Better Shading on Bw Display While Rendering Filled Surfaces

Better shading on BW display while rendering filled surfaces

I managed to get this working. I ended up with hardcoded 17 shade patterns of size 8x8 pixels (created manually in Paint). These are the shades:

17 shades

From there I just use x,y of rendered pixel mod 8 as coordinate in the LUT of shades. This is the result:

shaded cube

As you can see its much more better than PRNG based shading. Here the updated code:

//------------------------------------------------------------------------------------------
//--- SSD1306 I2C OLED LCD driver ver 2.001 ------------------------------------------------
//------------------------------------------------------------------------------------------
/*
[Notes]
+ I2C transfere size is not limitted on LCD side
- No memory address reset command present so any corruption while VRAM transfere permanently damages output untill power on/off but somehow after some timeout or overflow the adress is reseted
- UC3 HW I2C limits up to 255 bytes per packet
- UC3 glitch on SDA before ACK which confuse LCD to read instead of write operation and do not ACK
*/
#ifndef _LCD_SSD1306_I2C_h
#define _LCD_SSD1306_I2C_h
//------------------------------------------------------------------------------------------
#define SSD1306_SETCONTRAST 0x81
#define SSD1306_DISPLAYALLON_RESUME 0xA4
#define SSD1306_DISPLAYALLON 0xA5
#define SSD1306_NORMALDISPLAY 0xA6
#define SSD1306_INVERTDISPLAY 0xA7
#define SSD1306_DISPLAYOFF 0xAE
#define SSD1306_DISPLAYON 0xAF
#define SSD1306_SETDISPLAYOFFSET 0xD3
#define SSD1306_SETCOMPINS 0xDA
#define SSD1306_SETVCOMDETECT 0xDB
#define SSD1306_SETDISPLAYCLOCKDIV 0xD5
#define SSD1306_SETPRECHARGE 0xD9
#define SSD1306_SETMULTIPLEX 0xA8
#define SSD1306_SETLOWCOLUMN 0x00
#define SSD1306_SETHIGHCOLUMN 0x10
#define SSD1306_SETSTARTLINE 0x40
#define SSD1306_MEMORYMODE 0x20
#define SSD1306_COLUMNADDR 0x21
#define SSD1306_PAGEADDR 0x22
#define SSD1306_COMSCANINC 0xC0
#define SSD1306_COMSCANDEC 0xC8
#define SSD1306_SEGREMAP 0xA0
#define SSD1306_CHARGEPUMP 0x8D
#define SSD1306_SWITCHCAPVCC 0x2
// Scrolling #defines
#define SSD1306_ACTIVATE_SCROLL 0x2F
#define SSD1306_DEACTIVATE_SCROLL 0x2E
#define SSD1306_SET_VERTICAL_SCROLL_AREA 0xA3
#define SSD1306_RIGHT_HORIZONTAL_SCROLL 0x26
#define SSD1306_LEFT_HORIZONTAL_SCROLL 0x27
