WHat are some ways of making multi-colored sprites?

[Cybernoid]

I used the GPRIOR mode=0 technique in Jellybeans (http://atari.hobby-site.com/jellybeans). If you turn on the "overlapping players create third color mode" it is great. I believe that players 0 and 1 interact with playfield 0 and 1. Players 2 and 3 interact with playfield 2 and 3 the same way.

 

This mode is perfect for this type of puzzle drop game.

[atariksi]

It might be in De Re Atari.

 

Basketball was probably the first game to use a kernal to change PMG colours.

 

Interaction of colours depending on GPRIOR settings is described in the Hardware Manual.

 

Nothing thorough about all the GPRIOR combinations with various modes in any of the manuals I read. I had to find out by experiment that Player 5 replicated as overlay on GTIA mode 9 gives you a translucent effect (and 5-bit depth). [i guess Atari did not want competitors to know all their hidden secrets until many years later when the 8-bit computer wars were over and then they (or someone else) could show that they had the best 8-bit machine.]

[potatohead]

I think it's quite possible that Atari didn't know about all the cool stuff the chips would do. When you design stuff that incorporates software, interesting things happen over time. If Atari had let loose what they did have documented sooner, the hacks that get us the better displays would have happened sooner!

 

Swear there was a COMPUTE! article about the PM / Playfield interaction somewhere. I remember playing adventure on the VCS and understanding how the partially visible area in some rooms was done after reading it. The idea then was to make something visible with a player, and that implied the color interaction, but it was not stated. Don't remember if GTIA modes were ever mentioned at that time.

[ijor]

I agree that GPRIOR and player priority was, as best, sub-documented. AFAIK, priority on GTIA modes was never officially documented. Actually, I don't recall ever seeing any in-depth official GTIA (not CTIA) documentation. Closest thing was probably De Re Atari, which I guess it can be considered a "semi-official" documentation, but as I recall it doesn't cover priority on GTIA modes at all. We now have the GTIA datasheet, but it wasn't publicly available back then.

 

I think the sub-documentation of GPRIOR was just an oversight. Most other aspects were documented in the technical reference much better. I see no intentional reason for not doing the same with GPRIOR. But of course that the hardware manual doesn't cover GTIA at all, only CTIA.

[emkay]

This mode is perfect for this type of puzzle drop game.

 

And for some "128" colour version of Zak McKracken or else like below:

 

Really everything was possible there, because we don't need "50" frames per second.

Edited December 30, 2008 by emkay

[Heaven/TQA]

where is the GTIA doc mentioned here???

[Tezz]

I'm sure GPRIOR must have been documented somewhere? maybe not in the official Atari publications but it must have been in others. I've certainly used the overlap beween players and playfield registers back in the day so I must have learnt it from one of the regular sources? You've mentioned Mapping and De Re how about the technical reference notes? I guess if not it must have been covered in magazine articles back then.

[ijor]

I'm sure GPRIOR must have been documented somewhere? maybe not in the official Atari publications but it must have been in others. I've certainly used the overlap beween players and playfield registers back in the day so I must have learnt it from one of the regular sources? You've mentioned Mapping and De Re how about the technical reference notes? I guess if not it must have been covered in magazine articles back then.

 

I didn't say that GPRIOR was completely undocumented. I said that it wasn't fully documented, and that priority on GTIA modes was (AFAIK) completely undocumented. Documented/undocumented always refer, of course, to official documentation.

 

I mentioned the tech references. It is precisely there where it is not 100% completely documented, and it is precisely there where GTIA modes is not even mentioned in any way whatsoever. I didn't mention Mapping at all, in first place because it is not an official reference (it is an invaluable one, but not official), and GPRIOR covering on the Mapping is not precisely its strongest one.

 

where is the GTIA doc mentioned here???

 

Which one?

[Tezz]

... I said that it wasn't fully documented, and that priority on GTIA modes was (AFAIK) completely undocumented. Documented/undocumented always refer, of course, to official documentation.

Sorry yes I was referring to all documentation official or otherwise. Yes you are right that it wasn't documented by Atari officially. Although Atari documentation for the machine period was virtually non existant in the early years and their support later down the line in respect of technical documents and throughout it's commercial life was quite poor overall. I think in fact this has been discussed previously that Atari kept the technical details to themselves for a long time which of course held back 3rd party software developers struggling for information about the machine. I always wondered if those technical reference notes were just hashed out rough by Atari? The one I bought back then was supplied photocopied in a ring binder.

 

Luckily there were books and magazine publications around helping to fill in the gaps although it's poor that much wasn't officially documented.

