Top Games You'd Like to Have for Atari 8-Bits

[+MrFish]

That island deserves its own game x)

 

Sure... whenever he crash lands, we can have it go to the island

where he has to salvage his plane and deal with the natives until

he gets rescued.

 

 

I think you should test how it looks while moving, maybe an animated gif could be enough to see if there is any unanticipated "weirdness" going on..

 

I have a way to test it in Photoshop. I just checked it out and it

looks great. The mask becomes more defined as it moves over

the different parts of the water pattern, which it just what I

wanted to see; over the island it's even more so, since all colors

get blocked there.

 

It all looks much nicer than the impressions you get in the static

image. I tested it out on each object individually; and I like it.

 

 

But at least the bullets should use the mask, they look a lot better that way.. and they are small, so the rendering cost should be low.

 

The bullets are part of the reason why I started messing around

with masks in the first place. I always liked a lot of those NES

games where the bullets were done that way. It makes them

better defined and easier to see during combat.

Edited April 17, 2018 by MrFish

[popmilo]

That island deserves its own game x)

I think you should test how it looks while moving, maybe an animated gif could be enough to see if there is any unanticipated "weirdness" going on..

But at least the bullets should use the mask, they look a lot better that way.. and they are small, so the rendering cost should be low.

 

Now.. if Popmilo's test don't appear soon enough, someone else could be tempted to do them

That island sure looks like one of the best islands I've seen on 8bit so far

 

Here is a test without blue mask:

 

1942_no_mask.xex

 

And with blue mask:

 

1942_blue_mask.xex

 

Both versions have their good and bad sides. Masking looks good but costs cpu time.

Anyway, don't see any serious obstacle for making shooter similar to 1942-43.

[emkay]

How about renaming the thread? Somehow to "what impossible stuff technical demos do you want to have on the ATARI that makes you screaming for an other system, where such games run in that way" ?

[+MrFish]

That island sure looks like one of the best islands I've seen on 8bit so far

 

Here is a test without blue mask:

1942_no_mask.png

 

1942_no_mask.xex

 

And with blue mask:

1942_blue_mask.png

 

1942_blue_mask.xex

 

Both versions have their good and bad sides. Masking looks good but costs cpu time.

Anyway, don't see any serious obstacle for making shooter similar to 1942-43.

 

They look great Pop; thanks for taking the time to experiment with them!

 

Can you do a version that has half as many sprites for comparison?

 

[Edit] It also would be useful to see them using an up and down movement

pattern.

Edited April 18, 2018 by MrFish

[Rybags]

Hey look! We actually have Galaga and Exerion on the Atari computer! And an improved version of Galaxian.

 

Why anyone wants to misrepresent the actual system these games are on, no idea. Actual system is a NES.

 

[carlsson]

Hey, if YouTube had existed back when you and me were 9 years old and we got a chance to prank the world, I'm quite sure at least I would've tried to do it once or twice. Perhaps you had higher standards even as a kid. It looks very convincing with cartridge plugged in and everything.

[popmilo]

 

They look great Pop; thanks for taking the time to experiment with them!

 

Can you do a version that has half as many sprites for comparison?

 

[Edit] It also would be useful to see them using an up and down movement

pattern.

How do you mean up down movement ? Like in 1942-43 ?

ps. left all code on home pc.... Will code more examples it later today.

[+MrFish]

How do you mean up down movement ? Like in 1942-43 ?

ps. left all code on home pc.... Will code more examples it later today.

 

Yeah, just how a lot of the attack runs go down the screen and then back up.

I guess they're always at a slight angle, but even just straight up and down

would be a good example.

[leech]

The A8's are capable of better. I guess it's universally agreed that we

got only slightly updated versions of the 2600 games from Activision.

 

I suppose the music was revolutionary enough on a 2600, but for an

A8 that theme does get tiresome. I also think the minor key version of

the song that plays when your guy is sent back plays a little too long.

They could send him back a little quicker. I prefer not to dwell on the

fact that I screwed up in a game any longer than necessary.

I played that game enough that at the mere mention of it, the song is stuck in my head for days.

[popmilo]

 

Yeah, just how a lot of the attack runs go down the screen and then back up.

I guess they're always at a slight angle, but even just straight up and down

would be a good example.

Here is couple examples in 25 and 50 fps with enemies moving in tight ellipse.

 

1942_25fps_12sprites.xex

 

1942_50fps_6sprites.xex

 

I've reduced number of masked sprites to 6 till it got back to 50 fps. Just remember that this is just for visual test. In real situation these enemies would be drawn using Player gfx. Without need for masking or shifting it would be faster. That is the next test

[+MrFish]

Here is couple examples in 25 and 50 fps with enemies moving in tight ellipse.

 

1942_25fps_12sprites.xex

 

1942_50fps_6sprites.xex

 

I've reduced number of masked sprites to 6 till it got back to 50 fps. Just remember that this is just for visual test. In real situation these enemies would be drawn using Player gfx. Without need for masking or shifting it would be faster. That is the next test

 

Thanks, it looks great.

 

My idea for having less sprites was to see how much faster they could move on the screen.

Shouldn't 6 sprites be moving much quicker than 12 sprites?

[popmilo]

 

Thanks, it looks great.

 

My idea for having less sprites was to see how much faster they could move on the screen.

Shouldn't 6 sprites be moving much quicker than 12 sprites?

