Jet's terrain thread


#41

[quote=“Jetrel, post:39, topic:94”]Something from a week or two ago - We finally have an appropriate background/setting to use the rain fx on.

“All” we need to do is make some levels that actually use it.[/quote]

Jesus. How did I miss this? <3 I can’t wait to see this in game!


#42

One of the things that makes our dungeon feel markedly cheaper than the first parts of the game is how we don’t have a parallax background for it. I’m working to correct that; bear in mind this will be a while or so before it’s useable in-game, since this requires something we don’t have right now, to actually be useable: edge tiles for the dungeon walls.

Compared to drawing this tower, though, those should be relatively easy.



#43

It is a rather huge tower. :smiley:


#44

Did some work on the sky; the moon is done except for some cleanup on the interior, and the background now has stars. The tower was also colorshifted to fit better with the now more orange moon.



#45

:oLooks amazing. Per usual.


#46

Yeah - once I actually get edge tiles for the wall bricks done, I’m hoping we can do some castlevania-esque “walk on the castle ramparts” type scenes. This might include some railing tiles, and will definitely include things like pillars and merlons and such.

It ought to be really, really sweet. :-[ I can’t wait.


#47

[quote=“Jetrel, post:46, topic:94”][quote author=RyanReilly link=topic=74.msg1599#msg1599 date=1312103115]
:o Looks amazing. Per usual.
[/quote]

Yeah - once I actually get edge tiles for the wall bricks done, I’m hoping we can do some castlevania-esque “walk on the castle ramparts” type scenes. This might include some railing tiles, and will definitely include things like pillars and merlons and such.

It ought to be really, really sweet. :-[ I can’t wait.[/quote]

DELISH.


#48

Your post seemed as good an excuse as any to post an update. I’ve been working on this more or less nonstop, and I’ve gotten a lot done:



#49

I’m really in awe of your texturing skills. I mean, I’ve seen dozens of decently-blocked, decently-shaded pixel art pieces, but somehow most of them lack good texture (either too anti-aliased, too jagged, too dithered, etc.), which kind of spoils the effect of the piece. But you consistently nail it, no matter if you’re texturing rock, wood, metal, ceramic… it really elevates Frogatto’s graphics to a whole new level.


#50

[quote=“Jetrel, post:48, topic:94”]Your post seemed as good an excuse as any to post an update. I’ve been working on this more or less nonstop, and I’ve gotten a lot done:[/quote]You damn right you’ve been gettin it done. Eyes are tantalized, but they hunger for moarrrrrrrrrrrr


#51

:-[ Shucks. :smiley:

It honestly means a lot to hear you say that. I’m constantly aware of my faults, and sometimes forget that in my stronger suits, I’m really good. Maybe even one of the best out there, but pixel-art isn’t a glamorous enough field for me to have any reminders of the fact, besides nice comments from people like you.

My faults being … one of the things I find “doable but hard/intimidating” is doing really tight, clean stuff. Oversimplified shapes are one part, but it’s especially symbolic textures that I have trouble with (some strong examples of this would be Zelda: LttP, and Zelda: Minish Cap. Also possibly a bit in Advance Wars and newer editions of pokemon). Stuff where instead of doing grass by actually drawing a grass texture, you instead have, say, a solid field of color, and plop the occasional symbol of grass or a flower down to make it register as grass. Another good example would be the same thing with a brick wall, although that particular case feels less like magic because I’ve done it before.

:-\ This probably isn’t nearly as difficult as I crack it up to be, it’s just that that particular style of pixel art isn’t really my most-liked (which would be more like the style of stuff in, say, Castlevania 5, or demon’s crest, or whatnot), so I’ve had very, very little practice doing it. Once I get some practice with it, it’ll probably turn out really easy. In fact, once I bust the ice on this stuff, with my current skillset, a lot of it might be hilariously easy. I just keep tending to accidently skirt around it, because I keep falling into detailing stuff, and once I start, I keep going because there’s usually no hard-and-fast rule saying I’m supposed to not do so, and whatever I started detailing looks cool and I don’t want to throw it out. So, I need to be willing to butcher those sacred cows if I’m gonna ever get any practice. One area this really bites me in is monster design for frogatto. I’ll get over it.