#define SSD1306_VERTICAL_AND_RIGHT_HORIZONTAL_SCROLL 0x29
#define SSD1306_VERTICAL_AND_LEFT_HORIZONTAL_SCROLL 0x2A
//------------------------------------------------------------------------------------------
//#define I2C_send(adr,buf,siz){}
//------------------------------------------------------------------------------------------
#ifndef _brv8_tab
#define _brv8_tab
static const U8 brv8[256] = // 8 bit bit reversal
{
0,128,64,192,32,160,96,224,16,144,80,208,48,176,112,240,8,136,72,200,40,168,104,232,24,152,
88,216,56,184,120,248,4,132,68,196,36,164,100,228,20,148,84,212,52,180,116,244,12,140,76,204,
44,172,108,236,28,156,92,220,60,188,124,252,2,130,66,194,34,162,98,226,18,146,82,210,50,178,
114,242,10,138,74,202,42,170,106,234,26,154,90,218,58,186,122,250,6,134,70,198,38,166,102,230,
22,150,86,214,54,182,118,246,14,142,78,206,46,174,110,238,30,158,94,222,62,190,126,254,1,129,
65,193,33,161,97,225,17,145,81,209,49,177,113,241,9,137,73,201,41,169,105,233,25,153,89,217,57,
185,121,249,5,133,69,197,37,165,101,229,21,149,85,213,53,181,117,245,13,141,77,205,45,173,109,
237,29,157,93,221,61,189,125,253,3,131,67,195,35,163,99,227,19,147,83,211,51,179,115,243,11,139,
75,203,43,171,107,235,27,155,91,219,59,187,123,251,7,135,71,199,39,167,103,231,23,151,87,215,55,
183,119,247,15,143,79,207,47,175,111,239,31,159,95,223,63,191,127,255
};
#endif
static const U8 shade8x8[17*8]= // 17 shade patterns 8x8 pixels
{
0, 0, 0, 0, 0, 0, 0, 0,
17, 0, 0, 0, 17, 0, 0, 0,
17, 0, 68, 0, 17, 0, 68, 0,
68, 17, 68, 0, 68, 17, 68, 0,
85, 0,170, 0, 85, 0,170, 0,
68,170, 0,170, 68,170, 0,170,
68,170, 17,170, 68,170, 17,170,
85,136, 85,170, 85,136, 85,170,
85,170, 85,170, 85,170, 85,170,
85,238, 85,170,119,170, 85,170,
221,170,119,170,221,170,119,170,
221,170,255,170,221,170,255,170,
85,255,170,255, 85,255,170,255,
221,119,221,255,221,119,221,255,
119,255,221,255,119,255,221,255,
119,255,255,255,119,255,255,255,
255,255,255,255,255,255,255,255
};
//------------------------------------------------------------------------------------------
class LCD_SSD1306_I2C // max 96 lines
{
public:
// screen
int adr; // I2C adr
int xs,ys,sz; // resoluiton
U8 _scr[((128*96)>>3)+1]; // screen buffer
U8 *scr;
U8 *pyx[96]; // scan lines
// pattern
U32 pat,pat_m,pat_b; // binary pattern,max used bit mask,actual bit mask
// filling
int bufl[96];
int bufr[96];

