Street Fighter 1 - Scorpion Engine

[msmalik681]

If I can separate player parts I can save graphics data that is wasted on dead pixels.

[jotd]

that's the most annoying part of a fighting game: character "compression" and also collision masks (you have to create some hit zones and some invincible zones else if for example you get punched in the leg you're attacking with, you're hit...).

There's a good reverse of SF2 collision system around. It's a lot of work to input such data.

not to mention the A.I. that you have to reverse more or less faithfully or the game isn't going to look like the original. A fighting game is never a piece of cake, even with Scorpion, because the hard part is the logic, not the basic blitting or animations.

Personally after I'm done with Karate Champ (if I can make it work) I'll take a vacation redoing Pengo or Phoenix or some other very simple game...

[saimon69]

In case something goes very wrong here is a plan B for you: is indeed a different approach but if well done can wield excellent results
Snk vs Capcom for C64

[Retro1234]

Originally Posted by saimon69

In case something goes very wrong here is a plan B for you: is indeed a different approach but if well done can wield excellent results
Snk vs Capcom for C64

That looks so good

[malko]

This C64 game is really cute

[msmalik681]

Originally Posted by saimon69

In case something goes very wrong here is a plan B for you: is indeed a different approach but if well done can wield excellent results
Snk vs Capcom for C64

wow I had no idea the C64 could make such great fighting games nice work. I can't really help as I know nothing about low level coding.

I have started a github repo for this project: https://github.com/msmalik681/street_fighter_1_amiga as a update I now have all of ryu's moved completed but the game is almost hitting 1mb with just ryu so I am going to try to separate the upper and lower body see if that can cut down the size might have to start cutting down the animation frames too.

[str0m]

It's a port/based on/looks same as the Neo Geo Pocket Color game which is an excellent game https://cheapergamer.co.uk/snk-vs-ca...ennium-review/

[alain.treesong]

Originally Posted by msmalik681

I have started a github repo for this project: https://github.com/msmalik681/street_fighter_1_amiga as a update I now have all of ryu's moved completed but the game is almost hitting 1mb with just ryu so I am going to try to separate the upper and lower body see if that can cut down the size might have to start cutting down the animation frames too.

Maybe what I was talking about in some previous messages is more obvious now ? ;-)
You cannot have a lot of frames (so nice movements) with only 512Ko of chip in 16 colours screen. You have to use dual playfield (moreover with dual playfield you can have 50 fps)
With 1mb chip memory, it is more possible but a standard Amiga 500 has only 512kb.
With characters in 16 colors, why not doing a 1200 conversion (2mb of chip) ?

Thanks for the git repository.

[Tsak]

Originally Posted by alain.treesong

Maybe what I was talking about in some previous messages is more obvious now ? ;-)
You cannot have a lot of frames (so nice movements) with only 512Ko of chip in 16 colours screen. You have to use dual playfield (moreover with dual playfield you can have 50 fps)
With 1mb chip memory, it is more possible but a standard Amiga 500 has only 512kb.
With characters in 16 colors, why not doing a 1200 conversion (2mb of chip) ?

Thanks for the git repository.

Well, problem is Scorpion doesn't support dual playfield (yet). Other than that he might be able to squeeze them -regardless- if he starts cropping them as suggested. If that's not enought then going full modular might do the trick, however this is also not fully supported yet (might be soon though).

[acidbottle]

Originally Posted by Tsak

Well, problem is Scorpion doesn't support dual playfield (yet).

Are you saying that this feature is being seriously considered now I know earok has covered this before and at the time wasn't planning on it. Id love dual playfield implemented of course!

[Tsak]

Originally Posted by acidbottle

Are you saying that this feature is being seriously considered now I know earok has covered this before and at the time wasn't planning on it. Id love dual playfield implemented of course!

Well, I have recently suggested this to Erik and he's definitely considering to add it at some point. But hasn't since it's not a major requested feature by the community.

[saimon69]

Originally Posted by alain.treesong

Maybe what I was talking about in some previous messages is more obvious now ? ;-)
You cannot have a lot of frames (so nice movements) with only 512Ko of chip in 16 colours screen. You have to use dual playfield (moreover with dual playfield you can have 50 fps)
With 1mb chip memory, it is more possible but a standard Amiga 500 has only 512kb.
With characters in 16 colors, why not doing a 1200 conversion (2mb of chip) ?

Thanks for the git repository.