My idea was to see same movement difference in 50 vs 25 fps

 

Ok. Here is a non-slowed down version:

 

1942_50fps_6sprites_no_slowdown.xex

[+MrFish]

My idea was to see same movement difference in 50 vs 25 fps

 

Ok. Here is a non-slowed down version:

 

1942_50fps_6sprites_no_slowdown.xex

 

Cool; that's what I was looking for!

 

It looks a little bit jerky, though; maybe because of the step value used for the curve?

[popmilo]

 

Cool; that's what I was looking for!

 

It looks a little bit jerky, though; maybe because of the step value used for the curve?

Please try it on a real Atari and crt if you can. Emulators (even Altirra) sucks at showing 50fps movement.

[+MrFish]

Please try it on a real Atari and crt if you can. Emulators (even Altirra) sucks at showing 50fps movement.

 

That's not the issue. I slowed it down in Altirra and I can see why it looks that way.

The vertical step value of movement increases as the planes approach the middle

portion (vertically) of the ellipse (it looks like it goes from 1 pixel movement to about

4 pixel movement at the greatest point -- hard to see exactly with the patterns and

masks).

 

I would expect this for an ellipse, because of the difference between height and width.

You'd probably have to use a hand edited table in order to keep things all moving at

1 pixel throughout the entire ellipse.

 

I'm not saying it looks bad, in fact it's one good way to increase speed of objects

in a game -- within reasonable limits. I just wasn't sure what I was seeing at first.

 

BTW, it's not so noticeable on NTSC since the movement itself is slower there.

Edited April 20, 2018 by MrFish

[emkay]

That's not the issue. I slowed it down in Altirra and I can see why it looks that way.

The vertical step value of movement increases as the planes approach the middle

portion (vertically) of the ellipse (it looks like it goes from 1 pixel movement to about

4 pixel movement at the greatest point -- hard to see exactly with the patterns and

masks).

Welcome to the world of technical issues

Maybe on a computer, using 300Hz (and full HD resolution) for the screen updates, you can do a slow movement line by line and to have the fitting visual changes.

[popmilo]

I would expect this for an ellipse, because of the difference between height and width.

You'd probably have to use a hand edited table in order to keep things all moving at

1 pixel throughout the entire ellipse.

....

BTW, it's not so noticeable on NTSC since the movement itself is slower there.

Ok. Think it's like this:

- In Pal on emu, stuff running in 25 or 50 fps is not good looking on 60Hz PC display because of non-integer ratio.

- In Ntsc on emu, it looks fluent (30 or 60 fps on 60Hz display).

 

Shame is how much less cpu time there is on NTSC ! I knew it was less but I needed to drop from 6 to 4 sprites to stay on 50fps...

1942_4_sprites_60fps_on_ntsc.xex

 

And yeah, I've used simple SIN command in Mads compiler. Nothing fancy, no subpixel accuracy, no hand adjusted or precise curved paths. Just 256 steps around a central point. So "jumps" of more than single pixel are in there.

[+MrFish]

And yeah, I've used simple SIN command in Mads compiler. Nothing fancy, no subpixel accuracy, no hand adjusted or precise curved paths. Just 256 steps around a central point. So "jumps" of more than single pixel are in there.

 

Ah, I figured it was something like that. Precalc table is better (for speed); although I don't know

what MADS is producing for code using that function; maybe it's not too bad?

 

It's ok; no complaints, just thinking out loud here. As I said, jumps like that are a great idea for

speeding things up (in terms of velocity). So it's a nice little preview of what that looks like.

 

In terms of coding and overall speed it can be even better (of course) if some form of byte

boundaries or even character boundaries are used for jumping, as long as it looks good.

 

But that's another track. I like the results we're seeing using your engine.

[Bill Lange]

There are new Ultima-like games coming out for the Apple II (Nox Archaist) and the C64 (Unknown Realms). It would be nice to see either one of these ported to the A8.

[+MrFish]

Shame is how much less cpu time there is on NTSC ! I knew it was less but I needed to drop from 6 to 4 sprites to stay on 50fps...

1942_4_sprites_60fps_on_ntsc.xex

 

That looks very smooth and fluent, albeit only 4 sprites.

[emkay]

That looks very smooth and fluent, albeit only 4 sprites.

When scrolling happens and the gameplay is going on, you won't really recognize less fluent movement. The relevant point is to have the movement always stable synchronized to the scrolling.

[+MrFish]

Some ideas for the hi-res score area.

 

Also, showing a mixture of masks for large objects and no masks for small ones.

A different look, which might help to differentiate and also save on the amount

of objects that need to be masked. The bullets seem like they don't really need

it anyway, since there's not much too them and their colors are sufficient to

make them stand out.

 

Edited April 21, 2018 by MrFish

[emkay]

Just a reminder: Pictures with some colors might look impressive.

But it is even more impressive, if the action goes fullscreen...

 

[+Stephen]

Just a reminder: Pictures with some colors might look impressive.

But it is even more impressive, if the action goes fullscreen...

 

Nice mix of ANTIC Mode E at the top * bottom of screen, and Mode D for the center. How are they getting the single line graphics above the double line "sprites" though?

[emkay]

Nice mix of ANTIC Mode E at the top * bottom of screen, and Mode D for the center. How are they getting the single line graphics above the double line "sprites" though?

I guess, it's the missiles.

It is also a good demonstration that , after double scanline mode is used, the limit isn't generally at doubled scanlines. PMg always could use single scanlines for enhancing the details.