Jump to content
IGNORED

Atari v Commodore


stevelanc

Recommended Posts

It's a bit of a semantic issue; some people consider a multiplexor to need a sort, others consider any recycling of sprites to be multiplexing and there's loads of grey area inbetween...

Multiplexing is simply reusing an information channel. If PMs are reused, then that's multiplexing. I know that in the UK there is this "Andrew Braybrook definition of sprite multiplexing" which includes sorting

 

It wasn't really Braybrook who popularised the term, he was quite late in since Morpheus was his first go at a sorting system and during the Zzap! diary he was getting advice from Gary Liddon and others so presumably the term came with that. The way it was originally explained to me was that if it makes an "intelligent" attempt to juggle the actual hardware sprites around behind the scenes (that would include pre-sorted routines) so that the objects onscreen can move seamlessly from top to bottom then it's a multiplex, if some or even all of the objects are stuck within two vertical co-ordinates it's not... but like i said, y'mileage varies considerably.

 

but then again: Where to draw the line? You can have different sorting methods too, up to pre-sorted sprites which would also apply to Mirax Force.

 

i've got the same kind of splitting system in my current C64 project Co-Axis 2189, but the people i knew back in the olde days always referred to it as "zone splitting" (since the sprites are divided into zones) or occasionally sprite recycling rather than multiplexing and it sort of feels like cheating using that term for the system i've got...

Link to comment
Share on other sites

A comment to Mirax Force...

 

http://atari.fandal.cz/detail.php?files_id=736

 

 

In this demo you see full working 8 coloured hardware "Sprites" moving freely. You also see the tripple screen mode. If you know how it is done, you know that this trick doesn't leave much cpu power in that range.

Now imagine Mirax Force using one Software Sprite for the Player's ship and such multiplexing for the enemies.

A software routine for the given LMS scrolling background is not that complex and should be possible with 50fps as the Sprite routine already is. The Missiles are fully available for the shots...

 

So, please, no one tell me again, a game like Katakis is not possible on the Atari.

Link to comment
Share on other sites

Yes the TI did, but I think it was the one with the multiplexer in ROM.

 

Hey, all this thread made me think of something. Was reading it, and was also playing Final Fantasy XII on my PS2. We've got a 16:9 screen now. FF XII isn't one of the few PS2 HI DEF games, but it does do an interesting 16:9 mode. Basically, that mode uses almost a full NTSC frame, and squeezes all the graphics in the X direction, so that the user can hit the "fill" option for their display, and see things with the correct aspect ratio.

 

Seems to me, on the Atari, the 48 byte DMA could be used to a similar effect, providing for a cool game option on modern displays. When the 16:9 option is invoked, an alternate set of game graphics and maybe the extended DMA come into play. The extra time consumed would limit some special effects, but so what? Playing something cool, like Star Raiders, in the 16:9 would be totally worth it.

 

Keeping on topic to the thread, that's a clear differentiator in favor of the Atari, right?

 

Can either of these machines produce an interlaced display, BTW? Just wondering.

Link to comment
Share on other sites

In this demo you see full working 8 coloured hardware "Sprites" moving freely. You also see the tripple screen mode. If you know how it is done, you know that this trick doesn't leave much cpu power in that range.

 

That routine relies absolutely on the vertical positions of the various sprites always being fixed in relation to each other; that means that although the horizontal positions don't have to follow a sine curve the vertical ones are hardwired. Granted you could do a game using that but it'd never move in the way Katakis does although it could possibly look something like one of the C64 budget shoot 'em ups such as Tangent.

 

Now imagine Mirax Force using one Software Sprite for the Player's ship and such multiplexing for the enemies.

A software routine for the given LMS scrolling background is not that complex and should be possible with 50fps as the Sprite routine already is. The Missiles are fully available for the shots...

