Dissertation Idea

DKeymer3d

Hey PolyCount,

I'm rapidly approaching my final year at University studying computer games modeling and animation and I'm struggling to nail down a subject for my dissertation.
I've decided to go down the path of an environment artist and want my dissertation to be something that helps me in that area.

Here's are some ideas from my brainstorm so far:

1. Analyzing and comparing different types of lighting like say UDKs and Cryengines. This has both a practical element and a argument going as well as plenty to research on the aspects that make up each engines lighting but I need some help narrowing it down to a question.

2. Looking at how MEL script can benefit an environment artist workflow. (though I have found this to be a very limiting subject). Can someone who has experience with MEL give me an idea on how far you can take MEL script and how they have used it?

3. Looking into multi-pass rendering in maya to render a scene. I like this idea cose it has a practical element to it however multipass rendering scene in maya is more used in film than in games. Is there any way to turn this around? Does anybody know of any games that use multipass rendering for games?

4. To help me with environment techniques I would like to look more into particle effects. I know of UDKs kismet particle system and want to explore that further but I don't have anything to compare it to nor can I think of a question to go with the subject. Again this is good for me cose it has its practical element.

5. Looking at Unreal's lighting now and the future of Unreal's lighting. This has me interested as UE4 is just around the corner and would be helpful to know however there's very little information I can find other than what people say about it, not really anything I can analyse on a practical level

Which idea would be best to follow? Suggest some new innovations or techniques that are on the horizon or just coming out that haven't yet been looked at enough or at all. If anybody can help me based off of these ideas or suggest something new for an environment artist I'd be very grateful.

Kwramm

forget MEL - look at Python. Python is extremely powerful in Maya and you can do a lot with it, depending on your pipeline / task. If you want to go into that direction you can investigate script based scene assembly systems, how scripts can help with modular environment workflows or with baking maps and exporting.

leleuxart

I would skip number 1. The UDK we use now will probably be gone soon or updated to UE4.

I can't speak on number 2, not involved with coding languages at all :poly121:

Number 3 is a little vague to me, because GPUs do multiple passes. Each engine can also adjust where and how things are rendered. Crytek has done a lot of presentations on their rendering tricks for certain effects, like AA and SSAO/SSDO(You can find those on Crytek's site). I'm not sure if that's what you're after though.

I'm not sure of a topic for 4 either. Perhaps the complexity of particles now compared to older gens? The pros/cons of particles systems, like the amount of overdraw?

Number 5 would be good, because I think right now UDK isn't fully Deferred, but UE4 is? You could also speak about UDK's baked lighting and Lightmass volumes for GI, compared to UE4's Enlighten lightmap solution and the older Sparse Voxel Octree Global Illumination Epic originally implemented. I could be wrong about the Deferred stuff, since you can have a mix of Deferred and not, but there's plenty of info on both engines now I think, including the Deferred/Forward renderers.

I hope this helps a little. :thumbup:

Kwramm

number 2 might not necessarily require coding knowledge. Maybe an analysis of the possibilities, off the shelf solutions and custom solutions employed by studios would be enough. This could present the tools used, describe pros and cons, give examples of real world pipelines and do an analysis of possible savings. Finish up by talking about where we're heading in the future and maybe mention some experimental - not yet mainstream - approaches of streamlining / automating workflows in env. production. Could also talk about engines like UDK / Cry / Unity and how to optimize world building + level design, bringing engine and DCC app (e.g. Maya) closer together and integrate it more tightly.

jfeez

If you want to look at lighting/shading research into physically based materials, next gen seems to be all about them atm. Python is great, making a pretty nice tool and then writing about how it would help environment artists workflow would be awesome. Personally i would look into the materials and into the new lighting techniques because while they are separate things they go nicely together and if you want to be an environment artist learning the new texturing techniques that will be needed for physically based materials will be really helpful. If you wanna be a tech guy, do python because you will use it everyday.

Wesley

DKeymer3d wrote: »

Hey PolyCount,

I'm rapidly approaching my final year at University studying computer games modeling and animation and I'm struggling to nail down a subject for my dissertation.
I've decided to go down the path of an environment artist and want my dissertation to be something that helps me in that area.

Here's are some ideas from my brainstorm so far:

1. Analyzing and comparing different types of lighting like say UDKs and Cryengines. This has both a practical element and a argument going as well as plenty to research on the aspects that make up each engines lighting but I need some help narrowing it down to a question.

I'd go with this.

It's a great excuse to make a KICK ASS environment with high focus on lighting.

It doesn't need to be a question, it just needs to be a topic really. So. "ENGINE LIGHTING - AN INVESTIGATION INTO LIGHTING TECHNIQUES IN VIDEO GAME ENGINES" (all caps is super important).

So you have a third of your paper (but not really a third of your time) dedicated to dissecting lighting principles, techniques, evolution in video games and the current standards/practices. Then! You get to explore this further by creating an environment with a high focus on lighting and then implemented in both Cryengine and Unreal. You then have a bit that covers the making etc. THEN, you have a wrap up with conclusions, relating to the earlier research. Then you have "FUTURE" section where you talk about near-future advancements.

AT THE END OF THE DAY: you get a nice environment for your folio (which is what really matters... - only pick the best engine version for your folio obviously) and you have a perfectly good dissertation piece that lends itself well to research. LOOOOADS of academic stuff covers lighting. You could make it super technical, principles of lighting and uses in video games (such as directing player movement through lighting) and how that feeds into engine development.

God dissertations are super fun.

Jason Young

Agree with Wesley. Whatever you do, use it as an opportunity to create another portfolio piece.

Also agreeing with some of the others. With the next gen looming, I'd pick something relevant to that and the physically based stuff is a hot topic right now. Personally, considering cryengine is getting an update very soon with pbs, I'd probably choose that discuss how the change will affect environment art. Dealing with multiple engines means less time creating art for the project.

Depending on what you're interested in, you could also look into doing something more stylized. A lot of the physically based lighting/shading is shown off with realistic examples, but it'll definitely benefit exaggerated environments as well. Think about how the changes this new generation brings will help us get closer to feature film animation visuals. Just a thought.