772 replies to this topic

[grafixbmp]

could you perhaps make only one calculation in reference to both players goal post knowing that both sets reside at opposite ends of the same playing field and share a coordinate? Sorta calculate for one then invert the values and save it in case the other player's screen needs them for the opposite side? That is assuming your calculations are based on the coordinates of the game arena grid.

I'm sorry. I probably don't understand the full context. I'm gonna re-read the posts.

Edited by grafixbmp, Tue May 24, 2011 6:38 PM.

[roland p]

FYI, I've descided to rewrite the thing. I've became smarter the last 2 years so I liked to redo it. Code I write now will be more efficient than code I've written a few years ago when I started 6502 assembly. I hope I can show something nice soon.

[Random Terrain]

roland p, on Tue May 31, 2011 4:02 AM, said:

FYI, I've descided to rewrite the thing. I've became smarter the last 2 years so I liked to redo it. Code I write now will be more efficient than code I've written a few years ago when I started 6502 assembly. I hope I can show something nice soon.

Every year I think I'm not as stupid as the year before, but that means next year me will think this year me needs a helmet and a drool bucket.

[Keatah]

roland p, on Tue May 31, 2011 4:02 AM, said:

FYI, I've descided to rewrite the thing. I've became smarter the last 2 years so I liked to redo it. Code I write now will be more efficient than code I've written a few years ago when I started 6502 assembly. I hope I can show something nice soon.

I have no doubt that 2600 BallBlazer will be better than ever!

[Thomas Jentzsch]

roland p, on Tue May 31, 2011 4:02 AM, said:

FYI, I've descided to rewrite the thing. I've became smarter the last 2 years so I liked to redo it. Code I write now will be more efficient than code I've written a few years ago when I started 6502 assembly. I hope I can show something nice soon.

That's quite a brave decision. Especially, because you will miss a lot of feedback, since in the beginning you have nothing new to show.

I wish you good luck.

[roland p]

Thomas Jentzsch, on Wed Jun 1, 2011 4:53 AM, said:

That's quite a brave decision. Especially, because you will miss a lot of feedback, since in the beginning you have nothing new to show.

I wish you good luck.

Thanks,

At least I'm having fun in doing it, which is the most important.

[Propane13]

Let us know if you need anything.
That, or feel free to rant about stuff that's not working from time to time.
This can be a great forum for venting when your cycle counts don't seem to work.

-John

[SoundGammon]

I have a version for the NES...

[roland p]

Propane13, on Thu Jun 2, 2011 7:42 AM, said:

Let us know if you need anything.
That, or feel free to rant about stuff that's not working from time to time.
This can be a great forum for venting when your cycle counts don't seem to work.

-John

Thanks,

Currently I'm running some proof-of-concept kernal and I'm spreading out the code (or placeholders) over the banks of a 32k scheme. I also want to use superchip memory (F4SC). As long I can get away with this bankswitching scheme, I don't want to touch the fancy ones yet (DPC+). Also the f4sc is more 'authentic', less cheating. That said, I think the Harmony guys did a great job.

[Keatah]

SoundGammon, on Thu Jun 2, 2011 10:12 AM, said:

I have a version for the NES...

and??

[roland p]

Time for an update.

The new playfield is almost finished now.
It is now full width and 48 scanlines high (24 double lines)

The sky is also 24 double-scanlines high. During the sky, it will determine which lines to draw in the checkerboard kernel. It also does anti-aliassing. So no tile-setup code during overscan/vblank. It also does some skewing for the last 5 lines for movement of the first line of the checkerboard kernel (this moves 0-5 cycles). 73 cycles left in each (double)scanline now.

The checkerboard kernel is now split up into several mini kernels which can draw some border if needed. There are 10 kernels for the first line (1-5 brown tiles left, no brown tiles and 1-5 brown tiles right) and for the last part there are 5, and some in between.

There is a bit of physics, you can bump left and right. Don't go to much forward or backwards

Next thing to do is to push some variables into superchip mem and then the rotofoil display.

Attached Thumbnails

• 

Attached Files

•  rom.zip   3.98K   88 downloads

Edited by roland p, Fri Jun 24, 2011 6:22 AM.

[Propane13]

You are the man!

This looks really, really great.
Anything you're looking for comments on, or are you just showing off what you've done?

-John

Edited by Propane13, Fri Jun 24, 2011 6:29 AM.

[roland p]

Propane13, on Fri Jun 24, 2011 6:29 AM, said:

Anything you're looking for comments on, or are you just showing off what you've done?

