AuthorTopic: Commercial Critique Challenge - Tilevania: pixel's quest  (Read 75300 times)

Offline Arne

  • 0010
  • *
  • Posts: 431
  • Karma: +1/-0
  • Panties.
    • View Profile
    • AndroidArts

Re: Commercial Critique Challenge - Tilevania: pixel's quest

Reply #50 on: April 19, 2016, 04:22:58 pm
I ended up with a noisy turd.

Reasons:
I wanted a different brick size for the background (smaller to suggest distance), but this added extra noise. Could probably throw in more blank/flat tiles.

I didn't want to do leaf blobs for the plants, so I started dithering, thinking it might work on a TV.

Tried to keep the tile count and 4*4 layouts somewhat low, but sort of creeped away doing a few decorative tiles.

Tried to keep the original 3 palettes to the end, but had to lighten the sky to reduce noise and replace GUI palette red with brown as I used that for the lit ground tiles.



*Edited in some more flats. Plant dither still a mess.
« Last Edit: April 22, 2016, 10:52:51 pm by Arne »

Offline mobichan

  • 0001
  • *
  • Posts: 35
  • Karma: +0/-0
    • View Profile
    • Pixel Licker

Re: Commercial Critique Challenge - Tilevania: pixel's quest

Reply #51 on: April 22, 2016, 03:36:16 pm
I've been lurking in here for a while now and man are these impressive.  Sorry to nitpick, but I just wanted to point out to Phoenix that the castle gate/entrance can be jumped past to reveal a hidden item.  So you need to continue the floor past the entrance instead of stopping it at a bush wall.

Offline moket

  • 0001
  • *
  • Posts: 52
  • Karma: +1/-0
    • View Profile
    • human hacks

Re: Commercial Critique Challenge - Tilevania: pixel's quest

Reply #52 on: April 24, 2016, 08:37:05 pm
Inspiring thread  :y:

Gave a try at it, the lack of a saturated bright red in the NTSC palette makes it quite hard.
Went a bit overboard with the tile count...can't draw stairs and props anymore.


« Last Edit: April 24, 2016, 09:49:13 pm by moket »

Offline Probo

  • 0010
  • *
  • Posts: 317
  • Karma: +1/-0
    • View Profile

Re: Commercial Critique Challenge - Tilevania: pixel's quest

Reply #53 on: April 24, 2016, 10:17:47 pm
Wow some great edits in here. that ntsc palette that kasumi provided really lends itself to some grungy, dark visuals. I love that shit. However one of my favourite things about the pal nes I grew up with is bright, weird and sometimes garish palettes so I leant into that :D

I chose Death's level, looked like it could do with a lick of paint:



I had a go at the existing metasprites in the room too, trying to stick to their existing real estate in terms of 8x8 or 8x16 sprites.

I wanted to keep the HUD palette the same as its iconic to me, but I also used the HUD palette on the fire in the top left, which I assume could be animated by switching out tiles? maybe not in the cv1 engine? Kasumi help me XD. in a perfect world it'd be whippable too, like a candle.

enjoyed making the upgrade 'III' in the HUD a bit more readable, after so many years playing this game haha

I haven't checked for errors or tile count :S

Offline Kasumi

  • 0010
  • *
  • Posts: 275
  • Karma: +1/-0
    • View Profile

Re: Commercial Critique Challenge - Tilevania: pixel's quest

Reply #54 on: April 24, 2016, 11:56:23 pm
About Castlevania's Engine: I learned that Castlevania (at least appears to) hardcode which 8x8 tiles are collision enabled. (Most games tie it to larger structures.) So (quite laughably), what makes a lot of these not work is too many unique tiles that are meant to be collision enabled.

I wouldn't consider adding that restriction to this challenge, since it's conceivably a thing they'd have changed if they had art like this. But it's the main thing that made hacking art in no fun.

Probo: You've got 134 non unique tiles, and no palette errors. Getting under tile count would be pretty easy for you. You've got three unique floor brick structures. Could use just one.

As far as animating the fire (assuming its background), it's possible with CHR RAM (which Castlevania uses), but easier with CHR ROM. And I don't remember Castlevania updating tiles that way mid level anywhere. (Could be wrong. I should note it'd be a bit of a different thing than the breakable walls that exist in the game.) 
« Last Edit: April 24, 2016, 11:58:41 pm by Kasumi »
I make actual NES games. Thus, I'm the unofficial forum dealer of too much information about the NES

Offline Probo

  • 0010
  • *
  • Posts: 317
  • Karma: +1/-0
    • View Profile

Re: Commercial Critique Challenge - Tilevania: pixel's quest

Reply #55 on: April 25, 2016, 11:05:02 am
Oh that's interesting, so is that restrictive because it would have to hold a long list of collision objects in memory, memory that could otherwise be used for more important things? Or it'd slow performance because it has to check through the list every time it checks for collision or something?

