Coder shows possible way in which some elements of Street Fighter 2 might be possible

[lmimmfn]

This might be useful, shows the hit boxes
https://doctorguile.github.io/ssf2st/compare.html

Skip the AI and just make it 2 player

[alain.treesong]

@neoman
A bit off topic, but is it possible to continue with your engine to make a new orignal AGA game ?
A good SF2 version of AGA is possible as the great video or a game like shadow fighter OCS prove it. Honnestly a new "SF2 like" AGA game, even if there are only for example 6 chars + 1 at the end and 3 backgrounds + 1 at the end will be wonderfull.

[Retro1234]

in another thread I posted a link to SF2 logic it's not that complicated but would be time consuming.

Obviously the AI, amount of moves and frames etc are beyond Shadow Fighter and other games and this is part of what killed it on Amiga.

I personally think saying your gonna make the boat one big bob is laughable.

If you search you will find many of these SF2 threads and there mainly click bait except what I thought was more click bait master464 actually thought through the palette, how to use the copper and sprites, you commented on his youtube video so you've obviously see it.

[kremiso]

Looking into the old x68000 port can give some ideas in how to develop/solve the issues in the other stages too?

[ Show youtube player ]

[lesta_smsc]

Can the original game logic from Amiga version not be used with updated sprites?

[neoman]

@alain.treesong

Well yeah, let's see. This is mostly a graphics prototype based on my demo engine now. But we've lots of rastertime free, so I'm quite happy with it already. I really hope to continue to make a real game engine out of it.

Of course always nice to get some useful information about existing efforts to reverse-engineer SF2. Thanks =)

[Retro1234]

so what part of the video if any is your engine on Amiga? not PC animation.

sorry your think it might be possible to colour the background characters on the boat with the copper and the boat is gonna go up and down it's not going to work is it?

sorry I'm not happy with way you have gone about this, a fake video, you have some good credentials but this is not great is it.

[DanScott]

Originally Posted by Retro1234

so what part of the video if any is your engine on Amiga? not PC animation.

sorry your think it might be possible to colour the background characters on the boat with the copper and the boat is gonna go up and down it's not going to work is it?

sorry I'm not happy with way you have gone about this, a fake video, you have some good credentials but this is not great is it.

Well, as a coder, I understand what they want to do regaring colouring the boat and the technique involved using the copper, and it is actually possible

I'd like to know your technical analysis as to why it won't work..

[jotd]

changing palette in the middle of the line is doable. you need at least 8 pixels "time" to change one color but it is possible

dynamic hires images were done that way.

[Retro1234]

your gonna change SMALL sections of the copper to change colour both Horizontal and Vertically as the boat goes up and down, na not gonna happen.

[DanScott]

Originally Posted by Retro1234

your gonna change SMALL sections of the copper to change colour both Horizontal and Vertically as the boat goes up and down, na not gonna happen.

very easy to do.... next...

[Retro1234]

whats easy for you.......show me a game where this is done?

[DanScott]

Originally Posted by Retro1234

whats easy for you.......show me a game where this is done?

I can't show you a game where it's been done, because there hasn't been a game where it's been done (as far as I know), but the technique is easy, and very achievable on AGA on A1200

Remember that most AGA games released in the past were not really any technical marvel

[Retro1234]

I mean I can't argue with Dan can I, with his record how could I but you must see I'm very sceptical but like I say prove me wrong.
If a true demo such as Master464 demo but with all the collision checks and A.I was done then yes I would be impressed because I can't help feel there is a reason no one has pulled it off in the past.

[alain.treesong]

@neoman
What you show in the video and what I assume you have implemented is just great.
So it's cool that you're continuing your engine with maybe (or not) a new game in the future.
Personally I played a lot of SF2 at the time in the arcade and a lot of SF3 and Capcom vs snk on dreamscast (with the sega arcade controller of course ;-) .
What I can say is that the visual technique is certainly important but the most important thing (for me) is the gameplay. I prefer less scroll and more playability (so less time to code the technical aspect and more time to code the playability aspect).
I think that with 2mega of chip and only 16 colour characters, you can have a lot of frames. It will be necessary to manage the response time, the combination of moves between different characters...etc (of course you already thought about that). Indeed, getting information from SF2 is a good approach.
In any case, as I said at the beginning of the message, the video is really good and what you say in it seems to me sensible and feasible (thump up! ;-) )

[DanScott]

Just to be clear, I suggested the topic name change based on the contents of the video.

However, everything that was discussed in the video does make sense from a technical perspective. Obviously they will come up with some challenges implementing this, but overall I think that what is shown in the video (whether or not that actuall video imagery comes from an Amiga) is quite achievable.

As I've mentioned before in the past in some thread or other, the actual game logic including AI and collisisions for a PvP fighting game should only really be taking a minimal amount of CPU time, leaving plenty for the Blitter / CPU to work in harmony , moving around a good amount of graphic data in various ways

[Retro1234]

I'd also be interested in memory requirements for all the frames etc.

[jotd]

if you check the video you'll notice that the shadows of the fighters disappear from time to time. probably betrays sprite multiplexing issue.

I don't think that the video is fake at all. Why taking the hassle to create such a video if it's a fake? what's the point? Let's stop this "the man never went on the moon" skepticism...

About dynamic copper color change, it's not done too much in games because it requires special processing of images: you can't take any image and do this, not all images can be rendered with that technique.

Going up and down isn't an issue: just issue the same color change instruction at every line for this particular(s) color(s). Just a headache to make it right, but possible, specially with 16-color dual playfield, not 8.

Memory can be cheated on with fastram and decompressing on the fly. Painful, but doable. Needs 68030 but doable.

How many colors for Jim Power (ECS) ? just ... a lot. So on AGA it's even more possible.

great video, great explanations, just ... great. Noone said it was easy to do such a demo/game, someone just said it was possible. And I believe him.

[alain.treesong]

Perhaps a demo will be published soon ?

For the frames count, with (for example) 1mega of chip dedicated to them, we can have a lot of 16 colour frames. To make a simple comparison it is as if we had a 320 x 6250 image in 16 colours in memory on which the frames are arranged. So it's quite comfortable.

EDIT : and of course the blitter which is dedicated to the characters can of course flip some images to manage the double direction

[skyzoo73]

Until a few months ago no one thought it was possible to port the quality of TinyBobble, or a game like Bomjack Beer edition, or the fantastic conversion of Rygar. But with the right team it proved exactly the opposite (with some minor limitations here and there). So I don't see anything impossible if you have a lot of passion, knowledge of Amiga hardware, and support from the community. I personally think it is doable. Of course it will be a mammoth job to be able to make a 100% complete game. But it does not seem to me to be the goal of the author of the video (although I hope so). For conversions of old games as a target the Amiga 500 (1mega) is more than enough, for more complex and recent titles like SF2 surely the Amiga1200 is a good machine, and you could also add a little fastram to the minimum requests, at the time of all the owners of 1200 almost no one had stock, most with 020/030 and 4/8 fastram accelerator cards was the norm. So thinking about the feasibility of the port in question, everything could be easier on this type of configuration. However complicated and laborious, but feasible with quality equal to the snes and genesis / megadrive.