 

You've almost described what i'm half working on at the moment... but it's not locked down like the demo routine is. =-)

Link to comment
Share on other sites

(the c64 could have just as split gfxmodes, we have colors instead)

 

 

With that trick the A8 can show up to 88 colours per scanline. But this is not the point. And I really don't understand why you start comparing things again. No one was talking about C64 here.

 

It was about the usage of the PM graphics and that it is possible to have enough "sprites" to handle a fast and good looking sidescrolling game.

 

People always wrote games on the ATARI starting to put all PM Graphics together for the "Hero-Sprite"and to use several software sprites....

The better solution was to use one software sprite for the "hero" . So the PM was available for flexible attack wave programming.

A highly achieved coder could do additional parallax scrolling aswell...

Link to comment
Share on other sites

Seems to me, on the Atari, the 48 byte DMA could be used to a similar effect, providing for a cool game option on modern displays. When the 16:9 option is invoked, an alternate set of game graphics and maybe the extended DMA come into play.

 

At the resolutions both machines run at i doubt it'd be possible to come up with meaningfully different graphics sets for the two ratios... certainly it's not something i'd like to try. i'm not sure the extended DMA can be used in the way you described without a hardware mod to actually send it out, i know some extra width is sent but if it's the entire 48 bytes i'd be surprised... anyone who has actually tried it feel free to jump in though. =-)

 

The extra time consumed would limit some special effects, but so what?

 

Seriously... you'd be surprised what a difference it can make!

 

Can either of these machines produce an interlaced display, BTW? Just wondering.

 

Assuming you mean software interlace then yes (in the spirit of the thread, i won't mention that the C64 is better at horizontal interlace... oops. =-)

Link to comment
Share on other sites

It does send the whole 48 bytes. I did this as a kid. Had an old color CRT TV, that I performed a CRT swap on. Wasn't quite the right CRT for the thing, and after adjustments, I settled on seeing the entire NTSC frame, centered on the thing.

 

(TV Broadcasts have lots of stuff in the overscan, and the Atari has some noise in the first pixel or two, depending on what you ask it to do.)

 

It would only take a pixel or two smaller game graphic to make the difference from what I saw done on the PS2.

 

The DMA modes are 32, 40 and 48 bytes, where the 48, is nearly the whole frame, for a 384 pixel hi-res, 192 pixel standard res screen.

 

Re: Special Effects -- Of course they make a difference! No question. Still could split the screen, and with a blank scan or two, do stuff with the sprites, and change color sets. It's a much smaller CPU window. If you are looking for the most colors and such, this wouldn't be the way to go. On the other hand, if you are looking to fill the frame, and do game graphics everywhere, then it would! My point being, that's an Atari thing more than a C64 thing. Didn't matter much until recently where we have enough display devices out there to actually leverage it. Run some video through your PC capture card to see it all play out.

 

I'm thinking of a game like Star Raiders, for example. It's simple, classy, and the GFX would translate well to the wider aspect, and fill the frame on newer display devices, without having to use the ZOOM feature, and lose the upper and lower areas of the frame, vertically.

 

Re: Interlace. Yep. The C64 can do that well. For what it's worth, if you send the composite signal of an Atari into the Luma input of an S-Video, or Component video input, you can do a very nice 640 x 200 monochrome screen. The trick is to use opposite colors, a DLI, and two bitmaps. Addressing is goofy, but it all works nicely enough. Did that too, on the older Apple ][ monochrome monitors, hoping to complete a real 8 bits per char 80 col display. Never did.. Plotted some damn nice graphs though.

 

What I was wondering specifically, was vertical interlace. I don't know if either machine can output the half-scan necessary to do this.

Edited by potatohead
Link to comment
Share on other sites

Oh, and no offense here. I like both machines a lot, and for different reasons. When I see one of these monster threads, I always learn something new, and that's worth the rest.

 

We really should just get along :) 8 bits is 8 bits dammit!

 

For me, the Atari systems engineering was really a stand out. Device independent I/O, the whole SIO interface, lots of game I/O, paddle, light pen, etc... Excellent!

