Environment art test for NLG

adam

Hello hello
Completed another art test for a company out in Vancouver, BC. for a company called NLGames. The test is literally "model this environment based off the concept provided" with not other limitations. I browsed the companies site and noticed their working with the HL2 engine so I tried to make it look some what geared towards those limitations.





Wireshot.
Concept.

The env. isn't suppose to be textured.

Posted this in flaagan's thread but figured I'd make its own for some more concentrated C&C.

artgoon

Better watch out!

" A response will be sent only if there is a suitable position available. All materials sent to Next Level Games become property of Next Level Games Inc.
We appreciate your interest."

Might want to hold onto the scene file.

Looks dandy to me.

adam

All they got were .jpg's so I'm not terribly worried.

Matabus

Fuckin sweet, Adam. I love the mood the lights provide.

spacemonkey

Looks good Adam, I like

pogonip

I'd hire you ... looks good ?? How do you know they did not want High Res art ?? Did you ask ?? Maybe they are modifying the HL2 engine to use Normal Maps ..etc etc ?? Might wanna ask ehh ??

Jeff Parrott

Looking real good Brome. I think you could make some real cheap bolts and put them on the walls. That might make the appearance of more detail. Good luck with the test.

adam

Well I was more or less making it so little things (like bolts) would show through with a good skin.

I'm doing a texture art test for them where I have to create textures for this scene, but I have to use another concept for the textures and not my env. I've created here. Easy, but very weird...

adam

Here's one I've done:

Jeff Parrott

I would model the bolts and just place them on the texture. Just seems like they pop out a bit on the concept. But it looks good either way.

adam

Ah, and thanks for the nice comments

Mangled Poly

Adam,
I must say you have improved alot! that in its self deserves a pat on the back, But I have a few comments.
Firstly I would believe you may not get a job for posting this, not for showing off your work, but for post THERE concept art, which is totally wrong.

Second although the scene is nice you are at a wierd position. Its to many polys to be realtime, but not enough polys to trully show off what normal maps can do. I recomend pushing it farther, more bolt models, more panel grooves the works. Then go back in with a bump overlay map and add details like scratches, dents and other details, that are to frustrating to model, but add the final polish of detail then you will be somewhere!

Lastly i would strongly ripping it all to low poly normal maps, This step may not be needed, but not many people these days know how, or know how to do it well. This will sell you as a good technical artist, aswell as a good artist.

Best of luck!

MoP

Yeah, I was also wondering about the polycount... as Mangled Poly says, it's too poly-heavy to be a real-time environment (unless it's the corner of a room in a console beat-em-up), but nowhere near enough detail or smoothing modelled in to take advantage of normal maps. What was the target with it? You say they suggested "no limitations" ... you may have made a mistake by not making a decision early on to either go all-out highpoly, or all-out lowpoly with textures.
However, it does look pretty good, and accurately represents the concept.
Good luck.

Matabus

[ QUOTE ]
Firstly I would believe you may not get a job for posting this, not for showing off your work, but for post THERE concept art, which is totally wrong.

[/ QUOTE ]

This particular company has their art test concept work available for public viewing. You can just go and take a look at all of their tests if you'd like. He won't be in any trouble for this.
http://www.nlgames.com/Art_Test/index.html

adam

Mangled: I'll answer/comment on your questions/comments in order.

Their art tests are public, so I should be fine in that regard.

I know I'm in that funky middle area that isn't comfortable to be with a model. The low poly I did was just too low poly for me to feel OK with submitting, and figuring that all they wanted were .jpg's I felt OK with submitting these images. I have a local friend here who works with DICE and he went over this with me, he also suggested I do the same thing of going even higher poly and ripping it to normals. The test doesn't mention anything on the technology being used so I didn't bother. I will be keeping this piece as my test for when I go to do normals for the first time Ah, and thanks man - I feel like I'm improving with every piece I do.

MoP: Thanks for the comments Again I agree with the feeling of being in a funny place with the polycount. Like I told MP I'll go back and use this as my test subject for learning normals.

Thanks guys

poopinmymouth

I'd say the biggest critique for me, is that you didn't follow the concept close enough. If you look at the concept almost all of their details, rivets, grooves, and nurny bits are thick and chunky. Probably because it will be for normal maps, and those small grooves you have are not gonna pick up much. I noticed in the concept the upper paneling had nice rounded corners and that would make a really nice detail. It also feels like you skipped out on the medium scale detail. There is the big stuff, and there is the tiny stuff, but there isn't much medium scale detail.

Looking pretty cool overall, I like the lighting, but would have prefered one of them to be a bit more saturated for some stronger color play over the forms.

adam

Good points Ben, thanks

adam

I forgot to mention that TelekineticFrog lended me a HUGE hand in lighting this scene. I'm a lighting n00b (hell, I'm a 3D n00b!) and he offered his wisdom - it's obviously paid off

Thanks TF!

notman

I think that's one thing I've never seen is a lighting tutorial that offers up good suggestions. I'm with you adam, my lighting abilities suck.

cholden

You had help with that lighting? Honestly, it does nothing for me, almost to the point I'd say this scene would look better without that lighting.

Adam, one place you tend to fail at art tests is you never take it above and beyond the requirements.
I'd personally call this art test a failure, because as Ben said, you didn't follow the concept close enough. IMO, you need to finish this scene with full detail, texturing, lighting, and possibly embelish the scene to being a larger area with more detail that fits the concept. By taking your art test above and beyond (without breaking the limitation), it shows your creativity and ability to run with a concept. The key word here is TEST, they are testing you to make sure they won't have to hold your hand and baby you all through the project. Based on your work here, that's something they are going to have to do, and lowers your chance of getting work. So why not show you're an ARTist, and push this TEST to the next level?

sinistergfx

If you consider yourself a noob on any part of an art test and really need help to complete it, then it's not really time to be submitting art tests.

Hone your craft. A company is really going to want someone who knows what they're doing.

That being said, not too bad on the modelling. Details could probably be exaggerated a bit to get the most mileage out of your polies.

Texturing wise: I think you could have gotten much more mileage out of your texturing by using tiling textures and then overlay planes for any unique detail you might need (the "numbers" on the panels and any stains or other area specific detail). Savour every pixel.

Lighting wise: decent, but it would be nice if the "floaty ball area" emitted a different color of light than those lights on the panels outside. Use varying color to differentiate different types of light sources and create depth in your scene.

I could go a bit more in depth, but I'm a little tight on time right now.

Keep practicing.

adam

I'm not at home so this has to be short. cholden and sini - thank you. really guys, great points.

Home now: chris, sini and ben.. you're all 100% right and I'm going to go through this again once I finish what I currently have on my plate. It might not be for a month or so, but I'm definitely coming back to it. Thanks guys

/start goodkickintheass.exe