Art Test

Deoce

Hello all,

Like the title says its my art test for the job, the first one actually. Never done this kind of model before and it took quite some time and now its finished. Learned a lot trough the process of making it and now we hope for the best to get the job. :poly122:

Would like to hear your critiques/comments on this and the things i can improve on my future work.

8006 tris
2k textures (Diffuse, Specular, Normal)
Rendered inside marmoset toolbag

Sorry for my bad english.

Limewax

Interesting design....

but in your first image: having the wire frame behind the turn around images isn't clever, its confusing.

Deoce

Limewax wrote: »

Interesting design....

but in your first image: having the wire frame behind the turn around images isn't clever, its confusing.

Nothing that can not be fixed

[Deleted User]

The user and all related content has been deleted.

DWalker

From a technical perspective, there are some problems. The normals at the tips of the rotors are wrong - notice the shadow in the rotor tip in the second picture. Swapping triangles in that final quad would fix that particular problem, but the geometry or shading groups might be the underlying issue. The polygon count seems a bit high for what you have - some of the fuselage details can be moved into the normal map without losing quality. The normals along the wing also seem to have some issues, but it's harder to see with the images. The red light seems too bright for the scene's lighting.

When modelling airfoils - whether wings, propellers or rotor blades - you need to understand their form. They are generally thickest at the quarter chord point (that is, 25% along the length of the airfoil), are almost elliptical ahead of that point, and meet at a sharp point at the trailing edge. You can successfully model one starting with a 4-sided cylinder; place one point at the leading edge, one at the trailing edge, and the remaining at the quarter-chord point. For a six-sided cylinder, place the next two at ~12.5% chord, roughly half the thickness of the chord. This should provide enough shape definition for most smaller wings. The top of the airfoil should be 1 smoothing group, the bottom in another, and the two sides at the leading edge in both groups. This keeps the trailing edge sharp and the nose smooth.

From a design viewpoint, you need to understand the proportions of vehicles just as you need to understand human proportions when modelling characters. Even for sci-fi designs, you need to understand how things are placed in similar existing vehicles.

Helicopters get almost all of their lift and most of their control from the rotor disc. The engine and rotor shaft are placed to go through the center-of-gravity along the vertical axis. The blades should have the same chord length along most of the span, possibly changing shape near the tip; your design would lose almost half the lift. These designs are as exotic as modern blades get:

Mast mounted items are radially symmetric, whether mast mounted sights as on the Kiowa or a radar as on the Apache.

Forward swept wings are extremely rare, and your design would shift the center-of-gravity forward, and shift it radically every time a weapon is fired. The heaviest weapons are always placed closest to the fuselage, both for stability and structural reasons; place the missiles on the innermost pylons.

Deoce

Looks promising... what engine/platform?

I think your presentation could be improved - first thing I would change is the brown background and improve the lighting (cast shadow on the ground might help)

Thanks. engine is marmoset toolbag i listed it.

other then my 3d work i know i have to improve my presentation... i dont like marmoset floor i avoid it all the time. i might start making little scenes for the model presentation tho :poly124:

From a technical perspective, there are some problems. The normals at the tips of the rotors are wrong - notice the shadow in the rotor tip in the second picture. Swapping triangles in that final quad would fix that particular problem, but the geometry or shading groups might be the underlying issue. The polygon count seems a bit high for what you have - some of the fuselage details can be moved into the normal map without losing quality. The normals along the wing also seem to have some issues, but it's harder to see with the images. The red light seems too bright for the scene's lighting.

When modelling airfoils - whether wings, propellers or rotor blades - you need to understand their form. They are generally thickest at the quarter chord point (that is, 25% along the length of the airfoil), are almost elliptical ahead of that point, and meet at a sharp point at the trailing edge. You can successfully model one starting with a 4-sided cylinder; place one point at the leading edge, one at the trailing edge, and the remaining at the quarter-chord point. For a six-sided cylinder, place the next two at ~12.5% chord, roughly half the thickness of the chord. This should provide enough shape definition for most smaller wings. The top of the airfoil should be 1 smoothing group, the bottom in another, and the two sides at the leading edge in both groups. This keeps the trailing edge sharp and the nose smooth.

From a design viewpoint, you need to understand the proportions of vehicles just as you need to understand human proportions when modelling characters. Even for sci-fi designs, you need to understand how things are placed in similar existing vehicles.

Helicopters get almost all of their lift and most of their control from the rotor disc. The engine and rotor shaft are placed to go through the center-of-gravity along the vertical axis. The blades should have the same chord length along most of the span, possibly changing shape near the tip; your design would lose almost half the lift. These designs are as exotic as modern blades get:

Mast mounted items are radially symmetric, whether mast mounted sights as on the Kiowa or a radar as on the Apache.

Forward swept wings are extremely rare, and your design would shift the center-of-gravity forward, and shift it radically every time a weapon is fired. The heaviest weapons are always placed closest to the fuselage, both for stability and structural reasons; place the missiles on the innermost pylons.

basically what are you saying is that i have to know how to fly this thing before i start modelling it im just kidding.

anyway they sended me the concept art i didnt came up with this idea. the 8k tris count is what they requested i know it could go lower. some of the errors is just lack of the experience i have and like i said first time modeling this kind of stuff (vehicles) other then that i see i need to research more before starting a project and plan the model ahed so i dont end up with some of those normal errors. its something it has to be learn the hard way thanks a lot for the in depth comment.

DWalker

Silly non-flying concept artists. :P It helps if you can post the concept, but sometimes they like to keep those secret...

You should try to broaden your experience a bit - pick a real vehicle with plenty of good reference photos and model it with the same 8k limit. Then import it into an actual game engine so you can get some experience with the differences between the appearance in-game and in Max/Maya. The exact engine isn't critical; the process is similar enough in most modern engines and most companies will expect to train you a bit anyway. UDK is a good choice because it can be freely downloaded and is still fairly robust.

Deoce

Concept art can be found on the internet.



yea i was planing to do something like that... take some scifi stuff or more vehicles or similar and try to keep away from all the fantasy stuff cause i did a lot of those

I use UDK for more complex stuff where i have to use more objects and more textures.

most companies will expect to train you a bit anyway.

that sentence really helps a lot