Beyond Castlevania...

[+grafixbmp]

I'm sure many of you have seen or even posted in the castlevania topic and like the idea and even maany have said they will patiently wait for it to arive. But when I started this game port, I did so not just for having castlevania on atari but for a new game engine that would allow new types of games that have more detail and plently of action on screen to give a more elaborate experience to the atari. One that will keep the atari stronger than ever. Since the conception back in/around september of 08, I had wanted to keep the ROM size low and had hoped that extra RAM wouldn't be necessary, but as things began to be flushed out, this didn't seem possible. That being the case, I have decided that bankswitcing scheme F4SC would be not just the best but perfect in terms of expandability. With the extra 128 bytes, this is ideal for a frame buffer dealing with all sprite data.

 

I have been (over a long time now) trying to come up with a few graphics that may inspire others, for when the engine reaches a viable solution then others can use it for there own games or even other downgraded ports of popular games that many would enjoy on the atari. This is also one big reason why I encourage many to get involved as well.

 

Several games (granted they would need to be taylored for atari) that I have thought of could be done to some extent. These include but are not limited to:

 

Contra

Blaster Master

Mario style games

metroid

etc

 

With 32 k and 128 bytes of extra ram, This style of engine would give this bankswitch type more use and therefore more support. I KNOW it can be done. Who's with me. (I know and am glad for The Golden Bands assistance) and others who have helped.

 

I will periodicly be adding a few images here and there for inspiration to thoes of you out there. Here are a few for now. enjoy!

 

 

*NEW*

Here is a quick image for contra. He is just a bit big but that could easily change. For concept purposes, I used a missle to extend the gun out. 2 scanlines at 4 X and the third at 2X. I figured that if the missle ran full 60 Hz with no gun fire and at 15 Hz when fired so that would give 1/4 for the gun and 3/4 for bullets.

 

Here is another. you guys should be able to figure it out.

Edited May 5, 2009 by grafixbmp

[nonner242]

Contra would be sweet!

 

But I would love to see Ninja Gaiden,R-type,Final Fight,TRON & FRENZY.....I would think would be killer!

[+grafixbmp]

Contra would be sweet!

 

But I would love to see Ninja Gaiden,R-type,Final Fight,TRON & FRENZY.....I would think would be killer!

I once did some mock only work for R-type (my earliest posts). I figured that the game Vanguard could be changed tward an R-type look and feel. The game was great but as R-Type would be awesome.

Edited May 2, 2009 by grafixbmp

[ZylonBane]

I thought you'd gotten over the whole posting images as BMPs thing, your username notwithstanding.

[+grafixbmp]

I thought you'd gotten over the whole posting images as BMPs thing, your username notwithstanding.

Just to make you happy, I fixed them. but your right. lol

[+grafixbmp]

If anyone thinks that Contra would make a good game on the atari and would like to experiment with it some. I can work up a few graphics in any spare time I have left. With many similarities in the game engine, this could help numerous projects at the same time. Any takers? Or if there are other games that anyone would like better that would have similarities, I could look into thoes as well.

[nonner242]

If anyone thinks that Contra would make a good game on the atari and would like to experiment with it some. I can work up a few graphics in any spare time I have left. With many similarities in the game engine, this could help numerous projects at the same time. Any takers? Or if there are other games that anyone would like better that would have similarities, I could look into thoes as well.

 

[+grafixbmp]

I also had been playing ninja gaiden lately and I bet it would work on atari the easiest. The SNES trilogy game pack. Actualy all 3 would probably do well.

[+grafixbmp]

Another one for contra. did a jump spin graphic.

Edited May 10, 2009 by grafixbmp

[+grafixbmp]

OK I have an unfinished mock screen that seems rather detailed but the intention I wish to convey is a 30 hz playfield where the brown shades (rockface and tree trunks) are a single frame and the other frame has green, grey and for certain spots blue (grass, metal, water) these are interleaved. The real question is if scanline updates are going to be easy enough to do or just too complexed. but would love some feedback and if anyone else would give a try at their own mockup that would be awesome.

