| |
MRT
Account closed
Registered: Sep 2005
Posts: 149 |
How to disable default charsets in bank 0 and 2
Is there any way to disable the default charsets in video bank 0 and 2?
A similar question would be... Are al interlaced fli images put in bank 1 and 3?
I mean... for a normal fullscreen fli image you allready need $2000 for the bitmap data and $400 * 8 = $2000 for all the fli color data. Totals $4000, which is exactly the size of one video bank.
So, if you would do this interlaced then you would need two of these video banks, because you have to switch banks every rasterline and flip banks every frame.
So, two complete video banks are needed... Isn't it possible to use... say, bank 1 and 2 ($4000-$7fff and $8000-$bfff)?
Or are all interlaced fli images put at $4000-$7fff and $c000-$ffff???
And if so... Can you still use the last two bytes (sprite pointer registers OR bitmapdata in case of fli) as IRQ vectors??
Questions, questions... so many questions, so few answers... ;-) |
|
... 20 posts hidden. Click here to view all posts....
|
| |
JackAsser
Registered: Jun 2002
Posts: 2014 |
@WVL: Exactly what Graham said... and what I also mentioned in my post in the timer interrupt thread. =)
|
| |
MRT
Account closed
Registered: Sep 2005
Posts: 149 |
Quote: @WVL: $FFFE/$FFFF is displayed if you do VSP/AGSP stuff :)
What is VSP/AGSP stuff???
Do you mean opening the borders? That would display the $ffff byte when the VIC bank is set to 3 |
| |
Oswald
Registered: Apr 2002
Posts: 5094 |
$fffe also can be displayed in a sprite
just my 2 smartass cents :D |
| |
WVL
Registered: Mar 2002
Posts: 902 |
Quote: $fffe also can be displayed in a sprite
just my 2 smartass cents :D
to be even an übernerd smartass : $ffff can also be displayed in a sprite if you use dirty spritecrunching techniques.. yuck :)
@ graham : ofcourse I was talking about a normal FLI screen :).. tell me about all those problems with sprite-pointers showing up as colors.. blllrrr!! that's even way more horrific than $fffe/$ffff problems..
btw, you can easily-peasily solve $fffe/$ffff by using $0314/$0315 :) |
| |
Copyfault
Registered: Dec 2001
Posts: 478 |
Quote: "If this works (it should!) we could do some quatrolace-FLI-Pics (and let our eyes bleed, yeah)."
Fullscreen??? And where would you put your code then??? ;-)
If we're speaking about a "normal" FLI (without any VSP/AGSP-whatsoever-action), you have $c0 bytes at the end of each bitmap - this should be enough for an FLI-routine, even if it utilizes all four banks. A little bit more complicated is to use the mem around $0000-$0200 as the stack and the data direction reg is mapped there...
But basically it works. Just set $DE00 to #$0b before the FLI starts and reset it to #$08 after it (or take care of the IRQ-vectors as set by the RetroReplay), and u can use all of the RAM for the VIC. Note that your FLI_rout must run in some mem_location < $1000, as this is the only available RAM during Ultimax. Should work on AR, too, though I didn't test it. |
| |
chatGPZ
Registered: Dec 2001
Posts: 11386 |
mmmh that said, the compo machine on BP will have a retro replay....go and write a demo that exploits that feature :=P |
| |
Ninja
Registered: Jan 2002
Posts: 411 |
@smartass WVL: $3f means always end of sprite, even when sprite-crunching, no? ;) |
| |
JackAsser
Registered: Jun 2002
Posts: 2014 |
@ninja: if byte $3e (starting with $00) is detected on the third byte on a sprite line, then it's considered complete and will finish, otherwise it will just continue ALWAYS (cheap ass sprite multiplexing).
Perhaps it was this you meant considering you above all people should know this, considering the part in Darwin. =) |
| |
Ninja
Registered: Jan 2002
Posts: 411 |
The problem with extreme smartassing is that you sometimes don't see the woods because of so many trees ;)
|
| |
WVL
Registered: Mar 2002
Posts: 902 |
Quote: The problem with extreme smartassing is that you sometimes don't see the woods because of so many trees ;)
:D |
| Previous - 1 | 2 | 3 - Next |
