Low Poly Modeling and Art Pipeline Optimisation
In which a funny sounding girl talks about how artists get recommended for great jobs and how artists and producers of artists can make better games in less time (without sacrificing quality). As is always the case in my lectures I’m generally offensive (but in a really helpful way).
This one is for the Artists and the Producers of Artists – I’ve been both in the last 7 years. This is about getting yourself recommended for great jobs and helping us make better games faster (without sacrificing quality).
Three key takeaways:
- Don’t waste time or processing power on shit that don’t matter.
- Focus on the game, not the pixels (ART SERVES GAMEPLAY).
- Art pipeline optimisation starts with the Artist.
#1 Don’t Waste Time on Shit That Doesn’t Matter
What matters for Art in Game?
- Composition: Silhouette. Form. The shape of the thing in space in relation to other things in space. Negative Space.
- Functionality: Move-ability. How the art is meant to be USED. Clean. Easy to use for the next guy. Well named, exported, etc.
Why? Because games mimic life and in life we don’t register detail when we’re thinking about other things in the heat of the moment (like playing the fucking game).
We have evolved to register and read light, sound, colour, form and movement. Save your detail for pretty Marketing renders and box art.
THAT is why any good art teacher worth their salt will pull you off the damn computer and get you focusing on colour theory, light and shadow, life drawing, etc. Purely because those are your game design tools. That is how you affect play response to what’s on screen at any one time.
Tech and tools will always change. Every year there will be something new. The way the BRAIN WORKS is pretty damn constant. Work to that – learn how the human works and how the brain works. You can always spend a week learning the new tech later.
#2 Focus on the Game, Not the Pixels (Art Serves Gameplay)
Art, Sound, Tech – These are just vehicles to deliver gameplay. They are elements in an overall production.
The ingredient isn’t the meal. Don’t focus on any one thing at the exclusion of how it fits in with the whole.
Design to the player’s perspective. How and when are they going to be exposed to that element on screen – how does it affect what they think or feel at that EXACT moment? How does it direct the player? Improve the experience?
Art is a design tool. Use it to design a better game.
It people are praising your game for its art – for how pretty it is – chances are there wasn’t much else worth praising about it.
No one should be walking away saying “man, that art was fucking gorgeous”. They should be talking about what they DID. “Oh man, did you see it when I…I couldn’t believe how hard it was to…hah, it was awesome when I…”.
Anytime someone has the brain space to focus on the details and the individual elements – it means design and gameplay has lost them.
They should never have the brain space available to sit there and consider the resolution of your normal maps.
#3 Art Pipeline Optimisation Starts with the Artist.
Reduce the total number of steps taken from start of work to delivered asset. Model a character, animate a rig, light a shot as many times as you can to spot and eliminate waste in your workflow. Every additional, useless step taken decreases the amount of time available to focus on great gameplay.
And every producer knows there is NO MORE TIME. You don’t get an extension. You ship because it’s time to ship.
If you can make great art in less time – everyone gets a chance to make a better game.
Saving cleanup for last is fucking lazy. Mess in your mesh, rig or animation just increases the opportunities for mess and errors later. All costing you precious time. Cleanup crap as you work. No floating faces, extra vertices, etc.
Make it easy for other people to work with. You are not the only person who has to work with this asset. That animator or that level designer has to work with it as well and if you haven’t thought about them and HOW they need to work with it than you’re costing the pipeline time and cutting into gameplay. That’s not very nice, is it?
Names things correctly. Export correctly. Build your mesh to be easily rigged and fluidly animated. Always think about the next guy in the chain and how he has to work with your work.
Everything on screen is precious and costs something. Each polygon, vertex, bone, light and shader costs the system something to process and render. Use them wisely, use them sparingly. Get as much done with any one vertex/light/shader as you can.
If a stage designer in theater can use plywood and paper mache to make us believe we’re on the battlements with Hamlet – you can do the same with polygons and repeating textures.
The Final Sell
Last time I clocked myself creating a character it took 4 hours from cube to animation ready.
Thinking the same way as a lighter – my record was 30 shots lit in 2 days.
Producers, multiply that across the number of artists you have, assets you have to create, levels you need to design.
How much time did you just save that you can now dedicate to better gameplay?
Artists, this makes us DIRECTLY responsible for making better games.
And THAT is what gets you recommended for great jobs.