So I wonder what advice I can give on texturing…
One biggie is to always beware of banding, and related to this is to really try not to use multiple levels of AA. Multiple levels of AA can be appropriate in some situations (namely, gradual lighting changes across a curved surface), but a beginner’s mistake is to feel like multi-level AA is always necessary.

That is, AA becomes a reflex. You see pixels of different tones next to each other (greater than, say, 40L in Lab). So you make some intermediate tones, and following good general practice, you space these every 10L. The mistake is getting tunnel-vision on this, and first AAing an edge with one tone, then feeling the need to make another strip of AA with the next tone, and then yet another strip - and pretty soon you’ve got a string of AA that’s like … 5 pixels wide. Which might be fine if you’re shading a gradual curve (say, the dress on wesnoth’s elvish shaman), but if you’re shading any hard edge (like the edge of a cube) … this isn’t appropriate. It isn’t appropriate because it just can’t be 5 pixels wide and not look like anything but a gradual curve. If it’s gonna look like a hard edge at all, it absolutely, positively has to be as close to 1 pixel wide as you can make it.

I mentioned banding, and what comes out of this is - the more complex AA you have, the more impossible banding is to eliminate. If you’re just doing a tight, 1px wide AA, banding is fairly easy to dispel. If you’ve got a spaghetti-mess of AA, though, it becomes inevitable that you’ll have some banding, which besides reducing the resolution of your piece, also imparts a “griddy” texture to that area.

So you always want to fight AA. AA is necessary, but it’s like … uh… sheetrocking mud, or glue. You always want to apply just as much as necessary, and then clean away the excess.

I won’t lie and pretend that AA is something I can manage to fully put off to a complete afterthought - I often will fall into tweaking it early a bit (just like portrait artists will often fall into doing a tight-ish render of the facial expression to see if the whole picture will pan out, before detailing the rest), but I definitely aspire to putting it off till the final rendering stages, whenever I can. When doing stuff, I often will “sketch the whole thing”, then fully render some small sections to feel out if they’re gonna work, then “lightly render” everything, and then fully render everything.


#52

Something I’ve been meaning to do forever, but which I’ve only gotten the pixel-art skills to do very recently: foreground parallax stuff. Mind you of course, parallax is the key word, and that effect is invisible on a screenshot. Rather unorthodox stuff, but then again, that’s because 16-bit consoles didn’t have enough layers to do this (the most I’ve seen is a few really half-hearted efforts in some rpgs like chrono trigger).

There are a few related things I’d like to try (near-field background stuff), but the way this looks thus far is pretty darned rad.

We don’t have any sort of “diablo-style” translucency if something’s in front of the player algorithm, but even if we get one, I think half of the trick to responsible usage of something like this is that I never put these somewhere where they’ll obscure actors by more than a small amount; they’re always largely placed over solid ground.



#53

Ooh… something like in Jazz Jackrabbit 2? Neat. :slight_smile:


#54

People keep talking about these different PC/Amiga games they think I’m imitating, but I’ve never played them. Never played jazz jackrabbit, never played superfrog, never played Sleepwalker, barely played about 10min of commander keen … 3?

I have played a few excellent Mac/PC platformers back when I was a kid, like abuse, dark castle and ferazel’s wand, though. I am also rather influenced by the recent crop of indie titles, such as cave story, legend of iya (unfinished), owlboy (unfinished), and limbo. I’m just saying this so any “armchair videogame anthropologists” or whatever the hell we want to call “the study of which games influenced which” actually have the right ideas in their head about what we’re a spiritual successor to.


#55

I really didn’t want to offend you. I tend to try to be a smart-ass sometimes, sorry. :-[

Jazz Jackrabbit is one of my favorite games, and the sequel had a really neat level editor. I think I was just mocking your claim of being unorthodox. When I think about it, there really weren’t that many platform games with multiple layers of parallax background and foreground. JJ2 had some (I don’t remember the maximum number) layers one could use, in front and after the “main plane”, and for each there was a possibility to set the relative (parallax background/foreground) and absolute (falling leaves, flowing water) movement.