// system api
void init(int _adr,int _xs,int _ys); // initialize LCD: I2C_adr,xs,ys
void _command(U8 cmd); // *internal* do not cal it (sends command to LCD over I2C)
void rfsscr(); // copy actual screen buffer to LCD (by I2C)
// gfx rendering col = <0,1>
void clrscr(); // clear screen buffer
void rotate(int ang); // rotate 180 deg
void pixel(int x,int y,bool col); // set/res pixel
bool pixel(int x,int y); // get pixel
void line(int x0,int y0,int x1,int y1,bool col); // line
void triangle(int x0,int y0,int x1,int y1,int x2,int y2,bool col);// triangle
void quad(int x0,int y0,int x1,int y1,int x2,int y2,int x3,int y3,bool col);
void rect(int x0,int y0,int x1,int y1,bool col); // rectangle using diagonal points
// patern rendering
void pat_set(char *s); // set binary pattern from bianry number string MSB renders first
void pat_beg(); // set pattern state to start of pattern
void pat_line(int x0,int y0,int x1,int y1,bool col);
void pat_triangle(int x0,int y0,int x1,int y1,int x2,int y2,bool col);
void pat_quad(int x0,int y0,int x1,int y1,int x2,int y2,int x3,int y3,bool col);
void pat_rect(int x0,int y0,int x1,int y1,bool col);
// filled polygons col = <0,255>
void _fill_line(int x0,int y0,int x1,int y1); // *internal* do not call it (render line into bufl/bufr)
void _fill(int Y0,int Y1,U8 col); // *internal* do not call it (render bufl/bufr onto screen)
void fill_triangle(int x0,int y0,int x1,int y1,int x2,int y2,U8 col);
void fill_quad(int x0,int y0,int x1,int y1,int x2,int y2,int x3,int y3,U8 col);
// text rendering
void prnchr(int x,int y,char c); // render char at x,y (y is rounded to multiple of 8)
void prntxt(int x,int y,const char *txt); // render text at x,y (y is rounded to multiple of 8)
};
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::_command(U8 cmd)
{
U8 buf[2]=
{
0x00, // 0x40 data/command
cmd,
};
I2C_send(adr,buf,2);
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::init(int _adr,int _xs,int _ys)
{
int y;
adr=_adr;
xs=_xs;
ys=_ys;
sz=xs*(ys>>3);
const bool _external_Vcc=false;
// VRAM buffer
scr=_scr+1; // skip first Byte (VRAM/command selection)
for (y=0;y<ys;y++) pyx[y]=scr+((y>>3)*xs); // scanlines for fast direct pixel access
clrscr();
// Init sequence
U8 comPins = 0x02;
U8 contrast = 0x8F;
if((xs == 128) && (ys == 32)) { comPins = 0x02; contrast = 0x8F; }
else if((xs == 128) && (ys == 64)) { comPins = 0x12; contrast = (_external_Vcc) ? 0x9F : 0xCF; }
else if((xs == 96) && (ys == 16)) { comPins = 0x02; contrast = (_external_Vcc) ? 0x10 : 0xAF; }
else {} // Other screens
static U8 init0[27]=
{
0x00, // commands
SSD1306_DISPLAYOFF, // 0xAE
SSD1306_SETDISPLAYCLOCKDIV,0x80, // 0xD5
SSD1306_SETMULTIPLEX,ys-1, // 0xA8
SSD1306_SETDISPLAYOFFSET,0x00, // 0xD3 no offset
SSD1306_SETSTARTLINE | 0x0, // line 0
SSD1306_CHARGEPUMP,(_external_Vcc)?0x10:0x14, // 0x8D
SSD1306_MEMORYMODE,0x00, // 0x20 horizontal (scanlines)
SSD1306_SEGREMAP | 0x1,
SSD1306_COMSCANDEC,
SSD1306_SETCOMPINS,comPins,
SSD1306_SETCONTRAST,contrast,
SSD1306_SETPRECHARGE,(_external_Vcc)?0x22:0xF1, // 0xd9
SSD1306_SETVCOMDETECT,0x40, // 0xDB
SSD1306_DISPLAYALLON_RESUME, // 0xA4
SSD1306_NORMALDISPLAY, // 0xA6
SSD1306_DEACTIVATE_SCROLL,
SSD1306_DISPLAYON, // Main screen turn on
};
I2C_send(adr,init0,sizeof(init0));
// init default pattern
pat_set("111100100");
// clear filling buffers
for (y=0;y<96;y++)
{
bufl[y]=-1;
bufr[y]=-1;
}
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::clrscr()
{
for (int a=0;a<sz;a++) scr[a]=0x00;
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::rotate(int ang)
{
U8 a0,a1;
int x0,y0,x1,y1;
if (ang==180)
for (y0=0,y1=ys-8;y0<y1;y0+=8,y1-=8)
for (x0=0,x1=xs-1;x0<xs;x0++,x1--)
{
a0=brv8[pyx[y0][x0]];
a1=brv8[pyx[y1][x1]];
pyx[y0][x0]=a1;
pyx[y1][x1]=a0;
}
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::rfsscr()
{
static const U8 init1[] =
{
0x00, // commands
SSD1306_MEMORYMODE,0, // horizontal addresing mode
SSD1306_COLUMNADDR,0,xs-1, // Column start/end address (0/127 reset)
SSD1306_PAGEADDR,0,(ys>>3)-1, // Page start/end address (0 reset)
};
I2C_send(adr,(U8*)init1,sizeof(init1));