 

The sound on the C64 is great, and clearly it does better characters with the sprites. There remains cool tricks for both though!

 

I think it would be fun to see everything swapped for a year. Make all the Atari Alphas work on the C64, and the reverse and see what productions come of that.

Link to comment
Share on other sites

Keep in mind that the available 32,40,48 byte wide screen options provide increasingly lower available cycles for Antic to interupt the display so it's not quite as simple as changing the display to 48 wide. Unless what you are changing is very simple the Antic timing will need to be redone and anything pushing it will not be possible on a fully wide 48 byte playfield.

 

32 bytes wide gives us the most options in terms of overcoming hardware limitations

Edited by Tezz
Link to comment
Share on other sites

Again, it depends on WHICH limitations, doesn't it?

 

If you want to put meaningful graphics into most of the NTSC frame, then the wide DMA is the way to go. Hell, with no real effort, you get a 48 char text display, or nice wide, graphics field to then do sprites on.

 

That's outta the box, just asking for the right modes via the DMA control register, and a display list that sets up the display you want.

 

On newer display devices, the largely unused overscan doesn't have to really go unused. Classic Atari flexibility tradeoff. If you really want to flex the hardware for lots of colors, mode changes, sprite re-use, multiplexing, etc... then you do it @ 32 bytes. Split the middle at 40, and go the other way @ 48.

 

This is a hallmark of the Jay Miner design mindset. I really like it. It's an Atari thing.

Link to comment
Share on other sites

9 - FLAK

 

post-6191-1228070298_thumb.png

 

I'd forgotten about Flak - it totally blew Xevious out of the water ( on the 8bit/5200 at least )

 

"Blew Xevious out of the water"? Yea, I guess so. Although, that wasn't very hard to do. What happened to the flying enemies in Flak though?

 

Xevious should have been done a lot better. Shame on you Atari.

Link to comment
Share on other sites

That's interesting.. I guess you have two sets of multicolour players

p0 lum 4 , p1 lum 8 , p2 lum 6 ,p3 lum 10 ( and BAK lum 2 )

 

then the gradient is bak , p0 , p2 , p1 , p3, p0|p1(12) , p2|p3 (14)

 

pf0 - pf3 can be red , rainbow, black,white - total of 12 colours per scanline

 

.... or is it some other method?

Edited by Crazyace
Link to comment
Share on other sites

That's interesting.. I guess you have two sets of multicolour players

p0 lum 4 , p1 lum 8 , p2 lum 6 ,p3 lum 10 ( and BAK lum 2 )

 

then the gradient is bak , p0 , p2 , p1 , p3, p0|p1(12) , p2|p3 (14)

 

pf0 - pf3 can be red , rainbow, black,white - total of 12 colours per scanline

 

.... or is it some other method?

 

Hehe.... If you're familiar with the Atari and the graphics modes, you see it on the first sight. The pink colour reminds of the 5th colour register that comes into the bitmap graphics when switching the GTIA mode at the same scanline...

Link to comment
Share on other sites

Re: Special Effects -- Of course they make a difference!

 

We-ell, i said that it makes a difference meaning that going up to 48 bytes wide nicks a surprising amount of CPU grind - or at least it surprised the hell out of me a few weeks ago; changing up broke something i was working on recently because it needed to do too many splits per scanline for the ones where the DMA was fetching.

 

What I was wondering specifically, was vertical interlace. I don't know if either machine can output the half-scan necessary to do this.

 

No, not as far as i'm aware - the C128 might be able to, it's got a few interesting quirks in there but i don't know my way around the differences well enough.

 

Oh, and no offense here. I like both machines a lot, and for different reasons. When I see one of these monster threads, I always learn something new, and that's worth the rest.

 

We really should just get along :) 8 bits is 8 bits dammit!

 

Well, i'll play with just about anything that has a 6502 variant... one day i'll have another go at learning Z80 to play with more stuff too! =-)

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...