L System 256 colour demo + screen transitions

[cmonkey]

Here's my latest coding efforts on L System. The videos have full narration (if you can bare to listen to my drony voice for 17 minutes). Binaries attached, just in case you want to bore yourself silly looking at 256 rotating colours.

https://www.ukvac.com/forum/data/uploads/1497/L_System_256_colour_demo.zip

 

[joe34]

hey nice one, the chuckie egg stuff i thought was very interesting and cool what u are doing

the atari st screens look really nice although it sounds like you have your heart set on the speccy gfx, but: i dont know if u thought of this already, how about if you just changed the player 'sprite' gfx of the atari st version, to the original Harry from the speccy version (but with more colours for the Harry gfx instead of having him in one colour like the speccy version). maybe that would make the atari st version feel a bit more like you remember/prefer

ps. i could do the Harry sprite gfx for u if u wanted to do that, i'm not a coder though so if my idea is only going to introduce all sorts of problems like changes in the collision detection and so on, just ignore me ok

anyway i hope the project goes well, it looks like loads of fun

 

[Hurray Banana]

Just watched part one Ade nice bit of demo code there, like the adjustment timings.

Do you think you can do some tight timing to monkey (no pun intended

) about with the pallete ram to do proper raster bar work?

 

[Hurray Banana]

Just watched the 2nd demo. It's not crappy mate.

Love the fade effects. Are you titing about with tiles on the foreground tile layer to do the masking of the background tile layer?

Nice work Ade.

 

[cmonkey]

Cheers guys.

Eric - yeah, animated f/g tiles to mask/unmask the b/g tile layer. The pixel dissolve alone took 256 tiles to make it appear random (even though it isn't purely random). That's OK when you've got 4096 tiles to play with but 256 tiles would be the entire tile set of Pac-Man hardware!

I think I've figured out that there are 424 T-states per scan line (including the horizontal retrace). It really depends whether Taito are using the CPU clock as the pixel clock (which I'd imagine that they are).

If so then the calculations looks like this :-

6665280 (CPU clock) / 59.94 Hz = 111199.199 T-states per frame (although this probably gets rounded to 111200)

111199.199 / 262 (number of scan lines per screen) = 424.424 T-states per line (maybe rounded to 424)

I'm betting it's either 424 or somewhere very very near that figure. I played around with it a bit when I was coding the scan line delay routine and got it pretty much down so that the delay routine delayed for exactly 1 scan line.

With that in mind it *should* be possible to be some kind of raster bar effect, although not having pixel level access to the frame buffer would hurt.

@joe34 - I've already got a 4-colour version of Harry that I'm toying with using although I really am tempted to do a pure port (i.e. not altered in any way, shape or form) of the game to L System. I may come back to you if you're a good pixel artist though, it's always handy to know one. Problem with using the Spectrum version of Harry in a graphical style of the Atari ST version of the game is that Harry in the Spectrum game is 8 pixels wide by 16 pixels high but he's 16x16 pixels in the ST version. This means he'd look stupid climbing/descending ladders as the ladders are 16 pix wide in ST version but only 8 pix wide in the Spectrum version. I could use 8 pix wide ladders in an Atari ST styled version of the game but I'm not sure how that would look. Need to experiment when the game engine gets more advanced.

 

[Hurray Banana]

Cheers guys.

Eric - yeah, animated f/g tiles to mask/unmask the b/g tile layer. The pixel dissolve alone took 256 tiles to make it appear random (even though it isn't purely random). That's OK when you've got 4096 tiles to play with but 256 tiles would be the entire tile set of Pac-Man hardware!

I think I've figured out that there are 424 T-states per scan line (including the horizontal retrace). It really depends whether Taito are using the CPU clock as the pixel clock (which I'd imagine that they are).

If so then the calculations looks like this :-

6665280 (CPU clock) / 59.94 Hz = 111199.199 T-states per frame (although this probably gets rounded to 111200)

111199.199 / 262 (number of scan lines per screen) = 424.424 T-states per line (maybe rounded to 424)

I'm betting it's either 424 or somewhere very very near that figure. I played around with it a bit when I was coding the scan line delay routine and got it pretty much down so that the delay routine delayed for exactly 1 scan line.

With that in mind it *should* be possible to be some kind of raster bar effect, although not having pixel level access to the frame buffer would hurt.

Yeah 4096 tiles is a lovely luxury, that I intend to make use of when I finish playing with pac hardware (I don't think I'll ever stop playing with out tho') and move over to the L system.

 

[joe34]

@joe34 - [font="Times New Roman, Times, serif"]I've already got a 4-colour version of Harry that I'm toying with using although I really am tempted to do a pure port (i.e. not altered in any way, shape or form) of the game to L System. I may come back to you if you're a good pixel artist though, it's always handy to know one. Problem with using the Spectrum version of Harry in a graphical style of the Atari ST version of the game is that Harry in the Spectrum game is 8 pixels wide by 16 pixels high but he's 16x16 pixels in the ST version. This means he'd look stupid climbing/descending ladders as the ladders are 16 pix wide in ST version but only 8 pix wide in the Spectrum version. I could use 8 pix wide ladders in an Atari ST styled version of the game but I'm not sure how that would look. Need to experiment when the game engine gets more advanced.

