program mandelbrot1; {First demo of drawing a Mandelbrot Set. This program merely demonstrates the basics of calculating and plotting.} uses crt; {A unit that allows us to read single keypresses.} {Note to C programmers: the variable type "real" is equivalent to the C type "float".} var cx,cy: real; {Where do we want to center the brot?} scale: real; {This is the "zoom" factor.} limit: word; {Divergence check value.} lp: word; {Convergence check value.} a1,b1,a2,b2: real; {For calculating the iterations.} x,y: integer; {The pixel we are drawing.} ax,ay: real; {The actual position of (x,y) in relation to the Mandelbrot set.} key: char; {Dummy value for keypresses.} procedure init256; {Initialises a VGA mode 320x200 pixels, 256 colours. A little chunky but the extra colours are worth it, and SVGA is beyond the scope of this program.} begin asm mov ah,0 mov al,$13 int $10 end; {This is a little direct assembly language. Feel free to use this routine, and the others that follow, in your own code.} end; procedure end256; {Turns off the video mode and returns to standard 16-colour text mode.} begin asm mov ah,0 mov al,$03 int $10 end; end; procedure pixel(a,b: word; c: byte); {This procedure plots a pixel at screen coordinate (a,b) in colour c. Again, there is no need for you to understand the intrinsic workings of this routine - you are welcome to use it in your programs but do be aware that silly values for a and b could cause unexpected results.} var v: word; begin v:=a+b*320; mem[$A000:v]:=c; end; begin {Set up video mode.} init256; {Set up initial values for drawing. Try compiling the program with different values here if you like!} cx:=0; cy:=0; scale:=0.02; limit:=4; {Loop through all pixels on screen. For reasons that will become clear, I am counting not from (0,0) but from (-160,-100).} for x:=-160 to 159 do for y:=-100 to 100 do begin {What is the *mathematical* value of this point?} ax:=cx+x*scale; ay:=cy+y*scale; {And now for the magic formula!} a1:=ax; b1:=ay; lp:=0; repeat {Do one iteration.} lp:=lp+1; a2:=a1*a1-b1*b1+ax; b2:=2*a1*b1+ay; {This is indeed the square of a+bi, done component-wise.} a1:=a2; b1:=b2; until (lp>255) or ((a1*a1)+(b1*b1)>limit); {The first condition is satisfied if we have convergence. The second is satisfied if we have divergence.} {Define colour and draw pixel.} if lp>255 then lp:=0; {If the point converges, it is part of the brot and we draw it with colour 0, or black.} pixel(x+160,y+100,lp); end; {Wait for keypress and return to text mode.} key:=readkey; end256; {Note: newer versions of Turbo Pascal do not need the dummy value "key", but in case you are using a version that can't I've built in a little redundancy!} end.