LOW-POLY ART
polycounter lvl 19
This is the sweet spot for your low-poly models. Post 'em if you've got 'em!
Low-poly hasn't really been a requirement in the games industry for a long while now. This thread is for low-poly art style appreciation, so please take note of these rough guidelines:
Here's a handy list of ways to make your art look right in mainstream 3D software: Low-Poly Art Style Guide
Low-poly hasn't really been a requirement in the games industry for a long while now. This thread is for low-poly art style appreciation, so please take note of these rough guidelines:
- Keep models under 1,000 triangles.
- Scenes are fine, if all models are low poly.
Here's a handy list of ways to make your art look right in mainstream 3D software: Low-Poly Art Style Guide
Replies
i actually really like the vibrance you brought to the shadows. And it's done in a non-cliche tf2 ripoff way as well. You've managed to make fairly boring objects look aesthetically pleasing. Sick job
It seems you're quite the war buff
Thanks for the info though, I always appreciate new knowledge.
Given what you said, why do you think the trend changed towards the far more monotone and muted planes we see today? Because modern weapons tech can easily track a plane regardless, and the more muted, the less visible?
Hey snader, thanks for the reply. I experimented with per face tiling (or in this case, just stacking) actually as per your first example, you're right it can save a lot of UV space. The only downside though is it would often lead to using a few more polys, for example, the top and bottom sections would have to be cut in half, adding another 16 triangles. When you have an area with an even distribution of polygons, it's perfect though.
Looking at the car UVs now, you're right, it's ridiculous
Regarding floating geometry - I agree that it would be good to use when feasible, as it saves a lot of unnecessary splits. The reason I'm reticent to do it that much though, is that these models are being made for a software 3D engine with no z buffer, just a basic painter's algorithm, which can lead to z fighting etc between neighbouring polys, ESPECIALLY when they differ in size significantly, such as the lights and the car bonnet.
With smaller polys, and especially when they are of similar size, I shall use floating geometry though. I think your example would be fine, and if you notice, I didn't boolean the legs of the snowmobile to the skis, the skis are seperate. Incidentally, pro boolean seems to work much better and more consistently than just boolean, even for single objects, I have no idea why.
Regarding the initial creation of UVs, do you think pelt mapping would be unnecessary for these kind of polycounts? What about for characters? (which is something I'm going to try next).
My method is just select all faces, flatten mapping, then arrange them. I'm curious about all of your workflows for low poly UV mapping.
Looking pretty good, but you might want to lose the lines on the window. Right now it looks like you just overlayed those windows on top of the stone (because of the lines).
Well this is well offtopic, but to answer as succinctly as possible, modern trends focus on obscuring silhouette, not direction, although you could say the old designs were the same in that they were hiding the geometry of the object, it's just that they didn't also use the noise of the object's surroundings to further add to the confusion. "Modern style" camouflage did exist in the first worldwar, alongside this whacky shit, so i suppose they were just experimenting on both in parallel, before ultimately deciding one way was more or less superior
edit: and so this post isnt TOTALLY offtopic, i'm working on revamping an old texture to be more appropriate for custom paintwork
So I couldn't resist posting this one:
I made this one for the community of http://www.hiveworkshop.com.
You can actually download the model as a Warcraft 3 model to play with it ingame.
T.
Yeah I forgot I need to do stitches and contrast at the folds. I don't know what you're referring to with the hat, but I also have to do that loose stitching or whatever it is that goes around the rim of it. I'm having trouble getting the hat to fit. Are you talking about it being at an odd angle?
when it should look like:
?
Since I have never worked on a really, really lowspec game this leaves me a bit confused. Surely it just a matter of painting the thread bit of the texture so that it can be horizontally tiled and shove all your thread polys in there. This would work regardless of the length of the individual polys and would remove any need for extra splits.
There could be some limitation in tiling in your engine that I'm not aware of.
So in regards to tiling, I have to use pretiled sections. Which is slightly annoying, it would be nice to tile an entire map over say one wall of a building etc, just one tile piece as small as 128 or 64, which would be a seperate texture to the main model, and would tile many times within just 1 poly.
But having to use pretiled sections of a map (and thus, bigger textures, and more RAM usage) is still a much better trade off for the performance I gain by only having 1 texture, and even sharing textures between models.
On the map I posted, the track bit of the tex is far too big though you're right, it could be half that width, with them all stacked. Basically it only needs to be the same ratio as the longest polygon, which is the top and bottom of the tracks. The rest could be stacked within that space.
Let me give you an example to clarify what I'm talking about.
