What is good low-poly and what good are low-poly skills anymore anyhow? Many years ago, when then current generation games had progressed beyond low-poly models, I wondered if I should ever learn to do low-poly. What use does it have? But I did learn and am glad I did.
Good low-poly 3D does a lot with little
The limitations can be very strict, especially in the mobile and similiar platforms, so we play within the limits. Clever placement of polygons gets the shape down with minimal polycount while still maintaining animation capacity. Virtuoso-like uv-editing maximizes the use of texture. Great texturing makes one texture serve the same role as many. That’s good low-poly.
The low-poly I’m talking about in this article is really low-poly – work done with very few polygons. Don’t confuse this with ‘low-poly’ models often featured on forums such as Zbrushcentral. There, when artists model for games, they usually model for kick-ass consoles and strong PCs, and call the in-game model ‘low-poly’ even though it weighs somewhere above 7000 triangles. In that use ‘low-poly’ is just an expression, just a way to say ‘in-game model’.
Low-poly is anything but dead as a field of work
The demand keeps on growing. Mobile and similiar platform have recently acquired computing power to run low-poly scenes up to 1000 triangle-polygons and above, which could be full a scene with ground, a character and some props. It could actually be a small scene from a game production I’m in. Our game runs on game engine released in 2002 for PC. Now similiar stuff rocks on mobiles. Imagine that.
Other than on mobiles low-poly is used extensively in handheld devices and online-gaming – anywhere really where resources are limited and/or game features tons of characters at once. World of Warcraft too is low-poly.
We’ve established that low-poly takes skill to do well and that it is still viable as a field of work, and growing. Ok, so are the skills learned at low-poly useful anywhere else? Hell yes!
What does low-polygon work teach?
- Good 3D model is optimized for light-weight and for animation – polygons are placed only where needed to make shape or help animation deformation. Good model also has an easily read profile and personal look, good design.
- Good UV-mapping maximises the texture usage, re-uses parts of texture creatively so that you don’t notice it is the same bit repeating and weights important areas over others. Low-poly restrictions force people to explore these things.
- Low-poly texturing means painting everything to a colour-map: shadows, highlights – the works. It’s not unlike painting a picture. You need to learn about painting, how surfaces react to light, where shadows go and more.
- Low-poly animating puts attention to intention and the big picture. Small motions get lost with blocky models, and there will be no close-up of a facial pose, so it is better make clear and profilic motions and learn to get the message across using the whole body. Very good to know.
I think my point is made. If you don’t know how to do low-poly yet, don’t hesitate to learn. Have you avoided low-poly work? Or have you found it useful? If you do low-poly, what platform is it for? Please don’t hesitate to share your take and experiences.