 

This method could also combine 2 separate colors together for other shades. Not sure what colors could be created but with some experimentation, this could possibly be a very colorful little game/test/demo.

 

 

Tried for more detail and I think this could be a workable solution. The only thing left are the tops of the trees and that will be easy. Would love some feedback on the look and the practical doability of it as well. Please guys, feel free to try your own mockups. If others would like a contra game I will be willing to do the graphics but as far as the main work goes, well I am keeping that on castlevania.

 

 

Added a few sprites.

 

Edited May 31, 2009 by grafixbmp

[+grafixbmp]

I had some fun today. During my work breaks and such, I toyed with an image meant for atari 2600 displaying but I think is is unacheivable as is and would need to have alterations but it was sure fun and I wanted to share it with you guys. Feel free to take it and edit it so it could be more achievable on real hardware. I will also place the NES version for comparison.

 

 

[Corby]

now that is impressive

[roland p]

It would be nice to see an atari 2600 picture contest

[+grafixbmp]

It would be nice to see an atari 2600 picture contest

I would much rather have it running on real hardware. But a contest for elaborate still pictures running on 2600 hardware is an excellent idea.

[+grafixbmp]

I found the images I did for R-type on an old harddrive and thought I would share these few images.

 

Also, as you can see, here is also a fun mock of super mario bros.

[MausGames]

I'm really impressed with all of these mockup; Any chance you could do Marble Madness? Your design style seems like a perfect match for it.

[+grafixbmp]

I'm really impressed with all of these mockup; Any chance you could do Marble Madness? Your design style seems like a perfect match for it.

I guess I could. It is getting rather difficult working on castlevania I've had to restart the designs about 4 times now and it's frying my brain.

 

There is that one game with the koalla bear that has that isometric look. The one with the walking trees. I could base i off of that layout, heck it might even work well with that kernel.

[+grafixbmp]

Alright, I have given marble madness some thought and for starters, because of advanced collision detection being necessary, I think that a mirrored playfield without pf0 being used will be key and with ease of vertical scrolling will make this game concept workable. Like I said before about the other game (crystal castles btw) that kernel may be usable to some degree for a game like this. I also think that based on the wiki of the original arcade version this one should also be based around 30 fps but however use a dual graphics table thereby giving more tha 2 colors on screen for the playfield. Shadows will more than likely be impossible. and the ball can be done with a static ball one frame and the rotational dotted ring another. I also figured that the top first 8 to 16 scanlines could be cut off from the gameplay area for the time clock and that would also give less to be done on the screen. I should have a working concept soon.

[+grafixbmp]

Here is a quickie I did in an hour. Nothing much but eye candy.

 

 

I got a few snaps from crystal castles and a shot of the real marble madness.

 

The lowest downgraded port I could find of marble madness was the commodore 64 version.

 

 

I may be abe to base an atari 2600 version off of the way this one looks. Many things will need to be simpified just because of the lower resolution and lack of extra color. I can design the graphics rather well but I sure won't try to program it. I figured that the vertical position of the ball shoud stay at the same pace but the maze woud be the ony thing that moves. This will keep the ball from leaving the screen so fast like it does in the other ports when alot of speed is gathered. The C64 version also lacks shadowing so atleast there is a verion that already eiminated that. The best way to really get a good version with high resolution is with a chimera or other type of advanced cartridge. Otherwise, a high resolution map woud be next to impossible to achieve.

 

I would like to do one full level layout and then anyone else who wants to pursue a playable atari 2600 version can use it to design off of.

Edited June 12, 2009 by grafixbmp

[+grafixbmp]

For starters, check this out.

It could very well change and get better.

 

Added some more detail.

I will add more until I finish this level.

I would also like to do the practice as well sometime after this one is finished.