This is a model by my colleague, who is a graphic artist, learning modelling. See the wooden plank, well that is sharing the same texture as the castle and surroundings, using a pretiled section of the UV map.
I could give it it's own texture, just the square repeating wood texture, and tile it vertically, using say a 64 by 64 map or less.
But if it's sharing the UV map with the neighbouring objects, I have two options - what it's currently doing, which is using a pretiled section of the map, OR I could have just a single tile piece on the map using less space, but that would be mean subdividing the plank vertically into square polys, then stacking them over the tile. Which would mean an unacceptable amount of extra polys, so my preferred solution is to use a pretiled section.
By "pre-tiling" with the same repeating texture across a large area of space you ARE actually just wasting texture space and thus RAM, resources, etc, where you could be making better use of it.
That's pretty opinionated on my part, and perhaps you are going for a certain "look", but wasted UV space is wasted UV space...
Having said that, for some things I'll keep it very simple, using just a pretiled section for a stylistic choice (for an arcadey or Nintendo esque look). However, I'll at least bake or draw lighting onto it, to not make the extra UV space a complete waste.
here is my monk
DS - diffuse only
PSP - diffuse only
pre 3GS - (maybe pre-3G) diffuse only
Crackberry and Palm - 2D only?
Haha. yeah. absolutely agreed!
Besides the high resolution it looks like a lot of wasted (unused) space in your texture. At first you might use the same part of texture for the two sides of the blade. And then... more than half of the texture is designated for the hilt?
OK, it`s made with the focus on speed but I think this shouldn`t be a reason for wasting, right? ;-)
Cause its fun, challenging and it helps me get ideas or work out faster as I like to take my sweet time doing things.
In this case, i don't think it was worth it. You made an elongated cube (tweaked), that is cube mapped with basically no texture. It's not really pushing you in any real direction, as none of it is all that useful. Maybe that's just me though.
Those 20 minutes would have been better spend on modeling a silhouette only and not texturing at all, or on making a concept sketch.
If you want to improve, go in batches of an hour, about 20 minutes modeling, 40 minutes texturing.
THEN work on getting quick.
Anyways, this was just 10 minutes I had left over in my lunchbreak and I felt like modeling something and so I went with something random, with not much pre-planning and came out with this. A basic cube, some extrusions here and there and a quick automatic mapping to make some kind of sword. I didn't have any time left to texture so I thought I would try it at home in more or less the same amount of time. I wasn't going for anything serious with this, just some lighthearted fun.
Anyways, thanks for the comments guys. I'll be uploading something else later on.
edit: I was a bit influenced by this : http://media.photobucket.com/image/Riviera:%20The%20Promised%20Land/ahchess1337/wall_key_l.jpg
Working on the alien for the TD project :poly142:
that probably most of the users on this forum then
Do you have any wires, and a concept? I'm thinking you could distribute your polies a bit more efficiently (without messing up the unwrap much).
tone down the saturation in the gold and add some blue and white bands (overlay layer) that follow the form to fake reflections will look much better
Ask and ye' shall receive:
concept:
http://grungemedia.com/dkseries/My_Work/HordeZone/sketch.png
What he said, if you want to get a bit more optimized you can go a bit further:
-Simpler arm spikes, with a better resembling silhouette.
-Different.. ears, looks like this should about halve the polies spent there
-Hands which resemble the concept more.
Furthermore, you can take out the inside loops of the leg (under the knee and on top of the hip). The forehead indent doesn't need to be modeled, you can weld the middle of the pecs together, and the ends of the mustache can get away with being 3 sided.
This should give you a nice amount of polies back to spend on rounding out the legs, and bulking up the arms a bit.
191 tris, 256x256 texture, 100% self illum, lightning on the blade is kinda lol but eh
Slowly getting back into 3d stuffs
nice, i still remember when i made mine, its a bit more lowfi...
i really just cannot get enough of this kind of art style. mouth watering goodness everytime. amazing work Pierate!
also im new in this site and maya but i thought i would make a simple low poly sword to practice.
any advice on optermization? textures coming soon.
For optimization...I think you can at least halve the amount of segments on the handle without loosing any visible quality...
"v" snap to verts, put the merge and delete tools on your Marking menu/shelf
there isn't a need to triangulate everything, but convert your N-gons into quads/tris and optimize them. that sword can be 120 polys and still look just as good. give it a try and post in its own thread and people can help you better
ah yes wield the verts, i should have thought of that, duh haha. don't worry i shall not post anymore of it here till its done. thaaanks.
tryed that and the shape didnt change but the edges where gone haha i was scared of the fucked up geomatry