_scr[0]=0x40; // 0x40 VRAM
// SW I2C can pass whole VRAM in single packet
// I2C_send(adr,_scr,sz+1);

// HW I2C must use packets up to 255 bytes so 128+1 (as TWIM0 on UC3 has 8 bit counter)
int i,n=128; U8 *p=_scr,a;
for (i=0;i<sz;i+=n){ a=p[0]; p[0]=0x40; I2C_send(adr,p,n+1); p[0]=a; p+=n; }
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::pixel(int x, int y,bool col)
{
// clip to screen
if ((x<0)||(x>=xs)||(y<0)||(y>=ys)) return;
// add or remove bit
if (col) pyx[y][x] |= (1<<(y&7));
else pyx[y][x] &= (255)^(1<<(y&7));
}
//------------------------------------------------------------------------------------------
bool LCD_SSD1306_I2C::pixel(int x, int y)
{
// clip to screen
if ((x<0)||(x>=xs)||(y<0)||(y>=ys)) return false;
// get bit
return ((pyx[y][x]>>(y&7))&1);
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::line(int x0,int y0,int x1,int y1,bool col)
{
int i,n,cx,cy,sx,sy;
// line DDA parameters
x1-=x0; sx=0; if (x1>0) sx=+1; if (x1<0) { sx=-1; x1=-x1; } if (x1) x1++; n=x1;
y1-=y0; sy=0; if (y1>0) sy=+1; if (y1<0) { sy=-1; y1=-y1; } if (y1) y1++; if (n<y1) n=y1;
// single pixel (not a line)
if (!n){ pixel(x0,y0,col); return; }
// ND DDA algo i is parameter
for (cx=cy=n,i=0;i<n;i++)
{
pixel(x0,y0,col);
cx-=x1; if (cx<=0){ cx+=n; x0+=sx; }
cy-=y1; if (cy<=0){ cy+=n; y0+=sy; }
}
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::pat_line(int x0,int y0,int x1,int y1,bool col)
{
bool ccc;
int i,n,cx,cy,sx,sy;
// line DDA parameters
x1-=x0; sx=0; if (x1>0) sx=+1; if (x1<0) { sx=-1; x1=-x1; } if (x1) x1++; n=x1;
y1-=y0; sy=0; if (y1>0) sy=+1; if (y1<0) { sy=-1; y1=-y1; } if (y1) y1++; if (n<y1) n=y1;
// single pixel (not a line)
if (!n){ pixel(x0,y0,col); return; }
// ND DDA algo i is parameter
for (cx=cy=n,i=0;i<n;i++)
{
ccc=(pat&pat_b); ccc^=(!col);
pat_b>>=1; if (!pat_b) pat_b=pat_m;
pixel(x0,y0,ccc);
cx-=x1; if (cx<=0){ cx+=n; x0+=sx; }
cy-=y1; if (cy<=0){ cy+=n; y0+=sy; }
}
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::_fill_line(int x0,int y0,int x1,int y1)
{
int i,n,cx,cy,sx,sy,*buf;
// line DDA parameters
x1-=x0; sx=0; if (x1>0) sx=+1; if (x1<0) { sx=-1; x1=-x1; } if (x1) x1++; n=x1;
y1-=y0; sy=0; if (y1>0) sy=+1; if (y1<0) { sy=-1; y1=-y1; } if (y1) y1++; if (n<y1) n=y1;
// single pixel (not a line)
if (!n)
{
if ((y0>=0)&&(y0<ys))
{
bufl[y0]=x0;
bufr[y0]=x0;
}
return;
}
// target buffer depend on y direction
if (sy>0) buf=bufl; else buf=bufr;
// ND DDA algo i is parameter
for (cx=cy=n,i=0;i<n;i++)
{
if ((y0>=0)&&(y0<ys)) buf[y0]=x0;
cx-=x1; if (cx<=0){ cx+=n; x0+=sx; }
cy-=y1; if (cy<=0){ cy+=n; y0+=sy; }
}
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::_fill(int Y0,int Y1,U8 col)
{
U8 shd;
int x,y,X0,X1,i;
// select shade pattern
i=col;
if (i< 0) i=0;
if (i>17) i=17;
i<<=3;
// fill horizontal lines
for (y=Y0;y<=Y1;y++)
{
shd=shade8x8[i+(y&7)];
// x range to render
X0=bufl[y];
X1=bufr[y];
if (X0>X1){ x=X0; X0=X1; X1=x; }
// clip to screen in y axis
if ((X1<0)||(X0>=xs)) continue;
if (X0< 0) X0= 0;
if (X1>=xs) X1=xs-1;
if (col== 0) for (x=X0;x<=X1;x++) pixel(x,y,0);
else if (col==255) for (x=X0;x<=X1;x++) pixel(x,y,1);
else for (x=X0;x<=X1;x++) pixel(x,y,shd&(1<<(x&7)));
}
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::triangle(int x0,int y0,int x1,int y1,int x2,int y2,bool col)
{
