Priority_0 explanation

[José Pereira]

Hi.

Because of @rensoup talkings on the Sub Hunter thread:

 

Priority_0 gives OR(ings) on PMGs over certain PFs colours so works like this (using charmodes Antic4 and/or 5):

-> PM0 over BAK = PM0 colour;

       "      "    PF0 = PM0 OR PF0 colour;

       "      "    PF1 = PM1 OR PF1 colour;

-> PM1 over BAK = PM1 colour;

       "      "    PF0 = PM1 OR PF0 colour;

       "      "    PF1 = PM1 OR PF1 colour;

-> PM2 over BAK = PM2 colour;

       "      "    PF2 = PM2 OR PF2 colour;

       "      "    PF3 = PM2 OR PF3 colour;

-> PM3 over BAK = PM3 colour;

       "      "    PF2 = PM3 OR PF2 colour;

       "      "    PF3 = PM3 OR PF3 colour;

This way you can have this 12colours more the 5 (BAK/PF0/PF1/PF2/PF3) makes 17colours per scanline possible.

 

If you take of the M (Missiles off) and use only P (Players) is still the same but the mIssiles acts as the 'so called' 5th Player and take PF3 colour. This way the Missiles themselves doesn't OR anything but if P2 or P3 goes over them it'll OR them because they're be considered as PF3 colour.

 

On Mode D and E bitmap modes this same thing happens but you don't have PF3 but you can still OR it if you have the Missiles as 5th Player and P2/P3 OR them.

 

Also if you go to PMGs multicolour mode that is PM0 OR PM1 and PM2 OR PM3 to above you'll have to add:

-> PM0 OR PM1 over BAK: PM0 OR PM1 colour;

             "               "   PF0: (PM0 OR PM1) OR PF0 colour;

             "               "   PF1: (PM0 OR PM1) OR PF1 colour;

-> PM2 OR PM3 over BAK: PM2 OR PM3 colour;

             "               "   PF2: (PM2 OR PM3) OR PF2 colour;

             "               "   PF3: (PM2 OR PM3) OR PF3 colour;

This way to 17colours above we'll add more 6colours and that is why you may have read that A8 can display maximum of 23colours per scanline.

 

I hope this helps...

?

 

 

 

Edited August 19, 2019 by José Pereira
Title mispelling

[popmilo]

Yeah, simple

 

[José Pereira]

Just now, popmilo said:

Yeah, simple

 

Clear as water...

?

[rensoup]

Great explanation! Must be fun designing around those ?

[_The Doctor__]

It's not that your designing around those, it's that your designing with or for those.

[ivop]

6 hours ago, José Pereira said:

but the mIssiles acts as the 'so called' 5th Player

I admit I'm too lazy to look it up myself, but if the 5th player is activated, do you still need to set 4 x-positions, or does one (or each?) of the xpos registers control the whole player and have the three other missiles offset accordingly?

[Gibstov]

2 hours ago, ivop said:

I admit I'm too lazy to look it up myself, but if the 5th player is activated, do you still need to set 4 x-positions, or does one (or each?) of the xpos registers control the whole player and have the three other missiles offset accordingly?

You still have to set the x-position independently.  The only thing that combining into a 5th player gets you, is that the missiles all share the same color (PF3) instead of the color of their respective player.

[rensoup]

11 hours ago, _The Doctor__ said:

It's not that your designing around those, it's that your designing with or for those.

Thanks for the english lesson ?

 

(not a native english speaker)

 

7 hours ago, Gibstov said:

You still have to set the x-position independently.  The only thing that combining into a 5th player gets you, is that the missiles all share the same color (PF3) instead of the color of their respective player.

God damnit... I didn't even know that. That's ridiculous. Why couldn't they just give us 8 players ?

 

Does anyone know the actual reason ? Was it 8 originally and they just decided to cut the transistor count ?

[+Stephen]

9 hours ago, rensoup said:

Thanks for the english lesson ?

 

(not a native english speaker)

 

God damnit... I didn't even know that. That's ridiculous. Why couldn't they just give us 8 players ?

 

Does anyone know the actual reason ? Was it 8 originally and they just decided to cut the transistor count ?

I don't think it was ever meant to be 8, but yes, I am sure transistor count was the reason.  IMHO, an even worse decision, again surely due to transistor count, is that we have GTIA mode 10 which could do 16 colours, but the machine only has 9 colour registers.

[José Pereira]

13 minutes ago, Stephen said:

I don't think it was ever meant to be 8, but yes, I am sure transistor count was the reason.  IMHO, an even worse decision, again surely due to transistor count, is that we have GTIA mode 10 which could do 16 colours, but the machine only has 9 colour registers.

Probably a mode like GR.1 and 2 that having 64chars distributed by PF0->PF3 could also had GR.12/13 (Antic4/5) also with 64chars and BAK/PF0/PF1 common then PF2/PF3/PF4/PF5 that'll be 7 + 8colours of 8PMGs makes 15 and an independant Border colour on the 2:1 ratio modes (that A8 misses and other machines not) and 'et voilá' 16colours registers.

Probably 64chars aren't great (even 128chars aren't) but with multiple charsets DLIs it would be workable and I'll prefer than like is having less colours/registers.

 

Also Hi-res modes, at least the charmode ones have PF1/PF2 inverse could had at least another 2 like PF0/PF3 inverse or something similar...

