no way to compare when less than two revisions

Differences

This shows you the differences between two versions of the page.

@@ Line 1: / Line 1: @@
+====== Sprite Multiplexer ======
+By Fungus/Nostalgia.
+The sources are in Turbo Assembler format.
+You should be able to have 1 independant color per sprite, and also an independant image for each
+sprite. You could also add fore/background priority buffers and add that register to your plot code.
+I chose to go for shortness in this routine rather than pure speed. It can be speeded up about
+% by unrolling all the plotting loops and using self modifed code instead of the second set
+of plotting buffers. d010 and d012 storing and comparing could also be precalculated.
+Now, the goods.
+We are using 32 sprites, so we are going to call this variable (maxspr) from here on.
+All of our defined tables will be of length(maxspr). The first table we need, is our linked
+list. Our most important table, as we sort this linked list instead of all our tables.
+Lets define it.
+<code>
+maxspr = $1f
+sort   = $02
+       ldx #$00      ;init x with 00
+loop   txa           ;transfer x to a to create index vals
+       sta sort,x    ;store it in the sort table
+       inx
+       cpx maxspr+1  ;have we reached 32 yet?
+       bne loop
+</code>
+This routine puts numbers $00 to $1f (32) in the linked list table. Now we have our sort table defined.
+The sort buffers can be in any memory area. I chose zeropage for speed. Now lets set up some sprites
+and colors in our sprite data buffers.
+ each buffer takes (maxspr) number of bytes.
+<code>
+ypos = $c000    ;sprite y position frame buffer
+ybuf = $22      ;sprite y position raster buffer
+xpos = $c020    ;sprite x position frame buffer
+xbuf = $42      ;sprite x position raster buffer
+xmsb = $c040    ;sprite x msb frame buffer
+mbuf = $62      ;sprite x msb raster buffer
+sprc = $c060    ;sprite color frame buffer
+cbuf = $82      ;sprite color raster buffer
+sprp = $c080    ;sprite pointer frame buffer
+pbuf = $a2      ;sprite pointer raster buffer
+      jsr movspr     ;preset sprite coordinates
+      jsr anim       ;animate sprites
+      jsr colors     ;adjust sprite colors.
+</code>
+ Now that we have set up some sprite to show, lets go ahead and precalculate the first frame
+to prevent any update bugs, due to excessive sorting of the first frame. This is a special
+sorting algorythm, which I discovered on Cadaver's covert bit ops page. It comes from the
+Imagine games done for Konami and several others. It's nice too see that good old 64
+programmers came up with the idea of a prediction sort that long ago :) The speed of this
+sorting technique is unmatched, it uses a prediction, which is really just the previous
+frame's sorted index. Y positions generally do no change by a radical amount from frame to
+frame. This is ideal for Multiplexer sorting, as it takes less than 2000 cycles to sort 32 sprites.
+Otherwise the routine is a simple swap sorter, which is not very complicated at all :)
+ Here is the routine.
+<code>
+       ldx #$00         ;init x index to 00
+loop1  ldy sort+1,x     ;load y index from linked list plus 1
+loop2  lda ypos,y       ;load a with y position for that index
+       ldy sort,x       ;load y index from linked list
+       cmp ypos,y       ;compare yposition(index+1,x) to yposition(index,x)
+       bcc swap         ;if yposition(index,x) is less than yposition(index+1,x) then swap the indexes
+       inx              ;if more than, increment index
+       cpx maxspr+1     ;checked all ypositions?
+       bne loop1        ;no
+       beq end          ;yes
+swap
+       lda sort+1,x     ;swap yposition(index+1,x) with yposition(index,x)
+       sta sort,x
+       sty sort+1,x
+       tay
+       dex
+       bpl loop2        ;if not first sprite loop to main
+       inx              ;correct index
+       beq loop1        ;restart sort from sprite(index+1,x)
+end
+</code>
+Quite simple isnt it?
+ Ok, now lets setup our irq chian, so we can see this bumch of sprites weve just arranged.
+<code>
+       jsr setirq     ;setup irq.
+</code>
+In runtime code
+<code>
+main    lda timer     ;wait for signal, that the buffer swap is complete.
+        sta mloop+1
+mloop   lda +
+        beq mloop
+        jsr movspr    ;move sprites
+        jsr anim      ;animate sprites
+        jsr color     ;animate colors
+cont    jmp main      ;loop
+setirq
+      sei            ;set interrupt disable
+      lda #$1b
+      sta $d011      ;raster irq to 1st half of screen.