Just showing off But any comments are welcome.

I hope the anti-aliassing isn't 'too fancy'.

I saw there are some glitches when the border is appearing on the left, it has to do with setting the COLUBK register during horizontal blanking. They don't appear in z26, only in Stella. Also z26 doesn't recognise the f4sc (32kb + superchip) automatically, any help is welcome...

Edited by roland p, Fri Jun 24, 2011 6:42 AM.

[raindog]

roland p, on Fri Jun 24, 2011 6:41 AM, said:

I hope the anti-aliassing isn't 'too fancy'.

I don't think I'd call it "fancy" so much as "groundbreaking".

At this rate it might end up looking better than the C64 version

[Thomas Jentzsch]

AA on the good old 2600? I really have to try this one out!

[tschak909]

My eyes just popped out of my fucking head.



Whoa.

Wow, man. Incredible! Would love to see source code at some point!

-Thom

[atari2600land]

Just awesome! I haven't played Ballblazer before, but if this ever comes out, I would buy it just because it looks spectacular.
I hope at one point, the second player will control the bottom half, right?

[Nathan Strum]

That looks amazing! But I'm completely confused.

I turned on fixed debug colors in Stella... and there was nothing. Just background.

You're doing that all with just the background?!?

(This also has the makings of a 2600 version of Space Harrier...)

Thomas Jentzsch, on Fri Jun 24, 2011 9:09 AM, said:

AA on the good old 2600? I really have to try this one out!

Meltdown does sort-of an anti-aliasing effect as well. It's pretty cool.

[SoundGammon]

Keatah, on Fri Jun 3, 2011 2:53 PM, said:

SoundGammon, on Thu Jun 2, 2011 10:12 AM, said:

I have a version for the NES...

and??

...not as good as the 7800 version!

[roland p]

Nathan Strum, on Fri Jun 24, 2011 8:41 PM, said:

That looks amazing! But I'm completely confused.

I turned on fixed debug colors in Stella... and there was nothing. Just background.

You're doing that all with just the background?!?

Yes, there are only background color changes.
But that was also in the previous version, so not exactly new. The goal of this version, which is rewritten from scratch, is to be more optimised and to have cleaner code.

[potatohead]

Excellent!!

[Nathan Strum]

roland p, on Fri Jun 24, 2011 11:34 PM, said:

Yes, there are only background color changes.
But that was also in the previous version, so not exactly new. The goal of this version, which is rewritten from scratch, is to be more optimised and to have cleaner code.

Very cool. I'd looked at the previous version before, but I guess I'd never checked to see how it was drawn (or had forgotten). Pretty impressive stuff!

[cd-w]

Nathan Strum, on Sat Jun 25, 2011 3:49 AM, said:

roland p, on Fri Jun 24, 2011 11:34 PM, said:

Yes, there are only background color changes.
But that was also in the previous version, so not exactly new. The goal of this version, which is rewritten from scratch, is to be more optimised and to have cleaner code.

Very cool. I'd looked at the previous version before, but I guess I'd never checked to see how it was drawn (or had forgotten). Pretty impressive stuff!

It is quite a low-resolution technique, as the maximum speed that the colours can be changed is 3 cycles (9 pixels), starting on any 1 cycle (3 pixels) boundary. However, it works very nicely here and it leaves the playfield and sprites completely free for other purposes.

Nice work!

Chris

[Thomas Jentzsch]

Even with PAL colors it still looks impressive.

[Stephen]

roland p, on Fri Jun 24, 2011 6:13 AM, said:

Time for an update.

The new playfield is almost finished now.
It is now full width and 48 scanlines high (24 double lines)

The sky is also 24 double-scanlines high. During the sky, it will determine which lines to draw in the checkerboard kernel. It also does anti-aliassing. So no tile-setup code during overscan/vblank. It also does some skewing for the last 5 lines for movement of the first line of the checkerboard kernel (this moves 0-5 cycles). 73 cycles left in each (double)scanline now.

The checkerboard kernel is now split up into several mini kernels which can draw some border if needed. There are 10 kernels for the first line (1-5 brown tiles left, no brown tiles and 1-5 brown tiles right) and for the last part there are 5, and some in between.

There is a bit of physics, you can bump left and right. Don't go to much forward or backwards

Next thing to do is to push some variables into superchip mem and then the rotofoil display.

Awesome! Glad to see you are back to working on this. I have been watching the thread since day one.