Edited June 19, 2009 by grafixbmp

[+grafixbmp]

In regards to marble maddness I have a theory about how the game mechanic could theoreticaly work or atleast help to make one work. I had once said that the design needs to allow for a more elaborate collision detection system. well say with the way the screen is drawn and considering the movement of the joy stick in all 8 possible directions I think a special collision system in code that does pattern comparisons for a way to know what kind of area you are located at. I will list out a bit of what I mean

 

If the last 2 scanlines worth of player 0 where ever on the screen they happen to be were checked for a collision twice for one full frame of the game itself, then a detailed collision matrix could be constructed from the data.

 

Let me elaborate. The frame of this game consist of one bitmap done in a general tone grey while a second frame is done in the tone of the level. Each frame is marked 0 or 1 respectively. Each real frame is checked at the last 2 scanlines of the ball displyed on both frames.

 

Frame 0 possible outcomes 0, 1

 

Frame 1 possible outcomes 0, 1

 

With this there are atleast 4 possible outcomes considering both frame 1 and frame 2

 

Balls tend to roll a bit on their own in one of the possible 8 directions. This being true, left and right can be tracked by atleast every 4 H positions and up and down can be tracked based upon 2 thing THe first is not that the ball moved but tht the PF scrolls verticaly up and down. If the PF scrolls atleast 3 double sets of scanlines before starting over sorta speak, then a patern can be recognized indicating where the ball is currently as far as if it should fall off or not. As far as slopes go that may be another hurdle to overcome.

 

 

patterns for areas are aproached difrently depending on which way the pattern is recieved.

 

Falling off from behind can be:

 

BLUE

BLACK

GREY

 

or

 

BLACK

BLACK

GREY

 

These would be read going up

 

fall forward can be:

 

GREY

BLACK

BLUE

 

These would be read going down

 

 

There are also the areas where a back wall obstical can be detected in software

 

Going up would read

 

 

BLUE

BLUE

GREY

 

 

These are just a few examples and not every place (currently) reflects these examples. But with some basic design layouts depending on the patterns in certain orders nd the directin shifted up or down, certain crude detections can be acheived and this will give a wee bit of slack on control while keeping overall gameplay intresting. It needs fluc=shing out but what you guys think?

 

 

Remember that the current images DO NOT reflect the finished design as there are only 3 colors used and no blend of blue and grey at this time. This will give other possibilities.

 

 

With 3 double width scanlines with each having 4 possibilities and also depending on the order recieved based upon the movement up or down... a total including the ups and downs (lol!) 64! This may give many boundaries and switches that will indicate the ball's movement. However it does not reflect horizontal movements well. That is where others need to come in and conjure up some magic where I have no clue. I will have an almost finished level quite soon. yay!

 

I sure hope this has some merit but if not, I would be intrested in more advanced techniques for collision detection when dealing with terrain. That is to say if anyone knows any.

[+grafixbmp]

Sorry. Double post. My bad.

Edited June 19, 2009 by grafixbmp

[MausGames]

I think you've gone far above and beyond what I expected, that is instantly recognizable as marble madness. It looks like it would be very playable too; the isometric view usually isn't very nice with such limited horizontal resolution, but your idea to use three colors per line definitely makes up for it. Depth and boundaries are easy to make out.

 

As far as collision detection, that is way out of my league, but it sounds like your idea to take advantage of the patterns created by the multi-frame playfield is a good one.

[+grafixbmp]

Almost finished with the second level terrain. Hope you like it. When this gets done I may atleast start the practice round.

 

 

And now for the final draft. This finished level does not have the fourth color. But I am rather pleased with it.

 

 

Next will be practice...

Edited June 23, 2009 by grafixbmp

[cd-w]

And now for the final draft. This finished level does not have the fourth color. But I am rather pleased with it.

 

This is a very nice design. I assume you would get three colours per line by flickering on alternate frames? I wonder how it would look if you alternated colours on each line instead (without flickering) - there would be a gap between each colour, but it might still look OK?

 

Chris