+      lda #$fb
+      sta $d012      ;irq to happen at line #$fb
+      lda #<irq0
+      sta $fffe      ;hardware irq vector low byte
+      lda #>irq0
+      sta $ffff      ;hardware irq vector high byte
+      lda #$1f
+      sta $dc0d      ;turn off all types of cia irq/nmi.
+      sta $dd0d
+      lda #$01
+      sta $d01a      ;turn on raster irq.
+      lda #$35
+      sta $01        ;no basic or kernal
+      lda $dc0d      ;acknowledge any irq that has occured during setup.
+      lda $dd0d
+      inc $d019
+      cli            ;clear interrupt disable
+      rts            ;return from subroutine
+irq0
+      pha            ;use stack instead of zp to prevent bugs.
+      txa
+      pha
+      tya
+      pha
+      inc $d019      ;acknowledge irq
+      ldx #$03       ;wait a few cycles
+l1    dex
+      bpl
+      inx
+      stx $d015      ;sprites off = more raster time in top/bottom border
+slop  ldy sort+1,x   ;main index sort algo
+slep  lda ypos,y
+      ldy sort,x     ;this sorter uses the previous frame as a prediction buffer.
+      cmp ypos,y     ;as y position does not change much from frame to frame.
+      bcc swap       ;otherwise, it is a simple swap sort.
+      inx            ;our linked list (sort) is sorted in decending order, according
+      cpx #maxspr-1  ;to sprite y positions.
+      bne slop
+      beq end
+swap
+      lda sort+1,x
+      sta sort,x
+      sty sort+1,x
+      tay
+      dex
+      bpl slep
+      inx
+      beq slop
+end
+      ldy sort      ;re arrange frame buffers, into the raster buffers.
+      lda ypos,y    ;this is unrolled for speed.
+      sta ybuf      ;this allows us to use only 1 index pointer for our sprite plotter.
+      lda xpos,y    ;it is double buffered, to allow runtime code to calculate the sprite
+      sta xbuf      ;positions.
+      lda xmsb,y
+      sta mbuf
+      lda sprc,y
+      sta cbuf
+      lda sprp,y
+      sta pbuf
+      ldy sort+1
+      lda ypos,y
+      sta ybuf+1
+      lda xpos,y
+      sta xbuf+1
+      lda xmsb,y
+      sta mbuf+1
+      lda sprc,y
+      sta cbuf+1
+      lda sprp,y
+      sta pbuf+1
+</code>
+and so on until max sprites.
+<code>
+      ldx #$00     ;find # of used sprites (you can remove sprites by
+      stx sptr     ;placing #$ff into the ypos buffer for the corresponding
+maxc  lda ybuf,x   ;sprite. It will not be displayed by the raster routine.
+      cmp #$ff
+      beq mxs
+      inx
+      cpx maxspr
+      bne maxc
+maxs  stx cnt      ;max sprites this frame count.
+      cpx #$07     ;check if were doing more than 8
+      bcc maxm     ;if not, we want the plotter to stop after 1 irq.
+      ldx #$07
+maxm  stx mnt
+      lda #$ff    ;reset sprites to off screen.
+      sta $d001   ;prevents bugs.
+      sta $d003
+      sta $d005
+      sta $d007
+      sta $d009
+      sta $d00b
+      sta $d00d
+      sta $d00f
+      inc lsbtod   ;buffers are swapped, so we can do the next frame now.
+      lda #<irq1   ;irq chain for raster code. prolly want a routine before
+      sta $fffe    ;this one, to turn the sprites back on ;)
+      lda #>irq1   ;i.e. lda #$ff sta $d015
+      sta $ffff
+      lda #$28
+      sta $d012
+      jmp eirq
+</code>
+Since the buffers have been reordered into proper decending order, we can use unrolled loops for the sprite
+plotting. Each plot irq , is for each sprite in the order, 1,2,3,4,5,6,7,8 respectively.
+We have a counter (mnt) for 0-7 sprites, as we want to go ahead an plot the first 8 sprites all at
+once. We also have another counter (cnt) for the maximum number of sprites to display this frame.
+If (mnt) has not been reached yet, each irq branches to the next, to plot the first 8 sprites.