line(x0,y0,x1,y1,col);
line(x1,y1,x2,y2,col);
line(x2,y2,x0,y0,col);
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::pat_triangle(int x0,int y0,int x1,int y1,int x2,int y2,bool col)
{
pat_line(x0,y0,x1,y1,col);
pat_line(x1,y1,x2,y2,col);
pat_line(x2,y2,x0,y0,col);
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::fill_triangle(int x0,int y0,int x1,int y1,int x2,int y2,U8 col)
{
int y,Y0,Y1;
// y range to render
Y0=Y1=y0;
if (Y0>y1) Y0=y1;
if (Y1<y1) Y1=y1;
if (Y0>y2) Y0=y2;
if (Y1<y2) Y1=y2;
// clip to screen in y axis
if ((Y1<0)||(Y0>=ys)) return;
if (Y0< 0) Y0= 0;
if (Y1>=ys) Y1=ys-1;
// clear buffers
for (y=Y0;y<=Y1;y++)
{
bufl[y]=xs;
bufr[y]=-1;
}
// render circumference
_fill_line(x0,y0,x1,y1);
_fill_line(x1,y1,x2,y2);
_fill_line(x2,y2,x0,y0);
// fill horizontal lines
_fill(Y0,Y1,col);
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::quad(int x0,int y0,int x1,int y1,int x2,int y2,int x3,int y3,bool col)
{
line(x0,y0,x1,y1,col);
line(x1,y1,x2,y2,col);
line(x2,y2,x3,y3,col);
line(x3,y3,x0,y0,col);
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::pat_quad(int x0,int y0,int x1,int y1,int x2,int y2,int x3,int y3,bool col)
{
pat_line(x0,y0,x1,y1,col);
pat_line(x1,y1,x2,y2,col);
pat_line(x2,y2,x3,y3,col);
pat_line(x3,y3,x0,y0,col);
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::fill_quad(int x0,int y0,int x1,int y1,int x2,int y2,int x3,int y3,U8 col)
{
int y,Y0,Y1;
// y range to render
Y0=Y1=y0;
if (Y0>y1) Y0=y1;
if (Y1<y1) Y1=y1;
if (Y0>y2) Y0=y2;
if (Y1<y2) Y1=y2;
if (Y0>y3) Y0=y3;
if (Y1<y3) Y1=y3;
// clip to screen in y axis
if ((Y1<0)||(Y0>=ys)) return;
if (Y0< 0) Y0= 0;
if (Y1>=ys) Y1=ys-1;
// clear buffers
for (y=Y0;y<=Y1;y++)
{
bufl[y]=xs;
bufr[y]=-1;
}
// render circumference
_fill_line(x0,y0,x1,y1);
_fill_line(x1,y1,x2,y2);
_fill_line(x2,y2,x3,y3);
_fill_line(x3,y3,x0,y0);
// fill horizontal lines
_fill(Y0,Y1,col);
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::rect(int x0,int y0,int x1,int y1,bool col)
{
line(x0,y0,x1,y0,col);
line(x1,y0,x1,y1,col);
line(x1,y1,x0,y1,col);
line(x0,y1,x0,y0,col);
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::pat_rect(int x0,int y0,int x1,int y1,bool col)
{
pat_line(x0,y0,x1,y0,col);
pat_line(x1,y0,x1,y1,col);
pat_line(x1,y1,x0,y1,col);
pat_line(x0,y1,x0,y0,col);
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::prnchr(int x,int y,char c)
{
y&=0xFFFFFFF8; // multiple of 8
if ((y<0)||(y>ys-8)) return;
int i,a;
a=c; a<<=3;
for (i=0;i<8;i++,x++,a++)
if ((x>=0)&&(x<xs))
pyx[y][x]=font[a];
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::prntxt(int x,int y,const char *txt)
{
for (;*txt;txt++,x+=8) prnchr(x,y,*txt);
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::pat_set(char *s)
{
int i=1;
pat=0;
if (s!=NULL)
for (i=0;(*s)&&(i<32);s++,i++)
{
pat<<=1;
if (*s=='1') pat|=1;
}
if (!i) i=1;
pat_m=1<<(i-1);
pat_beg();
}
//------------------------------------------------------------------------------------------
void LCD_SSD1306_I2C::pat_beg()
{
pat_b=pat_m;
}
//------------------------------------------------------------------------------------------
#undef I2C_send
//------------------------------------------------------------------------------------------
#endif
//------------------------------------------------------------------------------------------