#56

Oh, no, no - I wasn’t at all offended. :wink: I just wanted people to know the actual truth, since it’s the immediate bystanders that inevitably keep the history (in our case, it’s postings on places like wikipedia, tvtropes, and any of the wide world of gaming forums out there - and not just you, but anyone who reads this topic). It’s amazing how much history gets distorted by “misunderstandings in primary sources”.

Typically speaking, snes games could afford roughly 2 layers, because about 2 other layers (of their four tile layers) got eaten up by the “standable tiles” and “heads up display/UI”, and the sprite layer was naturally always tapped out. So pretty much universally, they always seemed to devote both of these to the background. There are some exceptions who devoted their resources differently, but I can’t think of any that did foreground other than to do fog of some kind.

The tragedy of course is that the n64/playstation, which finally transcended these limitations, also ended up temporarily killing 2d-platforming, at least until the media and buying public came to their senses and wrapped their heads around the idea that “gameplay doesn’t go obsolete”. Vendors made some kickass 2d games, and … they were commercial failures, because everything was so hopelessly besotted by not 3d graphics, but 3d gameplay, which was the wild new frontier at the time. :frowning: I think there are actually some real gems from the era that I ought to check out (like Lomax), I just haven’t gotten around to doing the research, getting a good emulator(s), etc. (I have nothing against 3d games, mind you, I just lament that “learning our lesson” that 2d games weren’t suddenly obsolete, came at the cost of putting many of the most talented makers of them, out of business forever.)

I speak of orthodoxy because some (retarded) people have actually said bad things about frogatto for not being “absolutely purist” in design. What exactly that means seems to be different for everyone, but generally it means if I do anything innovative, I’m a bad person. ::slight_smile:


#57

[quote=“Jetrel, post:56, topic:94”]Typically speaking, snes games could afford roughly 2 layers, because about 2 other layers (of their four tile layers) got eaten up by the “standable tiles” and “heads up display/UI”, and the sprite layer was naturally always tapped out. So pretty much universally, they always seemed to devote both of these to the background. There are some exceptions who devoted their resources differently, but I can’t think of any that did foreground other than to do fog of some kind.

The tragedy of course is that the n64/playstation, which finally transcended these limitations, also ended up temporarily killing 2d-platforming, at least until the media and buying public came to their senses and wrapped their heads around the idea that “gameplay doesn’t go obsolete”. Vendors made some kickass 2d games, and … they were commercial failures, because everything was so hopelessly besotted by not 3d graphics, but 3d gameplay, which was the wild new frontier at the time. :frowning: I think there are actually some real gems from the era that I ought to check out (like Lomax), I just haven’t gotten around to doing the research, getting a good emulator(s), etc. (I have nothing against 3d games, mind you, I just lament that “learning our lesson” that 2d games weren’t suddenly obsolete, came at the cost of putting many of the most talented makers of them, out of business forever.)

I speak of orthodoxy because some (retarded) people have actually said bad things about frogatto for not being “absolutely purist” in design. What exactly that means seems to be different for everyone, but generally it means if I do anything innovative, I’m a bad person. ::)[/quote]
For the clarity: JJ2 was a Win9x platform game (didn’t work on XP and later systems, though)… But yeah, I should probably stop glorifying my favorite platformer already. :slight_smile: Although I do recommend seeing this gameplay video (actual gameplay starts around 0:40 mark). Especially the castle (dungeon) level shows multiple layers of parallax movement. They didn’t really defeat the grid graphically, but the visuals are still pleasant to the eye.


#58

I’ve partly-finished the third major rework of the forest houses, which is slightly embarrassing. Took me a while to really figure out my chops on this sort of thing, since I’ve never done large scenery art like this, but I think I’ve nicely settled on a way to do such things for frogatto.

So attached, you can see the very oldest one which is currently the only one public-visible in the current levels, a transitional one which I’m glad I didn’t waste the time to build a few levels around, and then the new one. The new one isn’t completely, totally rendered yet, but it’s almost done, and I’m pretty damn happy with it.



#59

Very cool! I’m impressed.


#60

It looks very cool. :smiley: