krzosa/software_rasterizer
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
fd8ce7d1a9176dbb7b1dfb0bce1cc6bbcc9e782c
software_rasterizer/main.cpp
Krzosa Karol fd8ce7d1a9 Merge with base
2022-02-21 17:39:19 +01:00

406 lines
13 KiB
C++
Raw Blame History

/* Things to do:
OK Drawing triangles
OK Drawing cubes and lines for testing
OK Y up coordinate system, left handed
OK Drawing a cube with perspective
OK Culling triangles facing away from camera
OK Texture mapping
? Basic math operations on Vec4 Mat4 - Muls, dot, cross etc.
OK Basic linear transformations - rotation, translation, scaling
OK Bilinear filtering of textures / subpixel precison
OK Fix the gaps between triangles (it also improved look of triangle edges)
* Perspective matrix vs simple perspective
OK Perspective correct interpolation
OK Depth buffer
KINDA_OK Gamma correct blending
OK Alpha blending??
OK Premultiplied alpha???
OK Merge with base
* Lightning
OK GLOBAL Ilumination
* LookAt Camera
* FPS Camera
OK Reading OBJ files
* Reading PMX files
* Rendering multiple objects, queue renderer
* Clipping
* Optimizations
* SIMD
* Multithreading
*
* Text rendering
* Basic UI
* Gamma correct and alpha blending
*/
#define OS_WINDOWS 1
#define PERSPECTIVE_CORRECT_INTERPOLATION 1
#define BILINEAR_BLEND 1
#define GAMMA_CORRECT_BLENDING 1
#define PREMULTIPLIED_ALPHA_BLENDING 1
#define PLATFORM
#include "base.h"
#include "math.h"
#include "stb_image.h"
#include "objparser.h"
#include <float.h>
//GLOBAL OS os = {};
GLOBAL bool draw_rects = 0;
GLOBAL bool draw_wireframe = 0;
FUNCTION
void draw_rect(Bitmap* dst, float X, float Y, float w, float h, U32 color) {
int max_x = (int)(MIN(X + w, dst->x) + 0.5f);
int max_y = (int)(MIN(Y + h, dst->y) + 0.5f);
int min_x = (int)(MAX(0, X) + 0.5f);
int min_y = (int)(MAX(0, Y) + 0.5f);
for (int y = min_y; y < max_y; y++) {
for (int x = min_x; x < max_x; x++) {
dst->pixels[x + y * dst->x] = color;
}
}
}
FUNCTION
float edge_function(Vec4 vecp0, Vec4 vecp1, Vec4 p) {
float result = (vecp1.y - vecp0.y) * (p.x - vecp0.x) - (vecp1.x - vecp0.x) * (p.y - vecp0.y);
return result;
}
FUNCTION
Vec4 srgb_to_almost_linear(Vec4 a) {
Vec4 result = {a.r*a.r, a.g*a.g, a.b*a.b, a.a};
return result; // @Note: Linear would be to power of 2.2
}
FUNCTION
Vec4 almost_linear_to_srgb(Vec4 a) {
Vec4 result = { sqrtf(a.r), sqrtf(a.g), sqrtf(a.b), a.a };
return result;
}
FUNCTION
void draw_bitmap(Bitmap* dst, Bitmap* src, Vec2 pos) {
I64 minx = (I64)(pos.x + 0.5);
I64 miny = (I64)(pos.y + 0.5);
I64 maxx = minx + src->x;
I64 maxy = miny + src->y;
I64 offsetx = 0;
I64 offsety = 0;
if (maxx > dst->x) {
maxx = dst->x;
}
if (maxy > dst->y) {
maxy = dst->y;
}
if (minx < 0) {
offsetx = -minx;
minx = 0;
}
if (miny < 0) {
offsety = -miny;
miny = 0;
}
for (I64 y = miny; y < maxy; y++) {
for (I64 x = minx; x < maxx; x++) {
I64 tx = x - minx + offsetx;
I64 ty = y - miny + offsety;
U32 *dst_pixel = dst->pixels + (x + y * dst->x);
U32 *pixel = src->pixels + (tx + ty * src->x);
Vec4 result_color = srgb_to_almost_linear(vec4abgr(*pixel));
Vec4 dst_color = srgb_to_almost_linear(vec4abgr(*dst_pixel));