[thorfdbg]

10 hours ago, rensoup said:

Does anyone know the actual reason ? Was it 8 originally and they just decided to cut the transistor count ?

Not only that. Most importantly, available slots for player DMA. 8 players would have required 3 additional DMA slots from Antic, and thus would have required to reduce the size of the playfield. Later on, in the Amiga, Jay did exactly that: Made the start of the playfield DMA configurable such that you could compromize between the number of sprites and the width of the playfield.

[emkay]

In theory it is very simple. If you keep the rule to use grays for the background, mixing has a logical sense. 

 

Here for someone to play around in G2F.

 

17 colors without DLI

color.zip

[Rybags]

Pin count also contributed to lack of possible colours - the AN bus only gave 8 possibilities and 3 of them were needed for sync control.

But mode 9 of course could have been given some sort of algorithmic mode like halfbright or something.

Available DMA cycles meant that they could have done a 16 colour 160 mode and more players but they were lifting the bar from a not very big height, remembering that nothing came near it for about 3 years.

[rensoup]

6 hours ago, emkay said:

In theory it is very simple. If you keep the rule to use grays for the background, mixing has a logical sense. 

 

Here for someone to play around in G2F.

 

17 colors without DLI

color.zip 1.04 kB · 5 downloads

That's not JP quality but that's a clear example.

 

I can see PM01 affecting PF01 and going over PF23 and PM23 affecting PF23 while going under PF01

 

[emkay]

8 hours ago, rensoup said:

That's not JP quality but that's a clear example.

 

I can see PM01 affecting PF01 and going over PF23 and PM23 affecting PF23 while going under PF01

 

 

You could use P3 and P4 at full width and filter the resolution with the PF, while P1 and P2 need to be shaped to the details they have to show.  This means P1 and P2 are for smaller objects , P3 a P4 for wider objects or coloring ...

For more variations it is more useful to have the Missiles using the Player colors. 

 

Imagine, if you use the full width of player 4 , using one of the playfield colors as "Background" , you could do attack waves that cross the Player to have the separated colors. You could do that on 1/4 of the screen width, and while the Attack wave is moving, you only need to adjust the x value of P4 and M4 .  

[Irgendwer]

Please be aware that there is a visualization bug in G2F (at least since version 4.0.2.3) in conjunction with PRIOR 0:

 

(examine the lower overlapping part of the green bar)

 

Reported to @tebe last September, but nothing has changed till then.

 

Edited August 22, 2019 by Irgendwer

[ijor]

On 8/20/2019 at 8:41 PM, Rybags said:

Pin count also contributed to lack of possible colours ... but they were lifting the bar from a not very big height, remembering that nothing came near it for about 3 years.

+1

 

Transistor budget was likely one of the reasons, and as thorfdb mentioned, DMA time availability is also an important issue. I could also add that more players would complicate priority and collision, almost exponentially. But I think they made a design considering what was available and what the technology could do at the time. And probably 4 players and 4 missiles was considered good enough, back then at design time. And nobody can blame them that they weren't ambitious enough

[Rybags]

The blame should be saved for later on.  Even CGIA was nothing special, if you look at Moore's Law then it's feasible that by the time the XL came out around '83 that a combined and enhanced Antic/GTIA could have provided more PMGs and more bits/DMA for the bitmap graphics modes in an upward compatible way.

[ijor]

Certainly they could have enhanced the chipset by the XL time. But guess Atari wasn't interested too much.

 

Nowadays, when they design consoles, they consider what would be the state of the art at the future, by the time it would be released, and not "now" at design time. Furthermore, they consider what would be the cost further in time, after the launch date. They don't mind selling the consoles at lose, for some time.

 

Of course, it was very different back then in Atari times.

[Gibstov]

2 hours ago, ijor said:

+1

 

Transistor budget was likely one of the reasons, and as thorfdb mentioned, DMA time availability is also an important issue. I could also add that more players would complicate priority and collision, almost exponentially. But I think they made a design considering what was available and what the technology could do at the time. And probably 4 players and 4 missiles was considered good enough, back then at design time. And nobody can blame them that they weren't ambitious enough

Do you think they could have done what the Atari 2600 did, and allow double or triple copies of a player? (NUSIZ)

 

Imagine what games could have been made with 12 players (3 copies of each player).

[+Stephen]

1 hour ago, Gibstov said:

Do you think they could have done what the Atari 2600 did, and allow double or triple copies of a player? (NUSIZ)

 

Imagine what games could have been made with 12 players (3 copies of each player).

I think the double and triple copy as well as mirroring features being removed was a wrong move.

[emkay]

All blaming "what they could have done" doesn't help. Things are as they are. 

 

The challenge is to get the best out of the given...

[globe]

Seeing this topic, here's a small color mixer with number of unique colors check, so you don't waste such precious resources to some 'bad' OR

 

mixer.xex

[emkay]

35 minutes ago, globe said:

Seeing this topic, here's a small color mixer with number of unique colors check, so you don't waste such precious resources to some 'bad' OR

 

mixer.xex 1.57 kB · 4 downloads

Nice tool.

But for some reason not all numbers "shift+" were accessible .  

[globe]

28 minutes ago, emkay said:

Nice tool.

But for some reason not all numbers "shift+" were accessible .  

Didn't check on real machine.

Before it was just a small routine where I directly changed color registers, then told WUSDN to compile and run to get my palette.

Will check now.