|
|
Atari coding BBS
| Re: Rock-steady HBL |
|---|
|
Posted by: evil
|
Jul,19.2009-10:53
|
Just a small question: With HBL do you use plain HBL or Timer-B ?
If not Timer-B, change to that and I think it will be OK, Timer-B interupts a lot earlier than the HBL which is essential for stable rasters/screen setting.
Anyway a bit of optimizing could be:
move.b #$A,$8205.w ;high
move.b #$B,$8207.w ;mid
move.b #$C,$8209.w ;low
move.b #$D,$8265.w ;hscrol
move.l #next_routine_address,$120.w
rte
That is, you have 200 routines (if you have 200 scanlines..) and each routine changes the vector for next scanline. This way you avoid having counters and avoid saving registers. The VBL need to update the values of each scanline routine though (selfmod code).
It measures up to 108 cycles including rte. Which with 200 scanlines is about 13.5% of the VBL.
Please note that .w addressing saves up 4 cycles per move.
--
Anders Eriksson
ae@dhs.nu
|
[All messages in this thread] [Start new thread]
|
Topic
|
Posted by
|
Date
|
|
Rock-steady HBL
|
Per Almered (XiA)
|
Jul,19.2009-10:22
|
|
Re: Rock-steady HBL
|
evil
|
Jul,19.2009-10:53
|
|
Re: Rock-steady HBL
|
evil
|
Jul,19.2009-11:11
|
|
Re: Rock-steady HBL
|
Per Almered (XiA)
|
Jul,19.2009-11:25
|
|
Re: Rock-steady HBL
|
evil
|
Jul,19.2009-11:52
|
|
Re: Rock-steady HBL
|
Per Almered (XiA)
|
Jul,19.2009-11:59
|
|
Re: Rock-steady HBL
|
Kalms
|
Jul,19.2009-12:47
|
|
Re: Rock-steady HBL
|
Per Almered (XiA)
|
Jul,19.2009-12:59
|
|
Re: Rock-steady HBL
|
Per Almered (XiA)
|
Jul,19.2009-13:59
|
|
Re: Rock-steady HBL
|
ggn
|
Jul,21.2009-08:09
|
|
Re: Rock-steady HBL
|
Nicolas Pomarede
|
Jul,20.2009-00:33
|
What's the anti-troll code?
That's your personal code to be able to add comments and messages on the dhs.nu site.
Don't have a code or forgot it? Fix it here.
|