[/font]
ah ok i thought there might be probs due to the atari st harry being a short little fat thing

i'd be happy to help out with any gfx u need for any of your projects, although if im honest, im better at hacking/changing/converting existing gfx, as opposed to actually doing my own gfx right from scratch

 

[cmonkey]

Yeah 4096 tiles is a lovely luxury, that I intend to make use of when I finish playing with pac hardware (I don't think I'll ever stop playing with out tho') and move over to the L system.

4096 is the base level configuration of L System (Plotting). If you move up to Puzznic you get 8192. Play Girls gives you 16384. Then the really big hitters like Raimais/Play Girls 2 offer you 32768 and Champion Wrestler a whopping 49152! Of course that storage space is split between tiles and sprites but it's still not bad for what is essentially an 8-bit system. I think you'd be hard pushed to find another 8-bit system that offers you that many tiles/sprites.

 

[Sokurah]

What a crappy demo. I mean "as crappy as those obscure games shown are actually obscure".

...so, not crappy.

Great demo. I really like the screen disolves.

Did you use PSG component of the YM2203 or the FM component for sound? Because it seems like you have the same problem with the sound as me.

 

[cmonkey]

Cheers Soren

I'm actually using the same sound effect playing routine as the one you sent me. I'm having the same problems you were, where some sounds play really loud and some are really quite. Yet in MAME they're all the same, perfectly normal, loudness level. One of the many joys of dev'ing in MAME and seeing totally different results when deployed to hardware! Grrrr!!

 

[Alpha1]

Looking good dude!

My fav is still the Gradius!

 

[cmonkey]

Haha! You and your Gradius!

 

[ben76]

Great stuff mate

If you need anything testing lmk.. thats about the only way I could help

- Ben

 

[muddymusic]

Just watched both vids. I'm trying to get a basic understand of how this arcade sh*t I like actually works, and these vids are very helpful!

 

[Hurray Banana]

Yeah 4096 tiles is a lovely luxury, that I intend to make use of when I finish playing with pac hardware (I don't think I'll ever stop playing with out tho') and move over to the L system.

4096 is the base level configuration of L System (Plotting). If you move up to Puzznic you get 8192. Play Girls gives you 16384. Then the really big hitters like Raimais/Play Girls 2 offer you 32768 and Champion Wrestler a whopping 49152! Of course that storage space is split between tiles and sprites but it's still not bad for what is essentially an 8-bit system. I think you'd be hard pushed to find another 8-bit system that offers you that many tiles/sprites.

Woot Woot I've got a puzznic

 

[ben76]

Yeah 4096 tiles is a lovely luxury, that I intend to make use of when I finish playing with pac hardware (I don't think I'll ever stop playing with out tho') and move over to the L system.

4096 is the base level configuration of L System (Plotting). If you move up to Puzznic you get 8192. Play Girls gives you 16384. Then the really big hitters like Raimais/Play Girls 2 offer you 32768 and Champion Wrestler a whopping 49152! Of course that storage space is split between tiles and sprites but it's still not bad for what is essentially an 8-bit system. I think you'd be hard pushed to find another 8-bit system that offers you that many tiles/sprites.

Woot Woot I've got a puzznic

I've got a champ wrestler

It's broke atm iirc though

never really wanted to bother trying to fix it.... till now maybe...

- Ben

 

[cmonkey]

Woot Woot I've got a puzznic

Is it Puzznic WITH naked chicks or WITHOUT naked chicks? If it's the former then you've have 27010 tile/sprite roms on board, which means 8192 tiles. If it's the latter then you'll have 27512 tile/sprite roms on board, which means 4096 tiles.

 

[cmonkey]

Great stuff mate

If you need anything testing lmk.. thats about the only way I could help

- Ben

Thanks Ben

I'd very much appreciate it if you could find 5 mins to test this demo on your Plotting/Flipull board for me. The palette rotation routine pushes the L System VERY HARD as it's constantly rotating all 512 bytes of palette ram every other frame (when palette rotation delay is set to its minimum value). I did notice that the PAL16L8 @ IC12 got *very* hot during extended execution testing (30+ mins), considerably hotter than it does when playing Plotting or any of my other homebrews.

With that in mind I'll fully understand if you don't want to test it, I wouldn't want to feel responsible if you damaged your board through the use of my code. I've no hesitations whatsoever in potentially damaging my own boards through homebrew experimentation but it's unfair to ask others to potentially damage their own boards in the same way. I'm a firm believer in pushing boundaries and experimentation, after all, if people didn't push boundaries or experiment the world would never evolve and advance.

I wouldn't recommend running this code for hours on end (why would you anyway, it's pretty boring to watch after the first minute or so!) but a couple of mins won't hurt.

 

[ben76]

So you want me to fire your latest demo up on my pcb? As I have no problems with doing that.. Will dig out the pcb... wherever the little bugger is lol

- Ben

 

[cmonkey]

I would very much appreciate that dude, but don't run it for more than a few mins. Let me know how it looks. Cheers!