I moved all the filling into member function _fill just see how the LUT shade8x8 is used ...

[Edit1] Floyd–Steinberg dithering

Thanks to Scheff I wanted to try this kind of dithering. Sadly his code is not usable for polygon rasterization (due to its specifics and lack of input image) so I implemented in on my own. I struggled a while to make it work properly and the only way ot worked as expected was when:

  • color threshold is 50% of max intensity
  • error propagation is done on signed integers

The second requirement brings huge performance hit as simple bitshift cant be used anymore however I think hardcoded LUTs will overcome this.

Here my current implementation with booth ditherings:

void LCD_SSD1306_I2C::_fill(int Y0,int Y1,U8 col)
{
int x,y,X0,X1,i;
const U8 colmax=17;
// bayer like dithering using precomputed shader patterns
if (dither_mode==0)
{
U8 shd;
// select shade pattern
i=col;
if (i< 0) i=0;
if (i>17) i=17;
i<<=3;
// fill horizontal lines
for (y=Y0;y<=Y1;y++)
{
shd=shade8x8[i+(y&7)];
// x range to render
X0=bufl[y];
X1=bufr[y];
if (X0>X1){ x=X0; X0=X1; X1=x; }
// clip to screen in y axis
if ((X1<0)||(X0>=xs)) continue;
if (X0< 0) X0= 0;
if (X1>=xs) X1=xs-1;
if (col== 0) for (x=X0;x<=X1;x++) pixel(x,y,0);
else if (col==colmax) for (x=X0;x<=X1;x++) pixel(x,y,1);
else for (x=X0;x<=X1;x++) pixel(x,y,shd&(1<<(x&7)));
}
}
// Floyd–Steinberg dithering
if (dither_mode==1)
{
int c0,c1,c2,rows[256+4],*r0=rows+1,*r1=rows+128+3,*rr,thr=colmax/2;
// clear error;
c0=0; for (x=0;x<256+4;x++) rows[x]=0;
// fill horizontal lines
for (y=Y0;y<=Y1;y++)
{
// x range to render
X0=bufl[y];
X1=bufr[y];
if (X0>X1){ x=X0; X0=X1; X1=x; }
// clip to screen in y axis
if ((X1<0)||(X0>=xs)) continue;
if (X0< 0) X0= 0;
if (X1>=xs) X1=xs-1;
if (col== 0) for (x=X0;x<=X1;x++) pixel(x,y,0);
else if (col==colmax) for (x=X0;x<=X1;x++) pixel(x,y,1);
else for (x=X0;x<=X1;x++)
{
c0=col; c0+=r0[x]; ; r0[x]=0;
if (c0>thr){ pixel(x,y,1); c0-=colmax; }
else pixel(x,y,0);
c2=c0;
c1=(3*c0)/16; r1[x-1] =c1; c2-=c1;
c1=(5*c0)/16; r1[x ] =c1; c2-=c1;
c1=( c0)/16; r1[x+1] =c1; c2-=c1;
r0[x+1]+=c2;
}
rr=r0; r0=r1; r1=rr;
}
}
}