result_color.r = result_color.r + (1-result_color.a) * dst_color.r;
result_color.g = result_color.g + (1-result_color.a) * dst_color.g;
result_color.b = result_color.b + (1-result_color.a) * dst_color.b;
result_color.a = result_color.a + dst_color.a - result_color.a*dst_color.a;
result_color = almost_linear_to_srgb(result_color);
U32 color32 = vec4_to_u32abgr(result_color);
*dst_pixel = color32;
}
}
}
FUNCTION
void draw_triangle(Bitmap* dst, float *depth_buffer, Bitmap *src, float light,
Vec4 p0, Vec4 p1, Vec4 p2,
Vec2 tex0, Vec2 tex1, Vec2 tex2) {
float min_x1 = (float)(MIN(p0.x, MIN(p1.x, p2.x)));
float min_y1 = (float)(MIN(p0.y, MIN(p1.y, p2.y)));
float max_x1 = (float)(MAX(p0.x, MAX(p1.x, p2.x)));
float max_y1 = (float)(MAX(p0.y, MAX(p1.y, p2.y)));
I64 min_x = (I64)MAX(0, floor(min_x1));
I64 min_y = (I64)MAX(0, floor(min_y1));
I64 max_x = (I64)MIN(dst->x, ceil(max_x1));
I64 max_y = (I64)MIN(dst->y, ceil(max_y1));
float area = edge_function(p0, p1, p2);
for (I64 y = min_y; y < max_y; y++) {
for (I64 x = min_x; x < max_x; x++) {
float edge1 = edge_function(p0, p1, { (float)x,(float)y });
float edge2 = edge_function(p1, p2, { (float)x,(float)y });
float edge3 = edge_function(p2, p0, { (float)x,(float)y });
if (edge1 >= 0 && edge2 >= 0 && edge3 >= 0) {
float w1 = edge2 / area;
float w2 = edge3 / area;
float w3 = edge1 / area;
float interpolated_z = (1.f / p0.w) * w1 + (1.f / p1.w) * w2 + (1.f / p2.w) * w3;
#if PERSPECTIVE_CORRECT_INTERPOLATION
float u = tex0.x * (w1 / p0.w) + tex1.x * (w2 / p1.w) + tex2.x * (w3 / p2.w);
float v = tex0.y * (w1 / p0.w) + tex1.y * (w2 / p1.w) + tex2.y * (w3 / p2.w);
u /= interpolated_z;
v /= interpolated_z;
#else
float u = tex0.x * w1 + tex1.x * w2 + tex2.x * w3;
float v = tex0.y * w1 + tex1.y * w2 + tex2.y * w3;
#endif
// @Note: We could do: interpolated_z = 1.f / interpolated_z to get proper depth
// but why waste an instruction, the smaller the depth value the farther the object
float* depth = depth_buffer + (x + y * dst->x);
if (*depth < interpolated_z) {
*depth = interpolated_z;
u = u * (src->x - 2);
v = v * (src->y - 2);
I64 ui = (I64)(u);
I64 vi = (I64)(v);
float udiff = u - (float)ui;
float vdiff = v - (float)vi;
// Origin UV (0,0) is in bottom left
U32 *pixel = src->pixels + (ui + (src->y - 1ll - vi) * src->x);
U32 *dst_pixel = dst->pixels + (x + y * dst->x);
#if BILINEAR_BLEND
Vec4 pixelx1y1 = vec4abgr(*pixel);
Vec4 pixelx2y1 = vec4abgr(*(pixel + 1));
Vec4 pixelx1y2 = vec4abgr(*(pixel - src->x));
Vec4 pixelx2y2 = vec4abgr(*(pixel + 1 - src->x));
#if GAMMA_CORRECT_BLENDING
pixelx1y1 = srgb_to_almost_linear(pixelx1y1);
pixelx2y1 = srgb_to_almost_linear(pixelx2y1);
pixelx1y2 = srgb_to_almost_linear(pixelx1y2);
pixelx2y2 = srgb_to_almost_linear(pixelx2y2);
#endif // GAMMA_CORRECT_BLENDING
Vec4 blendx1 = lerp(pixelx1y1, pixelx2y1, udiff);
Vec4 blendx2 = lerp(pixelx1y2, pixelx2y2, udiff);
Vec4 result_color = lerp(blendx1, blendx2, vdiff);
result_color.r *= light;
result_color.g *= light;
result_color.b *= light;
#if PREMULTIPLIED_ALPHA_BLENDING
Vec4 dst_color = vec4abgr(*dst_pixel);
#if GAMMA_CORRECT_BLENDING
