| |
Monte Carlos
Registered: Jun 2004
Posts: 351 |
Three pixel period pattern on upper/lower border
Any idea, how to create a hires pattern like this:
110110110110110110110110110110110110...
on the upper/lower border over the full width. Every trick allowed.
Btw: $3fff trick does not work, because then the pattern has a period of eight instead of three. Sprites cover only a width of 192 pixels!
Thank you Monte!
|
|
... 16 posts hidden. Click here to view all posts....
|
| |
Oswald
Registered: Apr 2002
Posts: 5017 |
looking at the above ascii art gave me an idea:
bb0aa0bb
0bb0aa0b
a0bb0aa0
aa0bb0aa
sprites are X stretched, aa is made of sprite '1', bb is made of sprite '2', 0 is background, change the X coords each line to get the tilt effect, and you're done!! :D
edit: oh, I do realize the above is not the wanted pattern :) ok how about then:
aa0bb0aa0bb
aa0bb0aa0bb
aa0bb0aa0bb
aa0bb0aa0bb
aa0bb0aa0bb
tech teching is not needed, even simpler, tho full width is not possible... :/
edit2: ok case SOLVED!!!!
aa is made of X stretched sprites, and bb is made of $3fff. the bb pattern is 8 bits: bb0000bb, and the aa pattern is 8 bits aswell aa0000aa, only the sprite X coords are used to push them in place.
I is teh win! :) |
| |
Hein
Registered: Apr 2004
Posts: 933 |
wow, was that your dream, Oswald? I'm affraid bb will shine through.. |
| |
WVL
Registered: Mar 2002
Posts: 886 |
oswald, what have you been using? :D
Think about your patterns again, and then draw how it should work for > 1 char please.. ;) |
| |
Oswald
Registered: Apr 2002
Posts: 5017 |
oh shit >1 char case is wrong indeed :D |
| |
JackAsser
Registered: Jun 2002
Posts: 1989 |
You can extend the correct pattern 7 or 8 lines into the lower border if you like. :) No further though. |
| |
Oswald
Registered: Apr 2002
Posts: 5017 |
12345678123456781234567812345678
aa0000aa0000aa0000aa0000aa
bb0000bb0000bb0000bb0000bb
12345678
aa0bb0aa0bb0aa0bb0aa0bb0aa0bb
needed patterns:
aa0000
bb0000
:/ |
| |
Monte Carlos
Registered: Jun 2004
Posts: 351 |
Oh wow, i'm really overwhelmed by all those replies.
Thought i have to repeat, that all the pixels are meant to be in a single rasterline. Anyway, a x size of 243 pixels would be enough instead of full width.
I make a short summary:
Skate: Use unexpanded hires sprites, leave 8 pixel gaps between sprites. -> The $3fff pattern is always black, the sprites are white or black. In any case, some pixels of the $3fff pattern must be anulled, some pixels in white must be added. As the sprite is single color, this does not work.
Monte Carlos: Use expanded single color sprites over $3fff -> If the sprites are white, the double pixels of the sprites cannot be half hidden by the $3fff raster at all positions.
If the sprites are black, no white pixels can be added -> failed
Oswald: Use shifted expanded single color sprites over $3fff pattern. Leave 4 pixel gaps and fill then in the middle -> The x size is not large enough. The intended $3fff pattern has a period of six instead of eight and is therefore not possible.
|
| |
Skate
Registered: Jul 2003
Posts: 490 |
@Monte Carlos: Thanks for the nice puzzle. You can have this pattern by using some kind of an interlace mode (toggling X position of expanded sprites by 3 pixel each frame). But that would look awful of course :) |
| |
Frantic
Registered: Mar 2003
Posts: 1627 |
I am not really a VIC coder, so I probably miss something, but couldn't you use unexpanded sprites and then 16 pixel gaps between the sprites, which shows two instances of the $3fff byte.
...and then you change the pattern of $3fff exactly at the cycle where the second $3fff byte in each pair of $3fff bytes would be shown, in order to make the first byte different from the second, in each pair. I think that would be each 5:th cycle (the sprite is 3 chars wide, and then there are 2 chars width of 3fff bytes = 5 chars = 5 cycles), so you should be able to do it in time. Yeah?
If so, then you would have 8*24 pixels from the sprite, and for each sprite there would also be 8*2*8 pixels from the 3ffff bytes:
(8*24)+(8*16) = 320 pixels
Perhaps someone said this already, or perhaps I did indeed miss something. I didn't really think hard now... |
| |
Skate
Registered: Jul 2003
Posts: 490 |
@Frantic: Your approach is the better version of mine but fails for the same reason. $3fff will be visible on sprite areas as well. So you cannot get that pattern correctly even if you can get the correct pattern for the gaps.
Btw, changing $3fff every 5th cycle is trickier than every 4th cycle.
lda #%11011011
ldx #%01101101
ldy #%10110110
sta $3fff-%01101101,x ; 5 cycles
stx $3fff ; 4 cycles
sty $3fff ; 4 cycles
It can be possible but harder. Illegal opcodes should help I guess. |
| Previous - 1 | 2 | 3 - Next |
