Lowpoly Arm Contest! Improve yourself! Win stuff! Exclamation!
polycounter lvl 15
TL;DR: Make a lowpoly (under 200 ▲) arm before the 12th, and win Portal 2: the final hours.
[excuse me if this is in the wrong place - I know there's the LPC forum but AFAIK that's only for official polycount challenges. you can move this if necessary]
Before I went on vacation I challenged people in the lowpoly thread to push their boundaries and make an arm within a limit of 200 polygons - and make it as detailed as possible. My post wasn't very clear however, and lacking in some important information. Such as... a deadline. In this thread I will clarify a bit, and I will also accept the entries. So with further ado, the challenge:
MAKE YOUR BEST LOWPOLY ARM
Nowadays we're hardly bound by polygon limits - even cellphones nowadays are becoming pretty powerful. But that doesn't mean that lowpoly is gone. First of all not all devices are fast yet. Secondly there are games like RTS which have tonnes of things on screen. And lastly - lowpoly is a very fun and educational experience.
I've learned that, often, the best way to improve is to push yourself. To go outside of your comfort zone. To try and work within very limited constraints. So that's what I challenge you to do. Take the narrow limit of 200 triangles and make a fully-featured arm!
ABOUT THE ARM
-it is humanoid, but does not have to be 100% human.
-you may use accessories like sleeves or bracelets, but remember the main goal is to make an arm.
-consists of at least a shoulder, upper arm, lower arm (and thus elbow joint) and a hand with 4 fingers + a thumb.
-make the shoulder so it fits on a torso/sleeveless shirt. Like these:
TECHNICAL STUFF
-polygon limit is 200 triangles. You may of course use less, but aim for lots of detail.
-the arm will have to be 1 continuous mesh. Accessories such as bracelets may be separate models.
-allowed textures: diffuse, alpha. Trying to keep the focus on modeling.
-no texture size limit, but I urge you to think realistic - a 128² is reasonable, but you can also use a 512² and only use a strip of it as if it were a whole character sheet.
HOW TO ENTER
-post a silhouette and a wireframe (like this) from several angles. A spinning *.gif would be great for these. The wireframe may use vertex colors, the silhouette needs to be 1 flat color.
-post an 'ingame' version with lighting and textures if you have them. Consider this your beauty shot.
-none of these images should use AA or raytracing. All simple viewport stuff.
-if you have a texture, post the flats as well.
ABOUT THE JUDGING/PRIZES
-deadline is Sunday the 11th. This gives you the weekend to finish up.
-the focus will be on modeling: polygon flow, animatability, silhouette and amount of detail.
-but you can definitely score some bonus points with texture, style, originality and presentation.
-I will choose at least one winner, possibly more (depends on amount of entries)
-I will then P.M. the winner(s) and gift them the item via their Steam account or via Email
-If you already have Portal 2: the final hours, a game of the same price will be gifted instead.
So in short:
~a complete arm
~in 200 triangles or less
~at least post a silhouette and a wireframe.
~???
~profit!
if anything is still confusing/unclear - ask!
[excuse me if this is in the wrong place - I know there's the LPC forum but AFAIK that's only for official polycount challenges. you can move this if necessary]
Before I went on vacation I challenged people in the lowpoly thread to push their boundaries and make an arm within a limit of 200 polygons - and make it as detailed as possible. My post wasn't very clear however, and lacking in some important information. Such as... a deadline. In this thread I will clarify a bit, and I will also accept the entries. So with further ado, the challenge:
MAKE YOUR BEST LOWPOLY ARM
Nowadays we're hardly bound by polygon limits - even cellphones nowadays are becoming pretty powerful. But that doesn't mean that lowpoly is gone. First of all not all devices are fast yet. Secondly there are games like RTS which have tonnes of things on screen. And lastly - lowpoly is a very fun and educational experience.
I've learned that, often, the best way to improve is to push yourself. To go outside of your comfort zone. To try and work within very limited constraints. So that's what I challenge you to do. Take the narrow limit of 200 triangles and make a fully-featured arm!
ABOUT THE ARM
-it is humanoid, but does not have to be 100% human.
-you may use accessories like sleeves or bracelets, but remember the main goal is to make an arm.
-consists of at least a shoulder, upper arm, lower arm (and thus elbow joint) and a hand with 4 fingers + a thumb.
-make the shoulder so it fits on a torso/sleeveless shirt. Like these:
TECHNICAL STUFF
-polygon limit is 200 triangles. You may of course use less, but aim for lots of detail.
-the arm will have to be 1 continuous mesh. Accessories such as bracelets may be separate models.
-allowed textures: diffuse, alpha. Trying to keep the focus on modeling.
-no texture size limit, but I urge you to think realistic - a 128² is reasonable, but you can also use a 512² and only use a strip of it as if it were a whole character sheet.
HOW TO ENTER
-post a silhouette and a wireframe (like this) from several angles. A spinning *.gif would be great for these. The wireframe may use vertex colors, the silhouette needs to be 1 flat color.
-post an 'ingame' version with lighting and textures if you have them. Consider this your beauty shot.
-none of these images should use AA or raytracing. All simple viewport stuff.
-if you have a texture, post the flats as well.
ABOUT THE JUDGING/PRIZES
-deadline is Sunday the 11th. This gives you the weekend to finish up.
-the focus will be on modeling: polygon flow, animatability, silhouette and amount of detail.
-but you can definitely score some bonus points with texture, style, originality and presentation.
-I will choose at least one winner, possibly more (depends on amount of entries)
-I will then P.M. the winner(s) and gift them the item via their Steam account or via Email
-If you already have Portal 2: the final hours, a game of the same price will be gifted instead.