I would try to recreate as much animations as possible with double body, you know like being hit, you can have the upper part move a bit instead of another frame, same with some of the kicks

By the way i wonder: if he store the characters in seven colors might be able to save some RAM?

[alain.treesong]

Originally Posted by saimon69

I would try to recreate as much animations as possible with double body, you know like being hit, you can have the upper part move a bit instead of another frame, same with some of the kicks

By the way i wonder: if he store the characters in seven colors might be able to save some RAM?

You mean that the characters are 2*8 different colors (including one transparent) ? On the disk you save some bytes but in RAM 4 bitplanes are used so you don't save some RAM.

For the disk it can be a good idea. Because there is another difficulties : how many disks (and disk swap and time loading) for this kind of game on a standard 500 and how to optimize that.

So, as pointed @jotd, it is difficult to code a good playability but for the 500 the technical thing is quite difficult too :-)

Difficult, but interesting

[msmalik681]

right now I am not targeting anything I just want to get the base of the game working then I will look into optimisations the obvious one being sprite size reduction. And yes reducing the colour count does reduce the used ram quite considerably but that is future talk.

just for feedback does anyone on this forum actually use a 512kb amiga I think targeting 1mb might be a better option but that kind of kills the whole point of this little test of mine.

[saimon69]

Originally Posted by alain.treesong

You mean that the characters are 2*8 different colors (including one transparent) ? On the disk you save some bytes but in RAM 4 bitplanes are used so you don't save some RAM.

For the disk it can be a good idea. Because there is another difficulties : how many disks (and disk swap and time loading) for this kind of game on a standard 500 and how to optimize that.

So, as pointed @jotd, it is difficult to code a good playability but for the 500 the technical thing is quite difficult too :-)

Difficult, but interesting

OK, the Tiertex version did suck but they put all of the characters and half the backgrounds in one disk at 16 colors, so a two disk version should be feasible on paper - with reduced assets of course

AND Final Fight was running in a 512k(!) amiga, as bad as it is was a miracle, i wonder if some optimization could be possible.
Maybe not in Scorpion, though

[msmalik681]

At the moment im working with arcade size sprites and arcade size backgrounds the 16 colors does help but might have to do some heavy frame cutting just for a 1mb version.

[alain.treesong]

Originally Posted by msmalik681

...
just for feedback does anyone on this forum actually use a 512kb amiga I think targeting 1mb might be a better option but that kind of kills the whole point of this little test of mine.

Originally Posted by msmalik681

At the moment im working with arcade size sprites and arcade size backgrounds the 16 colors does help but might have to do some heavy frame cutting just for a 1mb version.

If you do it for an amiga with 1mega chip, it's a 500+ or a 600. It's not really OCS...
At that time, might as well do something for the 1200. At least there are 2 megachips as standard and many people had a hard disk (it was a shame not to have sold a hard disk systematically in 1993) which will make the game more enjoyable and give an idea of what might have come out.
In addition you will have 50 fps easy and you can make the scenery in 16 separate colors (so the game will be in 32 colors while waiting for the dual playfield, which is 16x16 on the AGA). In fact you can use directly the original backgrounds (at least the front)
Very personal opinion of course :-)

[Tsak]

Originally Posted by msmalik681

just for feedback does anyone on this forum actually use a 512kb amiga I think targeting 1mb might be a better option but that kind of kills the whole point of this little test of mine.

The most common Amiga 500 configuration is 512k Chip + 512k Slow ram. Plain 512k Amigas has stopped being a thing since the early 90s, where (almost) every major game released required an additional 512k. So since then 512+512k has been considered a standard.

Now, if your question is about CHIP ram only then this is a different story. A500+ and A600 came with a solid 1MB for Chip but this is a rather uncommon configuration for plain a500s.

As for recent times, I've never (ever) seen anyone using unexpanded 512k Amigas or even asking for games that support this config.

[saimon69]

Originally Posted by Tsak

As for recent times, I've never (ever) seen anyone using unexpanded 512k Amigas or even asking for games that support this config.

Blocky Skies was available for the unexpanded A500, made by Alpine9000 in 2016, you can check with him

[mcgeezer]

Originally Posted by saimon69

Blocky Skies was available for the unexpanded A500, made by Alpine9000 in 2016, you can check with him

Originally Posted by tsak

"As for recent times, I've never (ever) seen anyone using unexpanded 512k Amigas or even asking for games that support this config."

But we do it anyway because as developers we understand the boundary of what was possible when the Amiga was mainstream. This is what makes the gamedev a challenge on the Amiga.