Thanks for checking the pic for me, I might edit it to bring it under later in the week.

Offline Kasumi

  • 0010
  • *
  • Posts: 275
  • Karma: +1/-0
    • View Profile

Re: Commercial Critique Challenge - Tilevania: pixel's quest

Reply #56 on: April 25, 2016, 10:54:38 pm
Collision detection on NES is the same as any other tile based collision. Choose a point on the screen, see if "what's there" is collision enabled. "What's there" being whatever structure contains if a thing is solid or not.

Castlevania's levels are built out of 32x32 (four by four 8x8 tiles) structures. They could have stored collision info with these structures. If having collision as fine as 8x8 was important to them, that means they "needed" a bit per tile (solid/nonsolid). Which is two extra bytes (16 bits at 8 bits per byte) per 32x32 structure.

Or... one could use the tile's number to store that information for "free." If the first 128 of the 256 tiles in the set are collision enabled, and the last 128 are not I can use the highest bit of the tile number itself to check if something is solid or not. Obvious caveat that you can't have more than 128 of either solid or non solid things. To get around the caveat, you could store a byte with every set to define the numbered range of collision enabled tiles, which is not "free" but is more versatile, and also way better size wise than two bytes per 32x32 structure.

Castlevania certainly appears to (I'm not getting out the debugger) tie its collision to 8x8 tiles in some way, which depending on how they did it might have saved them a fair bit of ROM. Less to do with RAM or performance since they (probably) didn't save a lookup.

But most games do store their collision information with some larger structure. For one thing, the same tile might need to be used in multiple contexts. (A solid color tile that is not solid/water in on structure, and solid/grass in another.) To do something like that, you'd need to have duplicate 8x8 tiles in your set. Generally you're storing more than solid/nonsolid, so even if you arrange tiles to save one way, it's tough/impossible to arrange them such that a range check will work for all the information about a tile. A lot of games store collision info with some 16x16 structure because NES palettes can't really be set per 8x8 (barring rare hardware). Since there already had to be definitions for palette at that size, why not put collision there too and be done?

So it didn't occur to me that Castlevania would... not do that until I was a bit of the way through putting graphics in, and then Simon kept falling through the floor. Changing the tile numbers of the floor alone seemed to make him fall, even though it was the same structure. May as well just post this:


The graphics I was working with were Phoenix849 and Prism's. Though Phoenix849's may not appear in the gif.

End note: Like I said, haven't got out the debugger to check the game. So if anyone wants to do that to tell me I'm wrong, or find out where/how it's storing collision info feel free. With that knowledge, hacking the game is... easier. I wish I had spent the time I spent hacking replicating the movement in my own personal engine. Because... getting the graphics in that was easier.
« Last Edit: April 25, 2016, 10:57:20 pm by Kasumi »
I make actual NES games. Thus, I'm the unofficial forum dealer of too much information about the NES

Offline Probo

  • 0010
  • *
  • Posts: 317
  • Karma: +1/-0
    • View Profile

Re: Commercial Critique Challenge - Tilevania: pixel's quest

Reply #57 on: April 26, 2016, 08:50:16 pm
that's really interesting, thanks. I love the idea of programming nes games but its way over my head! Maybe ill make a nes-faithful game in game maker one day :P

I need to make some more castle tiles for my game at some point, i got some ideas doing this!

Offline TOMOYA

  • 0001
  • *
  • Posts: 6
  • Karma: +0/-0
    • View Profile

Re: Commercial Critique Challenge - Tilevania: pixel's quest

Reply #58 on: April 27, 2016, 01:28:54 am
Great info there, Kasumi. Since I discovered this thread I was wondering about the posibility of getting these mockups into the real game, and how difficult it would be. It would make a great long-term project for the forum if we managed to get a lot of people interested on it.

I love the idea of programming nes games but its way over my head!
Totally agree with this. I also have an unfinished Castlevania engine that I started some years ago, after realizing how much time, work and how many headaches would take me to put all my ideas on an actual NES hack.

Offline SeanNoonan

  • 0001
  • *
  • Posts: 2
  • Karma: +0/-0
    • View Profile
    • Sean-Noonan.com

Re: Commercial Critique Challenge - Tilevania: pixel's quest

Reply #59 on: May 02, 2016, 09:43:59 pm
I don't think I'm ready to contribute just yet, but man, I'm seriously impressed with this stuff.

It would be great to see a remake like this :)