WIP Super ThunderCats: The Lost Eye of Thundera -BOR on DC

[Dakangel]

Will do! Is it a the more the merrier kind of thing? Should I try for 26 panels or just cut it up a bit, say 4 or 5 sections? Or is it another trial and error situation?

Panels were developed to easy create big and large stages without requeiring a lot of images for it recycling the already defined ones.
But the lazyness of the creators, the adition of walls, holes, platforms and the lack of proper tools to visualize what they are doing, it took them to just stamp the entire stage in one big panel regardless the implications in the memory allocation.

Its not a requirement nor an obligation but if you want ,you can optimize the stages with the reuse of certain parts that you see repeated in the panels;
If one chunk is equal to other just reuse it as many times as you see convenient, remember that panels can vary in size.

[marchegiano]

Bor format is just for the main sound track nothing else leave all the other wavs.. I included one ogg converted correct for the first level sound effects leave the way they where... The wav2bor i included does stereo and compiled to work on windows 64bit.. there is no more then 10 sound track files so not really need a batch convert for that.. I know mine works as all mine i have used to do what i show in the video..

Only replace the ogg files the rest can stay . Just this panel size it be fast and not run out of ram

Sorted, I had converted the wavs to oggs but in the text I changed the wav directory commands to bor instead of the ogg commands to bor. No clue why i did that, just wasn't thinking at the time.

I have fully converted music running perfectly now. That you so much!

Level 2 is still very slow, because I've done nothing about that massive panel yet, but level 1, 3, and 4 run smooth.

It's hard to see fps because of the hud panel but you can see enough to know i'm getting in the 120s mostly. It flash around a lot but the lowest figure I saw on level 1, 3, or 4 was around 70 fps.

C5D89DD3.0.png (49.6 KiB) Viewed 5193 times

C5D89DD3.1.png (48.94 KiB) Viewed 5193 times

C5D89DD3.2.png (70.09 KiB) Viewed 5193 times

C5D89DD3.3.png (54.94 KiB) Viewed 5193 times

C5D89DD3.4.png (42.13 KiB) Viewed 5193 times

The boss fight is "level 3". I forgot to screen the start and end of level 4 and didn't play 5 because I wanted to get on level 2's case before I got too far ahead of myself....also I lost pretty much all my lives pushing through the 3 fps garbage that is lvl 2.

I think I'm getting close to completed! This is super exciting! I think it's time I cut that panel up and get level 2 up to the 100s if possible


I never did completely re-pal the entire game either so it's probably time for that too. I re-palette-ed the game once and really effed everything up, then I did it, all of it, again and got to the point where I could see a lot of screens but because I had my translucent in my palette multiple times the hud panel was solid and not translucent above the hud. So I repaled that one image and then the next visual bug came at level 2 and I repaled it, all the images for level 2. But that's it, most of the game is still on my second attempt at palette. both the 2nd pal and 3rd pal are the same until the end. the only difference is the 3rd pal only has one instance of translucent in the color map where as the 2nd begins and ends with same color. So, I think, could be wrong, if I went through the entire game applying the 3rd pal that should get me some better performance as well, right?

Literally the only difference is the last five slots of color. Both start with the 255 0 255 color. the 2nd ends with have five slots of 255 0 255 and the 3rd I changed those five slots to all black, 0 0 0. The entire game is currently in the pal that begins and ends with 255 and if that's okay or making it end in 0 won't help me, you know, I'd rather not go through every single image again, if I can. If it can't be avoided and would definitely improve performance I'll just suck it up and get it done.

[marchegiano]

Will do! Is it a the more the merrier kind of thing? Should I try for 26 panels or just cut it up a bit, say 4 or 5 sections? Or is it another trial and error situation?

Panels were developed to easy create big and large stages without requeiring a lot of images for it recycling the already defined ones.
But the lazyness of the creators, the adition of walls, holes, platforms and the lack of proper tools to visualize what they are doing, it took them to just stamp the entire stage in one big panel regardless the implications in the memory allocation.

Its not a requirement nor an obligation but if you want ,you can optimize the stages with the reuse of certain parts that you see repeated in the panels;
If one chunk is equal to other just reuse it as many times as you see convenient, remember that panels can vary in size.

I'm a little bit guessing the details

Unaltered level.txt:
panel data/bgs/level02/panel.gif
background data/bgs/level02/background.gif

bglayer data/bgs/level02/bglayer.gif 0.4 0 1 1 0 0 -1 1 1 0 0 0 0 0 1

order a

-------------------------


What I need to be doing is cutting the original gif, to make this easy, let's just say in half for now, giving me a panelA.gif and a panelB.gif

Then I would change the panel line in the above txt from panel to panel a and panel b along with the order?

Like so:
panel a data/bgs/level02/panela.gif
panel b data/bgs/level02/panelb.gif

background data/bgs/level02/background.gif

bglayer data/bgs/level02/bglayer.gif 0.4 0 1 1 0 0 -1 1 1 0 0 0 0 0 1

order ab

----------------------------

Have I misunderstood anything?

[Dakangel]

What I need to be doing is cutting the original gif, to make this easy, let's just say in half for now, giving me a panelA.gif and a panelB.gif
Then I would change the panel line in the above txt from panel to panel a and panel b along with the order?

