Cube  Diff

{
	if (SOLID(o) || o->type == SEMISOLID) {
		float c1 = s->floor;
		float c2 = s->floor;
		if (s->type == FHF) {
			c1 -= d1->vdelta / 4.0f;
			c2 -= d2->vdelta / 4.0f;
		};
		}
		float f1 = s->ceil;
		float f2 = s->ceil;
		if (s->type == CHF) {
			f1 += d1->vdelta / 4.0f;
			f2 += d2->vdelta / 4.0f;
		};
		}
		// if(f1-c1<=0 && f2-c2<=0) return;
		render_square(o->wtex, c1, c2, f1, f2, x1 << mip, y1 << mip,
		    x2 << mip, y2 << mip, 1 << mip, d1, d2, topleft);
		return;
	};
	}
	{
		float f1 = s->floor;
		float f2 = s->floor;
		float c1 = o->floor;
		float c2 = o->floor;
		if (o->type == FHF && s->type != FHF) {
			c1 -= d1->vdelta / 4.0f;
			c2 -= d2->vdelta / 4.0f;
		}
		if (s->type == FHF && o->type != FHF) {
			f1 -= d1->vdelta / 4.0f;
			f2 -= d2->vdelta / 4.0f;
		}
		if (f1 >= c1 && f2 >= c2)
			goto skip;
		render_square(o->wtex, f1, f2, c1, c2, x1 << mip, y1 << mip,
		    x2 << mip, y2 << mip, 1 << mip, d1, d2, topleft);
	};
	}
skip: {
	float f1 = o->ceil;
	float f2 = o->ceil;
	float c1 = s->ceil;
	float c2 = s->ceil;
	if (o->type == CHF && s->type != CHF) {
		f1 += d1->vdelta / 4.0f;
		f2 += d2->vdelta / 4.0f;
	} else if (s->type == CHF && o->type != CHF) {
		c1 += d1->vdelta / 4.0f;
		c2 += d2->vdelta / 4.0f;
	}
	if (c1 <= f1 && c2 <= f2)
		return;
	render_square(o->utex, f1, f2, c1, c2, x1 << mip, y1 << mip, x2 << mip,
	    y2 << mip, 1 << mip, d1, d2, topleft);
};
}
}

const int MAX_MIP = 5; // 32x32 unit blocks
const int MIN_LOD = 2;
const int LOW_LOD = 25;
const int MAX_LOD = 1000;

		if (issemi(mip, x + x1, y + y1, x1, y1, x2, y2))
			return true;
	case CORNER:
	case SOLID:
		break;
	default:
		return true;
	};
	}
	switch (SWS(w, x + x2, y + y2, msize)->type) {
	case SEMISOLID:
		if (issemi(mip, x + x2, y + y2, x1, y1, x2, y2))
			return true;
	case CORNER:
	case SOLID:
		break;
	default:
		return true;
	};
	}
	return false;
}

bool render_floor, render_ceil;

// the core recursive function, renders a rect of cubes at a certain mip level
// from a viewer perspective call itself for lower mip levels, on most modern

	    vxx - lodleft; // these mark the rect inside the current rest that
	                   // we want to render using a lower mip level
	int ly = vyy - lodtop;
	int rx = vxx + lodright;
	int ry = vyy + lodbot;

	float fsize = (float)(1 << mip);
	for (int ox = x; ox < xs; ox++)
	for (int ox = x; ox < xs; ox++) {
		for (int oy = y; oy < ys;
		     oy++) // first collect occlusion information for this block
		// first collect occlusion information for this block
		{
		for (int oy = y; oy < ys; oy++) {
			SWS(w, ox, oy, sz)->occluded =
			    isoccluded(player1->o.x, player1->o.y,
			        (float)(ox << mip), (float)(oy << mip), fsize);
		};
		}
	}

	int pvx = (int)vx >> mip;
	int pvy = (int)vy >> mip;
	if (pvx >= 0 && pvy >= 0 && pvx < sz && pvy < sz) {
		// SWS(w,vxx,vyy,sz)->occluded = 0;
		SWS(w, pvx, pvy, sz)->occluded =
		    0; // player cell never occluded
	};
	}

#define df(x) s->floor - (x->vdelta / 4.0f)
#define dc(x) s->ceil + (x->vdelta / 4.0f)

	// loop through the rect 3 times (for floor/ceil/walls seperately, to
	// facilitate dynamic stripify) for each we skip occluded cubes
	// (occlusion at higher mip levels is a big time saver!). during the

		render_flat(s->ctex, xx << mip, yy << mip, 1 << mip, s->ceil, s,
		    t, u, v, true);
	if (s->type == CHF) // if(s->ceil>=vh)
		render_flatdelta(s->ctex, xx << mip, yy << mip, 1 << mip, dc(s),
		    dc(t), dc(u), dc(v), s, t, u, v, true);
}
}
;