dst_color = srgb_to_almost_linear(dst_color);
#endif
result_color.r = result_color.r + (1-result_color.a) * dst_color.r;
result_color.g = result_color.g + (1-result_color.a) * dst_color.g;
result_color.b = result_color.b + (1-result_color.a) * dst_color.b;
result_color.a = result_color.a + dst_color.a - result_color.a*dst_color.a;
#endif // PREMULTIPLIED_ALPHA_BLENDING
#if GAMMA_CORRECT_BLENDING
result_color = almost_linear_to_srgb(result_color);
//ASSERT(result_color.r <= 1 && result_color.g <= 1 && result_color.b <= 1);
#endif // GAMMA_CORRECT_BLENDING
U32 color32 = vec4_to_u32abgr(result_color);
#else // BILINEAR_BLEND
Vec4 result_color = srgb_to_almost_linear(vec4abgr(*pixel));
Vec4 dst_color = srgb_to_almost_linear(vec4abgr(*dst_pixel));
result_color.r = result_color.r + (1-result_color.a) * dst_color.r;
result_color.g = result_color.g + (1-result_color.a) * dst_color.g;
result_color.b = result_color.b + (1-result_color.a) * dst_color.b;
result_color.a = result_color.a + dst_color.a - result_color.a*dst_color.a;
result_color = almost_linear_to_srgb(result_color);
U32 color32 = vec4_to_u32abgr(result_color);
#endif // BILINEAR_BLEND
*dst_pixel = color32;
}
}
}
}
if (draw_rects) {
draw_rect(dst, p0.x-4, p0.y-4, 8,8, 0x00ff0000);
draw_rect(dst, p1.x-4, p1.y-4, 8,8, 0x0000ff00);
draw_rect(dst, p2.x-4, p2.y-4, 8,8, 0x000000ff);
}
}
FUNCTION
void draw_line(Bitmap *dst, float x0, float y0, float x1, float y1) {
float delta_x = (x1 - x0);
float delta_y = (y1 - y0);
float longest_side_length = (ABS(delta_x) >= ABS(delta_y)) ? ABS(delta_x) : ABS(delta_y);
float x_inc = delta_x / (float)longest_side_length;
float y_inc = delta_y / (float)longest_side_length;
float current_x = (float)x0;
float current_y = (float)y0;
for (int i = 0; i <= longest_side_length; i++) {
int x = (int)(current_x + 0.5f);
int y = (int)(current_y + 0.5f);
dst->pixels[x + y * dst->x] = 0xffffffff;
current_x += x_inc;
current_y += y_inc;
}
}
struct FaceA {
int vertex[3];
int tex[3];
int normal[3];
};
FN void null_terminate(Arena *arena) {
PUSH_SIZE(arena, 1);
}
FUNCTION
Obj load_obj(S8 file) {
Scratch scratch;
S8 data = os_read_file(scratch, file).error_is_fatal();
null_terminate(scratch);
char* memory = (char*)malloc(100000);
Obj result = obj::parse(memory, 100000, (char *)data.str);
return result;
}
FUNCTION
Bitmap load_image(const char* path) {
int x, y, n;
unsigned char* data = stbi_load(path, &x, &y, &n, 4);
Bitmap result = { (U32*)data, x, y };
#if PREMULTIPLIED_ALPHA_BLENDING
U32 *p = result.pixels;
for (int Y = 0; Y < y; Y++) {
for (int X = 0; X < x; X++) {
Vec4 color = vec4abgr(*p);
color.r *= color.a;
color.g *= color.a;
color.b *= color.a;
*p++ = vec4_to_u32abgr(color);
}
}
#endif
return result;
}
int main() {
obj::test();
os.window_size.x = 1280;
os.window_size.y = 720;
os_init();
os_init_software_render();
float rotation = 0;
Vec3 camera_pos = {0,0,-5};
Obj obj = load_obj(LIT("assets/f22.obj"));
Vec3* vertices = (Vec3 *)obj.vertices;
Vec2* tex_coords = (Vec2*)obj.texture;
Vec3 *normals = (Vec3 *)obj.normals;
FaceA* faces = (FaceA*)obj.indices;
I64 face_count = obj.indices_count;
Bitmap img = load_image("assets/bricksx64.png");
int screen_x = 320;
int screen_y = 180;
Bitmap screen320 = {(U32 *)malloc(screen_x*screen_y*sizeof(U32)), screen_x, screen_y};
float* depth320 = (float *)malloc(sizeof(float) * screen_x * screen_y);
while (os_game_loop()) {
Mat4 perspective = make_matrix_perspective(60.f, (float)screen->x, (float)screen->y, 0.1f, 100.f);
U32* p = screen320.pixels;
for (int y = 0; y < screen320.y; y++) {
for (int x = 0; x < screen320.x; x++) {
*p++ = 0x33333333;
}
}
float* dp = depth320;
for (int y = 0; y < screen320.y; y++) {
for (int x = 0; x < screen320.x; x++) {
*dp++ = -FLT_MAX;
}
}
Mat4 transform = make_matrix_rotation_z(rotation);
transform = transform * make_matrix_rotation_x(rotation);
if (os.key[Key_Escape].pressed) os.quit = true;
if (os.key[Key_O].down) rotation += 0.05f;
if (os.key[Key_P].down) rotation -= 0.05f;
if (os.key[Key_F1].pressed) draw_rects = !draw_rects;
if (os.key[Key_F2].pressed) draw_wireframe = !draw_wireframe;
for (int i = 0; i < face_count; i++) {
FaceA* face = faces + i;
Vec4 pos[3] = {
vec4(vertices[face->vertex[0] - 1], 1),
vec4(vertices[face->vertex[1] - 1], 1),
vec4(vertices[face->vertex[2] - 1], 1),
};
Vec2 tex[3] = {
tex_coords[face->tex[0] - 1],
tex_coords[face->tex[1] - 1],
tex_coords[face->tex[2] - 1],
};
Vec3 norm[3] = {
normals[face->normal[0] - 1],
normals[face->normal[1] - 1],
normals[face->normal[2] - 1],
};
//@Note: Transform
for (int j = 0; j < 3; j++) {
pos[j] = transform * pos[j];
}
//@Note: Cull
Vec3 p0_to_camera = camera_pos - pos[0].xyz;
Vec3 p0_to_p1 = pos[1].xyz - pos[0].xyz;
Vec3 p0_to_p2 = pos[2].xyz - pos[0].xyz;
Vec3 normal = normalize(cross(p0_to_p1, p0_to_p2));
float light = -dot(normal, vec3(0,1,0));
light = CLAMP(0.05f, light, 1.f);
if (dot(normal, p0_to_camera) > 0) {
for (int j = 0; j < 3; j++) {
//@Note: Camera
pos[j].xyz = pos[j].xyz - camera_pos;
//@Note: Perspective
pos[j] = perspective * pos[j];
pos[j].x = pos[j].x / pos[j].w;
pos[j].y = pos[j].y / pos[j].w;
pos[j].z = pos[j].z / pos[j].w;
//@Note: To pixel space
pos[j].x *= screen320.x / 2;
pos[j].y *= screen320.y / 2;
pos[j].x += screen320.x / 2;
pos[j].y += screen320.y / 2;
}
draw_triangle(&screen320, depth320, &img, light, pos[0], pos[1], pos[2], tex[0], tex[1], tex[2]);
for (int j = 0; j < 3; j++) {
pos[j].x += screen320.x / 8;
pos[j].y += screen320.y / 8;
}
draw_triangle(&screen320, depth320, &img, light, pos[0], pos[1], pos[2], tex[0], tex[1], tex[2]);
if (draw_wireframe) {
draw_line(&screen320, pos[0].x, pos[0].y, pos[1].x, pos[1].y);
draw_line(&screen320, pos[1].x, pos[1].y, pos[2].x, pos[2].y);
draw_line(&screen320, pos[2].x, pos[2].y, pos[0].x, pos[0].y);
}
}
}
// @Note: Draw 320screen to OS screen
U32* ptr = screen->pixels;
for (int y = 0; y < screen->y; y++) {
for (int x = 0; x < screen->x; x++) {
float u = (float)x / (float)screen->x;
float v = (float)y / (float)screen->y;
int tx = (int)(u * screen320.x );
int ty = (int)(v * screen320.y );
*ptr++ = screen320.pixels[tx + ty * (screen320.x)];
}
}
}
}
Reference in New Issue View Git Blame Copy Permalink