Lighting Artist
polycounter lvl 11
How common are the dedicated postitions in the games industry? I know smaller studios probably won't have a person dedicated to lighting, and even the more bigger studios seem have a smaller section dedicated to it, if any.
Whats the best way to show off a portfolio regarding a lighting artist?
Do I need to light full blown environments? Or can smaller scenes be better?
Breakdowns on each light/shadow and post-processing?
I know Naughty Dog does does have a lighting test, any others that anyone knows of that are open to give me an idea of what to expect?
There's pretty little info out there regarding this role, apart from the obvious expectations on having a good understanding of colour theory, lighting/shadow and hardware limitations.
Whats the best way to show off a portfolio regarding a lighting artist?
Do I need to light full blown environments? Or can smaller scenes be better?
Breakdowns on each light/shadow and post-processing?
I know Naughty Dog does does have a lighting test, any others that anyone knows of that are open to give me an idea of what to expect?
There's pretty little info out there regarding this role, apart from the obvious expectations on having a good understanding of colour theory, lighting/shadow and hardware limitations.
Replies
I'm not sure why you'd show break downs of each light or shadow.
Most engines and games will have their own tech limitations of what can be achieved and how expensive lighting is - so having experience working on environments in games that have shipped will help you a lot since you have a more astute understanding on how lights affect performance, and when it comes down to polish you'll know how to get the most out of the least.
You should also probably show that you understand color theory, composition, the ability to create a decent mood and possibly technical understanding - I know for one game I had to script shadow casting lights on/off in order to work around hard set limitations on how many lights could cast shadows, getting the most out of your shadowmap resolution, how player models look when being hit by dynamic lights and tweak the dyanmic lights to match static/baked ones and so on. You'll be lighting environments and characters, so understanding stage lighting would be useful I guess (there's more to it than just a basic 3 point lighting setup).
I know some lighting systems I've worked in the past use math algorithms for flickering and timing - so knowing a little bit of that to change speeds and behaviour would help.
Also showing organization will help for larger scenes (lots of lights, especially if there are seperate dynamic and static lights), and understanding pixel density will help (lightmap resolutions, keeping things looking consistent and understanding when to use vertex lighting etc).
Show you understand post-processing. Look up tables, light probes, AO, depth of field vs fog (height, distance) and so on.
Lighting for games, especially as a dedicated lighter, is more than simply placing lights around in space. Lighting also affects performance, so having gone through the motions of polish on a game and knowing what improves framerate will be useful - I know I've done profiling work in the past, and most in-house tech and even Unreal comes with tools for handling that.
Its becoming more common with this generation of game development. It'll probably be even more so in the next generation.
It's often not just a 'place and tweak lights' job though, as often you'll be dealing with or auditing lightmap UVs and props, adjusting materials and post/proc settings, regularly chatting with graphics engineers about budgets and workflow, likely working with tools that are proprietary and/or undocumented, and will likely have to present your work not just to a direct lead but also to many others (art director, producers, even story/narrative guys, etc).
A good first exercise is to light the same exact scene with different moods to show off what you are capable of.
You could start with an already completed scene from UDK example maps or something...but it doesn't have to be a complete level--just enough to show off what you're capable of.
Additionally, making a day night cycle animation of that set is a good portfolio piece.
Some art tests will have you lighting untextured stuff so the focus is on the lights, which is fine since the testing studio has a good idea of what to compare your results to, and its fine to do your initial lighting work in an untextured view mode, but I'd limit the amount of untextured final examples in your initial submission portfolio, as its the 'final scene' that really needs to be convincing, and oftentimes lighting and materials are REALLY closely intertwined.
Doing a break down of one or two shots isn't a bad idea, but only for a couple of pieces...no more than is needed to get a sense of your attention to detail and your process. It's better to spend the time filling out your portfolio with better and better examples.
Anyway, Goodluck!
http://portfolio.tomaslidstrom.com/
(but im biased since i work with him and he is awesome)
But yeah, lighting and FX often work pretty closely aswell, knowing a bit about shaders and postprocessing aswell as rendering pipeline etc is a good plus, not just "place lights make look good"
Any more info on what would be in a good portfolio, would be helpful.
1 or 2 of them might just be contract lighting artist, not sure. And we might have more than 5 or 6 too!
http://www.naughtydog.com/site/careers/lighting_artist1/
This is from our Career Page on what we look for in an lighting artist.
