91
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 20 Nov 2023
91 points (100.0% liked)
Games
31800 readers
430 users here now
Welcome to the largest gaming community on Lemmy! Discussion for all kinds of games. Video games, tabletop games, card games etc.
Weekly Threads:
Rules:
-
Submissions have to be related to games
-
No bigotry or harassment, be civil
-
No excessive self-promotion
-
Stay on-topic; no memes, funny videos, giveaways, reposts, or low-effort posts
-
Mark Spoilers and NSFW
-
No linking to piracy
More information about the community rules can be found here.
founded 1 year ago
MODERATORS
Well I have a lot of problems with how people design games so I don't really buy stuff anymore, plus I haven't really seen a lot of stuff that focuses on vector (esp textureless). In other words it's pretty niche even for indie, and discoverability generally isn't great even on the best day.
I'd probably have more luck doing it myself, I've done a few 2D things (meme made with Godot 3.X, 4.0 eye animation, not-yet-in-4.X test of someone elses' PR) but I'm not a dev and I don't have much energy or many ideas.
Ah. I think the problem there is that pure vectors can be much harder to work with. it's hard to make something that looks good with purely vector based approaches, especially as your scenes get more complex.
I don't think that's it. For 3D the workflow is already there and vertex colors are powerful (though usually used for shaders or other effects like terrain-based sounds). Even going for Spyro's approach (esp. grayscale textures that disappear with LoD so it's just color) wouldn't be too bad as I imagine its music/voice is actually what takes up the most space (newer audio compression or MIDI-like music would reduce that), though a more minimal/stylized look could make it a lot easier. Certainly some things are more suited for it than others.
I could say a lot of technical reasons for or against this workflow, but I think the biggest is just that it's something that people don't think about or would rather have photorealism or blocky pixels instead (or at least that's a large chunk of the market). Vertex lighting is cool but doesn't have much use over modern lighting (if it did, it'd be very niche) and developers often don't really care about optimization much, instead telling players 'upgrade your PC'.
(admittedly my experience with 2D vector seems less supported as far as editors and AA, though I'm not sure if Godot's clip children feature has an equivalent in 3D or if you'd just need to use meshes/rigging more cleverly... which is fair, I'm not aware of non-skeleton rigging tools in Godot's 2D either)
3D vectors can work well, but as an artist you are often better suited by going with the N64 approach. Due to technical limitations, textures were often used quite sparsely, with vertex shading providing the main colour and textures providing details. It was especially prevalent in games with cartoon art styles such as Mario 64 and Banjo Kazooie, but even games like GoldenEye used this trick to a degree.
The problem with pure vertex is, the more detail you add to stuff, the more calculations you are doing, and they can really add up. The same is kinda true of bitmap (in terms of resolution) but the problem doesn't scale the same way. With that said, it will work well for a cartoony/anime art style where massive amounts of detail isn't necessarily needed, and shaders can be used to complete the look.
I mean yeah we're mostly on the same page... but it should be clear that I'm not suggesting crazy detail everywhere, mostly just being a bit more intentional with model design when possible to integrate vertex color (or another old technique, use multiple objects when it means a simpler mesh). And I mentioned Spyro's texture/LoD system which is good, was going to mention sprite usage and also Crash having only 2 textures (shoes, back) but was too wordy (also Crash taking advantage of a linear camera for custom culling and view-specific models).
I'd say it's really good to give variation (and unique-ness) on detail and effects that way every tiny thing you decide to add isn't a fixed workload. Or in some cases the opposite approach, a more re-used/modular design for certain things like characters.
The problem with textures (aside from data w/high-res/high-color, resolution dependency, and workload) is that when you play an older game at modern resolutions (higher internal res or even just a Flash game) the elements that were designed for older resolutions/displays are really apparent next to the meshes that scale perfectly. Particularly if it's a GUI or pre-rendered cutscene (sometimes other random stuff). Textures on meshes can still be a really solid aesthetic for the environment/characters.
Also generated textures (see .kkrieger for an extreme example) might be a potential fix for the drawbacks, or something like textures that are designed to be used with an upscale filter (or in a similar way, maybe converting to SDF textures).
Yeah, we are definitely on the same page. What they did with classic Crash with minimal textures on the character, they also did the same thing with Mario in 64 and Banjo in Banjo-Kazooie.
You could do a similar approach to environmentals for a decent effect too, such as having the texture essentially act as a very restricted form of bump map for the vertex shaded polygon. Goldeneye did this a fair bit.
I would be interested in seeing this kind of art style make a return.