Smooth Linecrunching
(note: this article has a slight bug in that each crunched line scrolls the rest of the screen up by 7 pixels, not 8, so the example code actually scrolls the screen up by 7/8ths of the value read out of the sine table. Still works well enough for demonstration purposes, but something to keep in mind if you want to use it as the starting point for an AGSP routine or anything else that requires precise positioning).
Basically linecrunching is a VIC-trick that allows you to easily move graphics vertically on the screen. It is especially useful for bitmaps, where it would take far too much time to move it by copying.
On a more technical level, linecrunching is taking a full char-line of graphics and compressing it into a single raster-line. This is done by changing the YSCROLL value in $d011 on a badline, but before the VIC starts stealing CPU cycles (cycle 14 to be exact), to make the badline-condition no longer true. Simplified, this basically makes the VIC start reading from the next char-line.
If you continuously do this for several badlines, lines will continuously be crunched, and the VIC will eventually loop back to reading the start of the screen. It also can read the normally invisible last $18 bytes of the screen data. This causes the VIC to actually wrap to a little bit past the beginning of the screen data, causing a weird split-effect…
The problem with simply linecrunching is that it's pretty rough. If you tried scrolling it would look quite jumpy and just generally bad. We need to find some way to make the scrolling SMOOTH, but how??
Before rasterline $30, you must change the YSCROLL through $d011 to the soft-scroll value you want to use. But doing that will also change the line the first badline occurs on. We must account for this, and start linecrunching on the correct rasterline, via IRQ or some waitloop.
Now, getting the hardscroll value is easy, just divide by 8. This value is also how many char-lines to crunch.
Now there is one last problem- it looks smoother, but it is still quite jumpy… let's explain why. The greater the amount of crunched char-lines is, the more the screen scrolls up. BUT the greater the YSCROLL value is, the more the screen scrolls DOWN. Easy fix, all you have to do is flip the YSCROLL value.
All of the lines you crunched “build up” at the top of the screen, which can look quite bad… in the code example I used an illegal gfx mode to “cover” it up.
Finally, here is the code. 64tass syntax.
sin_index = $fb char_lines_to_crunch = $fc initial_yscroll = $fd * = $0801 .word $080b,0 ;basic stub .byte $9e,$32,$30,$36,$31 .byte 0,0,0 sei lda #$35 sta 1 lda #$1b sta $d011 lda #<nmi sta $fffa lda #<irqa sta $fffe lda #>irqa sta $fffb sta $ffff lda #$7f sta $dc0d sta $dd0d lda $dc0d lda $dd0d lda #1 sta $d01a ldx #$17 ;get rid of ugly post-screen chars lda #$20 - sta $07e8,x dex bpl - inc $d019 cli mainloop lda $d011 ;wait for new frame bpl *-3 lda $d011 bmi *-3 inc sin_index ldy sin_index ldx sintbl,y txa lsr lsr lsr sta char_lines_to_crunch txa and #7 eor #7 ;linecrunch scrolls up, but increasing the yscroll scrolls down ;therefore we must flip the yscroll to make it match sta initial_yscroll clc adc #$30-3 ;before linecrunching takes 3 rasterlines sta $d012 lda initial_yscroll ora #$50 sta $d011 lda #$18 ;use the invalid textmode to "cover up" the linecrunch bug area sta $d016 jmp mainloop ;stable raster routine. not exactly necessary for linecrunch, but might as well :) irqa tsx lda #<irqb sta $fffe inc $d012 inc $d019 cli .fill 40,$ea irqb txs ldx #8 dex bne *-1 .byte 4,$ea ;nop $ea lda $d012 cmp $d012 ;last cycle of CMP reads data from $d012 beq + ;add extra cycle if still the same line + ldx #9 dex bne *-1 ;this is the part that actually performs the linecrunch clc lda initial_yscroll ldx char_lines_to_crunch crunchloop adc #1 and #7 ora #$50 sta $d011 ldy #8 ;notice that there is lots of time left in the rasterline for more effects! dey bne *-1 nop nop .byte 4,$ea ;nop $ea dex bpl crunchloop and #%10111111 ;disable invalid gfx mode sta $d011 lda #8 sta $d016 lda #<irqa sta $fffe inc $d019 nmi rti ;precalculated sine-table .align $100 sintbl .byte 100,102,105,107,110,112,115,117,120,122,124,127,129,131,134,136,138,141,143,145,147,149,151,153,156,158,160,162,163,165,167,169,171,172,174,176,177,179,180,182,183,184,186,187,188,189,190,191,192,193,194,195,196,196,197,198,198,199,199,199,200,200,200,200,200,200,200,200,200,199,199,199,198,198,197,196,196,195,194,193,192,191,190,189,188,187,186,184,183,182,180,179,177,176,174,172,171,169,167,165,163,162,160,158,156,153,151,149,147,145,143,141,138,136,134,131,129,127,124,122,120,117,115,112,110,107,105,102,100,98,95,93,90,88,85,83,80,78,76,73,71,69,66,64,62,59,57,55,53,51,49,47,44,42,40,38,37,35,33,31,29,28,26,24,23,21,20,18,17,16,14,13,12,11,10,9,8,7,6,5,4,4,3,2,2,1,1,1,0,0,0,0,0,0,0,0,0,1,1,1,2,2,3,4,4,5,6,7,8,9,10,11,12,13,14,16,17,18,20,21,23,24,26,28,29,31,33,35,37,38,40,42,44,47,49,51,53,55,57,59,62,64,66,69,71,73,76,78,80,83,85,88,90,93,95,98