,
sub_tile_triangle_limit, shader, profile, profile_render,
sub_masks, pre_masks, profile_enabled, profile_compress,
acc_limit, scale_to_pre_tile_size, is_transparent, acc_to_linear_rgba,
} = self;
let profile_without_sub_masks = !sub_masks;
let profile_without_pre_masks = !pre_masks;
use rayon::prelude::*;
use std::sync::mpsc::channel;
use vecmath::row_mat4_mul;
let [sx, sy, sz] = flip_xyz;
let flip = [
[sx, 0.0, 0.0, 0.0],
[0.0, sy, 0.0, 0.0],
[0.0, 0.0, sz, 0.0],
[0.0, 0.0, 0.0, 1.0],
];
let view = mat4_transposed(cam.orthogonal());
let view = row_mat4_mul(flip, view);
let producer: &TransformProducer<_> = &TransformProducer {
matrix: view,
inner: producer,
};
let size = (size)(img);
let [w, h] = size;
let tile_size = TILE_SIZE as u32;
let grid = tile_grid(size, tile_size);
let start: Option<f64> = if profile_enabled {Some(now())} else {None};
if profile_without_pre_masks {
masks(
persp,
size,
tile_size,
producer,
compr_masks,
);
} else {
masks(
persp,
size,
tile_size * scale_to_pre_tile_size,
producer,
pre_compr_masks,
);
masks_with_pre_masks(
persp,
size,
tile_size,
scale_to_pre_tile_size,
producer,
compr_masks,
pre_compr_masks,
);
}
let koeff: u32 = sub_tile_triangle_limit;
if !profile_without_sub_masks {
row_sub_masks(&persp, size, tile_size, grid, koeff, producer,
compr_masks, sub_compr_masks);
}
profile_compress(profile, ProfileCompressData {tile_size, grid, compr_masks}, start);
let (tx, rx) = channel();
let start: Option<f64> = if profile_enabled {Some(now())} else {None};
(0..grid[1]).into_par_iter().for_each_with(tx, |tx, tj| {
let nh = (tj + 1) * tile_size;
let th = nh.min(h) - tj * tile_size;
let mut depth_buffer = &mut [[None; TILE_SIZE]; TILE_SIZE];
let mut acc = Accumulator::new(acc_data.clone());
let sm = &sub_compr_masks[tj as usize];
for (ti, val) in row_sub_tile_iter(tile_size, grid, tj, koeff, compr_masks) {
let masks = &compr_masks[(tj * grid[0] + ti) as usize];
let triangles = masks.count_ones() as u32;
if triangles == 0 {continue};
acc.clear();
let nw = (ti + 1) * tile_size;
let tw = nw.min(w) - ti * tile_size;
let pos = [ti * tile_size, tj * tile_size];
// Stores depth and colors that the ray accumulates.
let mut write = [[[0; 4]; TILE_SIZE]; TILE_SIZE];
*depth_buffer = [[
Some((0.0, IndexFlag::from_parts(0, false))); TILE_SIZE]; TILE_SIZE];
for _ in 0..acc_limit {
match (profile_without_sub_masks, val) {
(true, _) | (false, None) => {
if !render_tile_depth_all(&persp, size, pos,
producer, masks, &mut depth_buffer) {break};
}
(false, Some((st, offset))) => {
if !render_row_sub_tile_depth_all(&persp, size, pos, st,
producer, &sm[offset..], &mut depth_buffer) {break};
}
}
for j in 0..th {
for i in 0..tw {
let hit = &mut depth_buffer[j as usize][i as usize];
*hit = if let Some((depth, ind)) = *hit {
if !ind.flag() {continue};
let internal_offset = producer.to_internal(ind.index());
let (mut color, args) = scene_ray_color(
&scene, depth, internal_offset.unwrap());
let hit = ray_hit_all_to_ray_hit(*hit);
shader(&mut color, ShaderData {
hit: hit,
depth_linear: depth_linear(&persp, hit),
internal_offset,
args,
});
if !is_transparent(&color) {acc.upd(i, j, depth, color)};
Some((depth, IndexFlag::from_parts(ind.index() + 1, false)))
} else {None}
}
}
}
for j in 0..th {
for i in 0..tw {
write[j as usize][i as usize] =
rgba_to_u8(rgba_gamma_linear_to_srgb(acc_to_linear_rgba(acc.acc(i, j))));
}
}
let _ = tx.send(([ti * tile_size, tj * tile_size], write));
}
});
for y in 0..h {
for x in 0..w {
pxl(img, [x, y], [0; 4]);
}
}
for (offset, tile) in rx {
for j in 0..TILE_SIZE as u32 {
for i in 0..TILE_SIZE as u32 {
let color = tile[j as usize][i as usize];
let x = offset[0] + i;
let y = offset[1] + j;
if x >= w || y >= h {continue};
let y = h - y - 1;
pxl(img, [x, y], color);
}
}
}
profile_render(profile, start);
}
}