LOOPH continue; // floors
LOOPD
if ((s->type == SPACE || s->type == CHF) && s->floor <= vh && render_floor) {
	render_flat(s->ftex, xx << mip, yy << mip, 1 << mip, s->floor, s, t, u,
	    v, false);
	if (s->floor < hdr.waterlevel && !SOLID(s))
		addwaterquad(xx << mip, yy << mip, 1 << mip);
};
}
if (s->type == FHF) {
	render_flatdelta(s->ftex, xx << mip, yy << mip, 1 << mip, df(s), df(t),
	    df(u), df(v), s, t, u, v, false);
	if (s->floor - s->vdelta / 4.0f < hdr.waterlevel && !SOLID(s))
		addwaterquad(xx << mip, yy << mip, 1 << mip);
};
}
}
;
}

LOOPH continue; // walls
LOOPD
//  w
// zSt
//  vu

		if (SOLID(w)) {
			render_wall(w, h2 = s, xx + 1, yy, xx, yy + 1, mip, t,
			    v, false);
			topleft = false;
		} else if (SOLID(v)) {
			render_wall(v, h2 = s, xx, yy, xx + 1, yy + 1, mip, s,
			    u, false);
		};
		}
	} else if (SOLID(t)) {
		if (SOLID(w)) {
			render_wall(w, h1 = s, xx + 1, yy + 1, xx, yy, mip, u,
			    s, false);
		} else if (SOLID(v)) {
			render_wall(v, h1 = s, xx, yy + 1, xx + 1, yy, mip, v,
			    t, false);
			topleft = false;
		};
		}
	} else {
		normalwall = false;
		bool wv = w->ceil - w->floor < v->ceil - v->floor;
		if (z->ceil - z->floor < t->ceil - t->floor) {
			if (wv) {
				render_wall(h1 = s, h2 = v, xx + 1, yy, xx,
				    yy + 1, mip, t, v, false);
				topleft = false;
			} else {
				render_wall(h1 = s, h2 = w, xx, yy, xx + 1,
				    yy + 1, mip, s, u, false);
			};
			}
		} else {
			if (wv) {
				render_wall(h2 = s, h1 = v, xx + 1, yy + 1, xx,
				    yy, mip, u, s, false);
			} else {
				render_wall(h2 = s, h1 = w, xx, yy + 1, xx + 1,
				    yy, mip, v, t, false);
				topleft = false;
			};
		};
	};
			}
		}
	}
	render_tris(
	    xx << mip, yy << mip, 1 << mip, topleft, h1, h2, s, t, u, v);
}

if (normalwall) {
	bool inner = xx != sz - 1 && yy != sz - 1;

	    (!SOLID(s) || w->type != CORNER) &&
	    (w->type != SEMISOLID || issemi(mip, xx, yy - 1, 0, 1, 1, 1)))
		render_wall(s, w, xx, yy, xx + 1, yy, mip, s, t, false);
	if (yy <= vyy && inner && !SOLID(v) &&
	    (!SOLID(s) || v->type != CORNER) &&
	    (v->type != SEMISOLID || issemi(mip, xx, yy + 1, 0, 0, 1, 0)))
		render_wall(s, v, xx, yy + 1, xx + 1, yy + 1, mip, v, u, true);
};
}
}
;
}
;
}

void
distlod(int &low, int &high, int angle, float widef)
{
	float f = 90.0f / lod / widef;
	low = (int)((90 - angle) / f);
	high = (int)(angle / f);

    float vx, float vy, float vh, int yaw, int pitch, float fov, int w, int h)
{
	loopi(LARGEST_FACTOR) stats[i] = 0;
	min_lod = MIN_LOD + abs(pitch) / 12;
	yaw = 360 - yaw;
	float widef = fov / 75.0f;
	int cdist = abs(yaw % 90 - 45);
	if (cdist < 7) // hack to avoid popup at high fovs at 45 yaw
	// hack to avoid popup at high fovs at 45 yaw
	{
		min_lod = max(min_lod,
	if (cdist < 7) {
		// less if lod worked better
		min_lod =
		    (int)(MIN_LOD +
		        (10 - cdist) / 1.0f *
		    max(min_lod, (int)(MIN_LOD + (10 - cdist) / 1.0f * widef));
		            widef)); // less if lod worked better
		widef = 1.0f;
	};
	}
	lod = MAX_LOD;
	lodtop = lodbot = lodleft = lodright = min_lod;
	if (yaw > 45 && yaw <= 135) {
		lodleft = lod;
		distlod(lodtop, lodbot, yaw - 45, widef);
	} else if (yaw > 135 && yaw <= 225) {
		lodbot = lod;
		distlod(lodleft, lodright, yaw - 135, widef);
	} else if (yaw > 225 && yaw <= 315) {
		lodright = lod;
		distlod(lodbot, lodtop, yaw - 225, widef);
	} else {
		lodtop = lod;
		distlod(
		    lodright, lodleft, yaw <= 45 ? yaw + 45 : yaw - 315, widef);
	};
	}
	float hyfov = fov * h / w / 2;
	render_floor = pitch < hyfov;
	render_ceil = -pitch < hyfov;

	render_seg_new(
	    vx, vy, vh, MAX_MIP, 0, 0, ssize >> MAX_MIP, ssize >> MAX_MIP);
	mipstats(stats[0], stats[1], stats[2]);
}