Edited December 30, 2008 by Tezz

[Cybernoid]

And for some "128" colour version of Zak McKracken or else like below:

 

 

Yeah, that is definitely cool...

[MaPa]

IMHO the GPRIOR mixing PMGs with playfield is just case of undefined state of GTIA chip.. something like undocumented 6502 instruction. It does something, but it was not intended, just simple behaviour of tanzistors layout and connection inside the chip. I didn't know about that GPRIOR 0 feature till this year (ok, I never knew much about atari, but never read about this).

Edited December 30, 2008 by MaPa

[+Philsan]

This mode is perfect for this type of puzzle drop game.

And for some "128" colour version of Zak McKracken or else like below:

[ijor]

I always wondered if those technical reference notes were just hashed out rough by Atari? The one I bought back then was supplied photocopied in a ring binder. Luckily there were books and magazine publications around helping to fill in the gaps although it's poor that much wasn't officially documented.

 

I didn't mean that "that much" was undocumented. I think the technical reference is extremely well done and comprehensive. There are a few missing links, but not too much. The GTIA missing is probably because the hardware manual was written before it was released (possibly, even before it was ready). I know Atari published it much later, but my guess is that it was available since day one for "selected" developers under NDA.

 

Btw, it doesn't have the highest quality printing, but they were not photocopies, and as I recall it wasn't sold in a binder.

 

Yes, third party docs were great. But still regarding this GPRIOR topic, I'm not sure it was ever fully described. I can't rule out there was some magazine's article that I didn't see, but I don't recall ever reading one (again, in all the glory details as we know now).

 

IMHO the GPRIOR mixing PMGs with playfield is just case of undefined state of GTIA chip.. something like undocumented 6502 instruction. It does something, but it was not intended, just simple behaviour of tanzistors layout and connection inside the chip. I didn't know about that GPRIOR 0 feature till this year (ok, I never knew much about atari, but never read about this).

 

I'm not sure they are comparable. Yes, conceivable this feature was not by design. But one thing is "not-by-design", another one is undocumented. There are several aspects of the A8 that obviously are not by design (such as the Pokey divisor formula), yet they are fully documented.

 

And even from the point of view of the hardware design, I doubt this is an undefined state as the illegal 6502 opcodes. Possibly the exact behavior wasn't in the original blackboard design, but I suspect it was well know and verified at later stages of the chip design. This is very different from the 6502 "undefined" states, where probably even Peddle himself (6502 designer) wasn't fully aware about what exactly they do.

[atariksi]

IMHO the GPRIOR mixing PMGs with playfield is just case of undefined state of GTIA chip.. something like undocumented 6502 instruction. It does something, but it was not intended, just simple behaviour of tanzistors layout and connection inside the chip. I didn't know about that GPRIOR 0 feature till this year (ok, I never knew much about atari, but never read about this).

 

It's highly organized where PF0/PF1 OR with P0/P1 and PF2/PF3 OR with P2/P3 and unlike 6502 where different versions exist, the GTIA is consistent for all it's chips so one can optimize knowing it will work with all GTIA chips.

[atariksi]

I agree that GPRIOR and player priority was, as best, sub-documented. AFAIK, priority on GTIA modes was never officially documented. Actually, I don't recall ever seeing any in-depth official GTIA (not CTIA) documentation. Closest thing was probably De Re Atari, which I guess it can be considered a "semi-official" documentation, but as I recall it doesn't cover priority on GTIA modes at all. We now have the GTIA datasheet, but it wasn't publicly available back then.

 

I think the sub-documentation of GPRIOR was just an oversight. Most other aspects were documented in the technical reference much better. I see no intentional reason for not doing the same with GPRIOR. But of course that the hardware manual doesn't cover GTIA at all, only CTIA.

 

There's also this talk about BLACK on conflicting priorities which seems to be misdocumented in places.

[MaPa]

IMHO the GPRIOR mixing PMGs with playfield is just case of undefined state of GTIA chip.. something like undocumented 6502 instruction. It does something, but it was not intended, just simple behaviour of tanzistors layout and connection inside the chip. I didn't know about that GPRIOR 0 feature till this year (ok, I never knew much about atari, but never read about this).

 

It's highly organized where PF0/PF1 OR with P0/P1 and PF2/PF3 OR with P2/P3 and unlike 6502 where different versions exist, the GTIA is consistent for all it's chips so one can optimize knowing it will work with all GTIA chips.