+After that, cnt is checked to see if were done plotting sprites. If not, each irq checks the postion of the
+sprite to see if it has finished displaying, and to see if its time to be done yet, if it is, it branches to the next irq.
+If it is not time yet, it calculates the next raster irq position and sets up the next irq.
+By using the previous sprites position, you can properly chain the irqs down the screen, for much less
+"overlapping" bugs.
+<code>
+irq1
+      pha           ;save registers
+      txa
+      pha
+      tya
+      pha
+      inc $d019     ;acknowledge irq
+      ldx sptr      ;get current sprite index
+hlop1 lda ybuf,x    ;get sprite y position
+      sta $d001     ;store sprite y postion.
+      lda xbuf,x    ;get sprite x position.
+      sta $d000     ;sta sprite x position.
+      lda mbuf,x    ;get sprite x position msb
+      bne no1       ;set msb register
+      lda $d010
+      ora #%00000001
+      bne yes1
+no1   lda $d010
+      and #%11111110
+yes1  sta $d010
+      lda pbuf,x    ;get sprite image pointer
+      sta $63f8     ;store it, double buffered screen.
+      sta $67f8
+      lda cbuf,x    ;get sprite color
+      sta $d027     ;store sprite color
+      inx           ;next sprite index
+      cpx mnt       ;lets go to next plot, if < then 8 yet.
+      bcc hlop2
+      cpx cnt       ;no more sprites?
+      bne ok1
+      jmp done      ;no more sprites.
+ok1   stx sptr      ;save sprite index
+      lda $d003     ;get last position of next sprite
+      clc
+      adc #$15      ;add 21 lines
+      cmp $d012     ;we there yet?
+      bcc hlop2     ;yeah, so plot next sprite
+      adc #$02      ;no, so calculate next irq position (+3)
+      sta $d012     ;set it
+      lda #<irq2    ;irq for next sprite.
+      sta $fffe
+      lda #>irq2
+      sta $ffff
+      jmp eirq
+irq2
+      pha           ;and so on
+      txa
+      pha
+      tya
+      pha
+      inc $d019
+      ldx sptr
+hlop2 lda ybuf,x
+      sta $d003
+      lda xbuf,x
+      sta $d002
+      lda mbuf,x
+      bne no2
+      lda $d010
+      ora #%00000010
+      bne yes2
+no2   lda $d010
+      and #%11111101
+yes2  sta $d010
+      lda pbuf,x
+      sta $63f9
+      sta $67f9
+      lda cbuf,x
+      sta $d028
+      inx
+      cpx mnt
+      bcc hlop3
+      cpx cnt
+      bne ok2
+      jmp done
+ok2   stx sptr
+      lda $d005
+      clc
+      adc #$15
+      cmp $d012
+      bcc hlop3
+      adc #$02
+      sta $d012
+      lda #<irq3
+      sta $fffe
+      lda #>irq3
+      sta $ffff
+      jmp eirq
+irq3
+      pha
+      txa
+      pha
+      tya
+      pha
+      inc $d019
+      ldx sptr
+hlop3 lda ybuf,x
+      sta $d005
+      lda xbuf,x
+      sta $d004
+      lda mbuf,x
+      bne no3
+      lda $d010
+      ora #%00000100
+      bne yes3
+no3   lda $d010
+      and #%11111011
+yes3  sta $d010
+      lda pbuf,x
+      sta $63fa
+      sta $67fa
+      lda cbuf,x
+      sta $d029
+      inx
+      cpx mnt
+      bcc hlop4
+      cpx cnt
+      bne ok3
+      jmp done
+ok3   stx sptr
+      lda $d007
+      clc
+      adc #$15
+      cmp $d012
+      bcc hlop4
+      adc #$02
+      sta $d012
+      lda #<irq4
+      sta $fffe
+      lda #>irq4
+      sta $ffff
+      jmp eirq
+irq4
+      pha
+      txa
+      pha
+      tya
+      pha
+      inc $d019
+      ldx sptr
+hlop4 lda ybuf,x
+      sta $d007
+      lda xbuf,x
+      sta $d006
+      lda mbuf,x
+      bne no4
+      lda $d010