Have I misunderstood anything?

From the official Openbor manual:
https://dcemulation.org/?title=OpenBORM ... l_Settings:

"You can have up to 26 panels in a stage. They are labelled by OpenBoR from a to z.
This is how OpenBoR thinks of them, don't actually put those letters in the panel declaration.
{panel#} is a letter a through z which corresponds to a panel.
There should not be spaces between the panel declarations (ex. order abcabcada, not order a b c a b c a d a).
The same panel can be used more than once."
__
The transparent color is taken only from the very first palette slot, if you have it in more spaces you only are wasting slots for other usable colors.
The module itself seems to not use more than 128 colors or even can be abruptly reduced to 64 colors for the entire game.
That is why you require a master collage image ,to take from it the global pallete.

[marchegiano]

Tomorrow I have to go back to work so my updates are going to slow down a bit but I'll be working on it as I can.

I sliced those panels up all sorts of ways and got them to load all sorts of ways too, but, what I didn't think about is how changing the panel is going to effect everything else on the level. No matter how I chopped them up or loaded them in I didn't get the panels to line up with collision and background effects.

I went a bit a head and looked files for the later stages and it looks like not only does the majority this game rely on massive single file stages but also the next five or so games I opened up just to look at did too.

Definitely something anyone looking to optimize for DC is going to have to learn.


One of the times I got a semi-successful multi-paneled stage was with panel then directories with only a space to separate them. Like this:

panel directory/example/panel.gif directory/example/panel2.gif directory/example/panel3.gif

order abc

but this too seemed to work just as effectively:

panel directory/example/panel.gif
panel directory/example/panel2.gif
panel directory/example/panel3.gif

order abc



Outside of that, I repaled all the gifs. Nothing left to do but learn how level design works then tinker with it to load better on DC.

[marchegiano]

I've been tinkering, don't want anyone to think I gave up, I'm just a bit busy at the moment. This will be an underwhelming update.

I haven't quite figured out the panel bits but I did learn some interesting things about how they all come together. The very last panel is always fine but the panels before are either discolored or floating or a bit of both. No matter how I've loaded them they don't seem to be actually bringing up my fps. This particular level has a few floating elements so I thought I'd try eliminating them and running on a more basic level then bringing them back in once I understood better but eliminating them gave me no better fps either. I'll keep cracking at it anyway, level design is interesting and I would rather understand it than not. Once I get it sorted if it still hasn't helped performance I'll focus back on fps and what other things I can do to increase performance.

Outside of that. All this porting work is giving me an understanding of the engine itself I did not originally anticipate which has me thinking I might start trying to make games myself. It'd be cool if there was some OpenBOR devs who focused on Dreamcast. So maybe Thundercats has become a short-term goal, or like a schooling, and the long term goal will become developing for DC. If I can, I will. We'll see how it all goes, at the moment I think it's best I don't get distracted and just crack on learning to optimize made games for DC.

Thanks again for all the help everyone, will keep you posted.

[marchegiano]

I got my issue with discolored panels sorted. I didn't realize as I was cutting them up and saving them Gimp was changing the palette of each of them except of course the last one because the last one was always just cuts away from where as the other before I had to paste and did not follow up with a pal check or swap.

I didn't realize how much art and understanding art tools would go into this. I'm learning way more than just games. It may seem very basic to most of you guys but I did not realize cutting an image might change its palette when you paste the cut to a new canvas.

[Ian Micheal]

I got my issue with discolored panels sorted. I didn't realize as I was cutting them up and saving them Gimp was changing the palette of each of them except of course the last one because the last one was always just cuts away from where as the other before I had to paste and did not follow up with a pal check or swap.

I didn't realize how much art and understanding art tools would go into this. I'm learning way more than just games. It may seem very basic to most of you guys but I did not realize cutting an image might change its palette when you paste the cut to a new canvas.

Yes i had a lot of trouble with gimp doing this .. But you should see quite a bit of speed up compare to when you started i did with simple running in 8 bit pal

Levels.txt for each level must have some spaming of functions that is slowing it

Great to see you doing this

Lot more work then people think

[dubcity]

If you need help, let me know after the game conversion Eternal Champions: The Thin Strings of Fate and G.I. Joe: The Power of Cobra

Si necesita ayuda me lo deja saber despues que termine la conversion del juego Eternal Champions: The Thin Strings of Fate y G.I. Joe: The Power of Cobra

I've been working on eternal champions also. did you get the 60mb of opening scenes to load up? I took some of the scenes out and it still is hanging on loading screen.

[dubcity]

I've been working on eternal champions also. did you get the 60mb of opening scenes to load up? I took some of the scenes out and it still is hanging on loading screen.

The game is 100% functional, the whole code was optimized, without losing content since Sergio "Krauser" Nuñez asked me not to eliminate anything. This game will be launched on October 10, 2021 on my website.

El juego es 100% funcional, se optimizo todo el codigo, sin perder contenido ya que Sergio "Krauser" Nuñez me pidio que no eliminara nada. Este juego se lanzara el 10 de octubre de 2021 en mi pagina web.

cool. look forward to seeing it.