So in short:
~a complete arm
~in 200 triangles or less
~at least post a silhouette and a wireframe.
~???
~profit!
if anything is still confusing/unclear - ask!
Replies
nightshade - yes. Accessories are to be within the 200 tri limit.
I'll give this a shot!
cool I want to try my 'hand' at this one
Sorry, I had to repost the images, I forgot AA was not allowed.
gilesruscoe - I think the finger joint loops might be overkill. Too bad you lost the file (but you can make a completely new one, you've got over a week left
Cap Hotkill - looks good, but you're lacking a bit in the shoulder department. Pretty realistic result so far though (perhaps the nails need a bit different texture).
Capt, i think the finger joints are inside out? surely you should have collapsing geo on the inside of a bend?
@ Snader: I made a little rework some hours ago just before i posted that, I will post the model with the elbow and painted nails ;P.
EDIT: here is the reworked version with Elbow and new finger nail colour.
Now it sits exactly at 200 polygons.
The world's first lowpoly hand: made by Pixar in 1972.
http://vimeo.com/16292363
Here is mine, I will see about texturing it before the deadline.
The goal was to create a hand with each finger having the correct number of joints with acceptable deformation on both hand and arm within the 200 tri budget. Pretty fun challenge, forced me to think about every single tri in the mesh. Great idea Snader.
[edit] how do I upload a working animated GIF? all my attempts result in just the first frame being shown.
Installed dropbox and that sorted me
@Cap, yeah havn't had much time for 3d for a while, hopefully will be on more now.
Question, snader; why do you have set the requirements to include ALL five seperate fingers?
guess i am just playing with tattoos here is mine. I am not a low poly expert.
Here's my (virtual) arm
It's exactly 200 tris.
-because the challenge was based on this model
-lowpoly people get more detail in their models, with the same budget.
-highpoly people learn to prioritize and conserve polygons (this is a skill that some schools forget about)
-everyone gets to push themselves a bit more, and get better.
-it's easier for me to judge
-it'll give some nice things to analyze
It started out as a challenge to improve on a model in the Low Poly thread, which had 5 fingers. I could have removed that aspect in the contest. But I wanted to keep the 5 fingers to push people beyond their comfort zone a bit. Making a mitt-hand in 200▲ is a piece of cake, making one with fingers is more challenging. But entirely possible, as shown by all the entries.
It will improve people's lowpoly modeling. Those that would normally be okay with 200 tris for a mitt-hand, are now making complete fingers. This means that other parts of their arm will have to be more optimized. So in the future, they can make a mitt-hand with almost the same visual quality, in only 150 triangles.
For those that normally do lowpoly and simplified hands only, this is a chance to get acquainted with fingers and thumbs, which could lead them to try a current-gen character at some point. Besides, the industry is moving fast and in a few years even phones can render 1000▲+ characters. (yes, some can do that now, but I'm talking 'average' phones)
Those that normally do console-spec art will now have very few polygons to work with, so they really have to focus on making a nice silhouette. I believe silhouette is one of the most important aspects of a character, and as such - making good silhouettes is a crucial skill.
And to be entirely honest, it's my first time judging and it's easier to judge 3 finger-hands against eachother than 1 mitt, 1 fistblock and 1 fingerhand. Next time I do a challenge like this there'll probably a bit more freedom.
Also, since starting the thread I have become curious to see what kind of solutions people come up with to have 5 fingers, but enough polygons for some nice muscles. It seems most people take the straightforward route and make 5 4-sided fingers with 3 digits on each - something that's quite costly.
It would be interesting to see individual people come up with their own ways to get the same visual detail at a lower price. I've found one way, but I'm sure there are more. In fact, I've already seen one.
So this is also a sort of research. What problems arise from a low polycount? How do people handle these problems? Are there solutions that can be applied to multiple situations? And to gather a variety of hands and arms and analyze them (this will take a while though). I'd also like to add some of these arms to the Wiki, if the creators agree.
So basically, it started as a sillyness, but then grew into more.
Thanks for the mini comp Snader, definately been fun.
[edit]
Here it is as 150 tris but as I said, my goal was to have all the joints so this is just for fun.
You're right though, 9Skulls. This is taking too long (I'm still writing the article - I'm not a great writer but I want it to be a good read) so I'll hand out the prizes now. People shouldn't have to wait more than a month to find out what happened. I'll separate writings from winnings in the future.
Without further ado:
The self-improvement prize goes to Justin Hrala,
because he "took this as an opportunity to explore seam elimination techniques" which is a step beyond what I asked, and seam-elimination is a very useful skill in all projects. So he's making the most of his time, very efficient.
Furthermore, his model was decent (not too fond of the shoulder, but it works well enough) with nice fingers, that allow for a silhouette with 3 phalanges, while saving triangles on the inside of the hand.
The texture is okay, though rather yellow, and does it's job. It could have been better, but the main goals of this challenge were modeling and self-improvement.
The best-model prize goes to Phan Thanh Truc
I'm really liking the shapes on this mesh. The shoulder is nice, full and round, and even properly separated from the biceps and triceps. I also like the mild downward bulge of the upper arm.
The fingers, and in particular the thumb, feel a bit thin and spindly, but the topology of them is still okay. They would also have benefited from fingers like Justin's, but that wouldn't have fit inside the budget along with the bracelets.
While the texture quality isn't great, and rather pixely/blocky (a compression error perhaps?) it does show nice musculature, and has a fair amount of detail. The bracelets could use some shading on the bottom side. A clever way to hide the texture seam between the arm and hand, though.