+      ora #%00001000
+      bne yes4
+no4   lda $d010
+      and #%11110111
+yes4  sta $d010
+      lda pbuf,x
+      sta $63fb
+      sta $67fb
+      lda cbuf,x
+      sta $d02a
+      inx
+      cpx mnt
+      bcc hlop5
+      cpx cnt
+      bne ok4
+      jmp done
+ok4   stx sptr
+      lda $d009
+      clc
+      adc #$15
+      cmp $d012
+      bcc hlop5
+      adc #$02
+      sta $d012
+      lda #<irq5
+      sta $fffe
+      lda #>irq5
+      sta $ffff
+      jmp eirq
+irq5
+      pha
+      txa
+      pha
+      tya
+      pha
+      inc $d019
+      ldx sptr
+hlop5 lda ypos,x
+      sta $d009
+      lda xpos,x
+      sta $d008
+      lda mbuf,x
+      bne no5
+      lda $d010
+      ora #%00010000
+      bne yes5
+no5   lda $d010
+      and #%11101111
+yes5  sta $d010
+      lda pbuf,x
+      sta $63fc
+      sta $67fc
+      lda cbuf,x
+      sta $d02b
+      inx
+      cpx mnt
+      bcc hlop6
+      cpx cnt
+      bne ok5
+      jmp done
+ok5   stx sptr
+      lda $d00b
+      clc
+      adc #$15
+      cmp $d012
+      bcc hlop6
+      adc #$02
+      sta $d012
+      lda #<irq6
+      sta $fffe
+      lda #>irq6
+      sta $ffff
+      jmp eirq
+irq6
+      pha
+      txa
+      pha
+      tya
+      pha
+      inc $d019
+      ldx sptr
+hlop6 lda ybuf,x
+      sta $d00b
+      lda xbuf,x
+      sta $d00a
+      lda mbuf,x
+      bne no6
+      lda $d010
+      ora #%00100000
+      bne yes6
+no6   lda $d010
+      and #%11011111
+yes6  sta $d010
+      lda pbuf,x
+      sta $63fd
+      sta $67fd
+      lda cbuf,x
+      sta $d02c
+      inx
+      cpx mnt
+      bcc hlop7
+      cpx cnt
+      bne ok6
+      jmp done
+ok6   stx sptr
+      lda $d00d
+      clc
+      adc #$15
+      cmp $d012
+      bcc hlop7
+      adc #$02
+      sta $d012
+      lda #<irq7
+      sta $fffe
+      lda #>irq7
+      sta $ffff
+      jmp eirq
+irq7
+      pha
+      txa
+      pha
+      tya
+      pha
+      inc $d019
+      ldx sptr
+hlop7 lda ybuf,x
+      sta $d00d
+      lda xbuf,x
+      sta $d00c
+      lda mbuf,x
+      bne no7
+      lda $d010
+      ora #%01000000
+      bne yes7
+no7   lda $d010
+      and #%10111111
+yes7  sta $d010
+      lda pbuf,x
+      sta $63fe
+      sta $67fe
+      lda cbuf,x
+      sta $d02d
+      inx
+      cpx mnt
+      bcc hlop8
+      cpx cnt
+      bne ok7
+      jmp done
+ok7   stx sptr
+      lda $d00f
+      clc
+      adc #$15
+      cmp $d012
+      bcc hlop8
+      adc #$02
+      sta $d012
+      lda #<irq8
+      sta $fffe
+      lda #>irq8
+      sta $ffff
+      jmp eirq
+irq8
+      pha
+      txa
+      pha
+      tya
+      pha
+      inc $d019
+      ldx sptr
+hlop8 lda ybuf,x
+      sta $d00f
+      lda xbuf,x
+      sta $d00e
+      lda mbuf,x
+      bne no8
+      lda $d010
+      ora #%10000000
+      bne yes8
+no8   lda $d010
+      and #%01111111
+yes8  sta $d010
+      lda pbuf,x
+      sta $63ff
+      sta $67ff
+      lda cbuf,x
+      sta $d02e
+      inx
+      cpx mnt
+      bcc hlop9
+      cpx cnt
+      bne ok8
+      jmp done
+ok8   stx sptr
+      lda $d001
+      clc
+      adc #$15
+      cmp $d012
+      bcc hlop9
+      adc #$02
+      sta $d012
+      lda #<irq1
+      sta $fffe
+      lda #>irq1
+      sta $ffff
+      jmp eirq
+hlop9 jmp hlop1
+done  lda #<irq0
+      sta $fffe
+      lda #>irq0
+      sta $ffff
+      lda #$fb
+      sta $d012
+eirq
+      pla
+      tay
+      pla
+      tax
+      pla
+      rti
+END
+</code>