The dither_mode is just int deciding which dithering to use. Here preview for both side by side. On the left is Bayer ditheringlike (with mine painted pattern) and right is the Floyd–Steinberg dithering:

side by side

The Floyd–Steinberg dithering sometimes create ugly (parallel lines) pattern in some rotations while the constant patterns look smooth in all cases So I will stick with my original rendering.

Resizing single 1 pixel wide bitmap strip - faster than this example? (for Raycaster algorithm)

  1. You do not need floats at all

    You can use DDA on integers without multiplication and division. These days floating is not that slow as it used to but your conversion between float and int might be ... See these QAs (both use this kind of DDA:

    • DDA line with subpixel
    • DDA based rendering routines
  2. use LUT for applying Intensity

    Looks like each color channel c is 8 bit and intensity i is fixed point in range <0,1> so you can precompute every combination into something like this:

    u_int8 LUT[256][256]
    for (int c=0;c<256;c++)
    for (int i=0;i<256;i++)
    LUT[c][i]=((c*i)>>8)
  3. use pointers or union to access RGB channels instead of bit operations

    My favorite is union:

    union color
    {
    u_int32 dd; // 1x 32bit RGBA
    u_int16 dw[2]; // 2x 16bit
    u_int8 db[4]; // 4x 8bit (individual channels)
    };
  4. texture coordinates

    Again looks like you are doing too many operations. for example [RC_texture_size_i * tex_y_safe + tex_x] if your texture size is 128 you can bitshift lef by 7 bits instead of multiplication. Yes on modern CPUs is this not an issue however the whole thing can be replaced by simple LUT. You can remember pointer to each horizontal ScanLine of texture and rewrite to [tex_y_safe][tex_x]

So based on #2,#3 rewrite your color computation to this:

color c;
c.dd=texture_current_pixel;
c.db[0]=LUT[c.db[0]][intensity_value];
c.db[1]=LUT[c.db[1]][intensity_value];
c.db[2]=LUT[c.db[2]][intensity_value];
output_buffer[output_pixel_index]=c.dd;

As you can see its just bunch of memory transfers instead of multiple bit-shifts,bit-masks and bit-or operations. You can also use pointer of color instead of texture_current_pixel and output_buffer[output_pixel_index] to speed up little more.

And finally see this:

  • Ray Casting with different height size

Which is my version of the raycast using VCL.

Now before changing anything measure the performance you got now by measuring the time it needs to render. Then after each change in the code measure if it actually improve performance or not. In case it didn't use old version of code as predicting what is fast on nowadays platforms is sometimes hard.

Also for resize much better visual results are obtained by using mipmaps ... that usually eliminates the weird noise while moving

Display Monitors With More Than 256 Shades of Gray

Use OpenGL. Docs from AMD and Nvidia.

In a Qt app, I'm guessing (haven't tried) it ought to be possible to obtain a high bit-depth area for the application by creating a QGLWidget with an appropriately set up QGLFormat (set the bit depths with setRedBufferSize etc). Note that things like QPainter work very well on QGLWidget, so it's may not be necessary to port all your Qt code to OpenGL calls, only the bits you actually want to access the extra precision. The limited formats supported by QImage is certainly a bit of a weakness here though.



Related Topics



Leave a reply



Submit