Yep. Good lighting can save what might be a very stale or empty environment. Bad lighting can ruin a great one.
Ark - full environments that are lit and have good use of post would be a good start IMO. Perhaps have an unlit version for "before" and "after".
Game Development dude, it's awesome.
Seems I'm gonna have to start building some environments though before I can eventually get to actually lighting some scenes.
I literally had producers going from "we can't ship this level like this - do something" to "we got word from HQ they are in love with the level - it's awesome!", when all I did was new lighting. So yeah, hands down #1 important thing for me.
And speaking about a portfolio It might help having a few concept-art thumbs just showing your different mood ideas. I mean sure having a scene or two lit in engine is great, but if I can see that you're a cornucopia of looks and moods some minor technical blanks will probably seem unimportant and fixable.
P.s. Tor, you're totally unbiased. Gears 3 had some of the best lighting in a game ever. Really on par with Uncharted 3 in my book. Coming from a series that started out brown, heh?) So yeah, tell Tomas he's awesome from me please
Also... there needs to be a lot of technical understanding, some code maybe to get to do post processing and lighting properly. Specially night and day. So maybe a more tech artist approach.
Usually you would make them yourself. Knowing what makes a well made asset is part of being a lighting artist. Especially when it comes to textures, shaders, and lightmaps.
Yeah love his work.
Would be cool to hear from any lighting artists out there what a typical day is made up of?
When I was doing lighting, my day went something like this:
Sync to perforce, review the log and look at any changes that were made to the layout or environment since I last looked at it - both in Maya and on console. Since the tech used shadow casting geometry that was used for dynamic lights, if the level was modified I'd have to modify it to reflect any changes made to the actual environment.
Reviewed any lighting that I had noted down as incomplete or unfinished. I'd also check to see what the colors of the lights I was using and eyeball it to make sure they were all the right hues, and no intensities were out of control. I'd also check and make sure any dynamic lights that were animated were different enough if they were close to each other (this is where organization and a good naming convention will help you identify where lights are based on the name, and what they're being used for). Once that was done, I'd look at the shadowmap resolution and make sure any lights turn on in a given area was not going over the budget - if it was an out door map I'd look at the directional light and make sure it was sliced / LOD'd correctly and as best as it could so the transitions were not noticeable and as smooth as possible.
I'd also turn off all dynamic lights and make sure any light in the bake was not washing out any of the colors or killing the normals on the textures. Then I'd flip it out and turn off the lightmap and just look at the dynamic lights to see if they were looking OK. Then both on with no post and so on.
Check my post-volumes in-game and in Maya to make sure that they were transitioning properly and naturally. Check shadow casting lights to make sure they were turning on and off correctly to avoid and pop-ins or crashes in the level. I'd look and see if anything was named in correctly or was not needed and any kind of cleanup to the scene.
Take screenshots and adjust in photoshop to see if tweaking color balance or luminance would help make a scene pop out more. And so on.
Then it could have been any of the following depending on what stage of the process I was :
I'd take a look at it unlit ingame and look at the color key provided (or review the one I made, which would be a paintover).
Then I'd set basic lights and do preview renders or bakes to see how the interior would look. This can take a while, days depending on scale or even longer.
Once that was done I'd pop in dynamic lights and AI / player meshes and make sure the lighting on the characters matches the bake, and that it was reacting to spec, helping the normals pop out and reacting well with the player weapon (FPS game). Usually I'd have to take the baked lights and make them darker to help play up the dynamic lights, but it really depends on the tech since they'll typically be different per studio I guess. But the point is I'd be going back and forth and doing lots of test renders and builds to make sure things looked right.
Then I'd set up volumes / triggers / scripts to turn off lights that were expensive and not in view, or transition between two post-process settings or changing shadowmap resolution to make things crisper (why have a directional sun using that shadowmap if you're in-doors or underground?)
If it was nearing the end of the game production, a lot of my day would be spent on clean-up, level optimization, profiling, staring at GPU/CPU load graphs and seeing what I could do: get rid of geo, clean-up lights, reduce geo , swap out more expensive lights with cheaper pointlights, condense materials/textures, turn off shadows, be more aggressive with scripts and turning off shadow casting lights and so on.