 

It seems organized, but it doesn't prove it's intended. And P2/P3/PF2/PF3 are below P0/P1 and if 5th player is set than it's a little strange. It's below P0/P1 but above all playfields and mixed with P2/P3, but only with P2/P3 even if there is already P2/P3 mixed with PF2/PF3. The playfield has no effect on final color.

Edited December 30, 2008 by MaPa

[atariksi]

IMHO the GPRIOR mixing PMGs with playfield is just case of undefined state of GTIA chip.. something like undocumented 6502 instruction. It does something, but it was not intended, just simple behaviour of tanzistors layout and connection inside the chip. I didn't know about that GPRIOR 0 feature till this year (ok, I never knew much about atari, but never read about this).

 

It's highly organized where PF0/PF1 OR with P0/P1 and PF2/PF3 OR with P2/P3 and unlike 6502 where different versions exist, the GTIA is consistent for all it's chips so one can optimize knowing it will work with all GTIA chips.

 

It seems organized, but it doesn't prove it's intended. And P2/P3/PF2/PF3 are below P0/P1 and if 5th player is set than it's a little strange. It's below P0/P1 but above all playfields and mixed with P2/P3, but only with P2/P3 even if there is already P2/P3 mixed with PF2/PF3. The playfield has no effect on final color.

 

In general when there's an organized consistent way it works, there's some intent; otherwise, you could speculate that even things that are intentional may not be. Player #5 mapping to PF3 definitely looks intentional (perhaps did not want to add additional color registers) and then the P2/P3 ORing with PF3 (5ht player) follows correctly as it should (PF3 is highest playfield). It's hard to make a statement that mixing PMGs w/PF is an undefined state when it's just as organized as other priorities and features.

[ijor]

It seems organized, but it doesn't prove it's intended. And P2/P3/PF2/PF3 are below P0/P1 and if 5th player is set than it's a little strange. It's below P0/P1 but above all playfields and mixed with P2/P3, but only with P2/P3 even if there is already P2/P3 mixed with PF2/PF3. The playfield has no effect on final color.

 

The 5th player breaks the symmetry somewhat, so it would always be slightly different no matter what. But it is not that strange, you are just looking at this with a strange interpretation. The behavior is surely by design, and it is actually as following:

 

PF3 always (disregarding any GPRIOR setting) has priority above all other playfields. Point, end of the story regarding the hardware implementation.

 

Obviously priority among playfields could only happen with 5th player, which in turn depends on a GPRIOR bit. And as atariski is saying, P5 then follows the natural PF3 mixing/oring. PF2 would be masked out if PF3 (5th player) is present, and that's why it won't mix.

[danwinslow]

Wow, over my head here, not that thats hard.

What do you guys mean by 'playfield' as in pf2/3, etc? Do you mean the playfield colors..IIRC as in set by COLOR/SETCOLOR statements in basic? What is player 5? is that PMG missile or the 'inverse' fifth color in ANTIC 4/5?

 

The 'or-ing' thats being talked about is occuring between a plotted playfied dot and a PMG color at the same screen location, right?

[Rybags]

Playfield = playfield colours. For the 4-colour modes, PF0= color 1, PF1=color 2 etc.

 

Player 5 is the missiles if you set "5th player enable" in GPRIOR. It uses the PF3 colour.

 

PF3 isn't seen in most modes, it's only available in OS modes 1, 2, 10, 12 and 13.

 

The ORing occurs among the resultant colours that would go to the screeen.

 

So, e.g. Playfield 2 colour = $94 (default) ORed with Player 2 colour set to $38 (bright red), the result would be $BC (bright green).

Edited December 31, 2008 by Rybags

[+CharlieChaplin]

This mode is perfect for this type of puzzle drop game.

 

And for some "128" colour version of Zak McKracken or else like below:

 

Really everything was possible there, because we don't need "50" frames per second.

 

Well,

looks like a mockup (or a pic from another computer), not like a real A8 picture...

-Andreas Koch.

[emkay]

Well,

looks like a mockup (or a pic from another computer), not like a real A8 picture...

-Andreas Koch.

 

Ofcourse it is a mockup.

Or do you know a graphics tool that is allowing full GRPIOR features?

Give one to me and you see even better pictures on the Atari.

[atariksi]

Well,

looks like a mockup (or a pic from another computer), not like a real A8 picture...

-Andreas Koch.

 

Ofcourse it is a mockup.

Or do you know a graphics tool that is allowing full GRPIOR features?

Give one to me and you see even better pictures on the Atari.

 

So was the image viewable on Atari or from another computer even at 25/30Hz? It seems CharlieChaplin is claiming it's from another machine.