Home 3D Art Showcase & Critiques

Character Model, Would Like Feedback

So here is my second character model, admittedly it still has some work that needs to get done but I wanted to post some progress on it and get some outside opinions on how to improve it. The concept is a robotic cowboy who has seen better times but still maintains the personality of when he was first made. I created this in a combination of Maya and Zbrush, mainly used maya to model and create most of it and used Zbrush for the detailing and whatnot. Would really enjoy some feedback from you guys.

Polycount: 18,834
VertCount: 18,901
Rendered in Mental Ray
Utilized Diffuse, Specular, and Normal Maps

Sincerely,
Logan Shepard

fB480VH.jpg
DsFW5IZ.jpg
3tV6yAC.jpg
89YQkzX.jpg
TOoeFhy.jpg
hqSGfW6.jpg
T2QeyCm.jpg

Replies

  • Brian "Panda" Choi
    Offline / Send Message
    Brian "Panda" Choi high dynamic range
    is this for a film or a game?
  • LShepard
    The intention is for this to be for games.
  • neilberard
    Offline / Send Message
    neilberard polycounter lvl 18
    Woah, there is a lot of stuff going on there. Not really sure where to start...There is so much noise and contrast that it's really difficult to read the overall forms. I like the idea of a robot cowboy, but I think you need to get some reference and plan this out better.
  • LShepard
    How would you recommend planning it out better? What sort of changes would you like to see?
  • Brian "Panda" Choi
    Offline / Send Message
    Brian "Panda" Choi high dynamic range
    I'm going to divide this into two categories, since you took time to sit down and concept the character out.

    ==Concept==

    I'm finding the design busy, and this speaks to the final 3D model as well.

    The only thing about the silhouette of this character that speaks cowboy is the hat. But it feels out of place as the rest of the robot's body doesn't speak to a cowboy role.

    What I would recommend on the concept front is to sit down again, research 1607 – 1912 American Frontier (and Mexican, Brazilian, etc) cowboy wear and farming equipment, gather references of both cowboy gard and Industrial Age equipment, and create a crazy number (like a minimum of 50) thumbnail sketches, focusing only on the silhouette, of this character.

    In games, first reads are priority. Some games have different leeways, but there's good reasons why games like Starcraft 2, even if players are zipping across their screen incredibly fast, on-screen character are individuall distinguishable by their sikhouettes. This holds true for most game-art.

    Focus on a good silhouette that read clearly, and a huge chunk of your design is taken care of.

    ==3D==

    As stated above, the read across the materials are too uniform, my eyes aren't drawn immediately to one particular area. I'd recommend isolating the internal glow to the chest or the head. Vary where and how the metal wear and tear is. His hands, for example, are probably more worn down than his abs (when would a robot really sit that frequently)?

    My eyes have a very hard time finding a focal point on the character model. Make one.

    Also, render in a real-time renderer if this is a game asset. Mental Ray does not render in real time the way Unity or Unreal Engine 4 does. What better proof of "This works in engine" if it's popped into a game engine or a realtime renderer?
  • TophT
    Offline / Send Message
    TophT polycounter lvl 10
    I think you need to stop and think about what you are trying to convey with this piece. Concept artists and modelers are different people and have different job descriptions. It's all fine and dandy to make something from your imagination, but, there has to be some reason for the design choices that are made. Don't just make things up as you go, that will result in art that is aimless and amateurish. Spend the time to create a solid concept before even touching 3D. Make sure that things you have placed on your model are there for a reason. When thinking robot you need to think about how it is going to move, what kind of joints are required to get the robot to move the way you want. And, all that has to be done after you have created, like Panda said, a solid silhouette.
    If you aren't capable of designing something that fits the image in your head or the vision you had for you model, find another artist that can. Work with a concept artist and model what they draw for you. This is very important for a couple reasons, the first being that in the industry, that's the way it works. A character artist does not make things up and put them in the game. Character designs are created through several iterations from the concept team and then given to the modelers. The second reason to work this way is that when your portfolio is being review, if the reviewer does not like the idea of your model, the concept, they will have a negative taste in their mouths before they even look at what you did in terms of polygonal layout, UV space and textures.

    If you are absolutely hell bent on doing your own concepts, then do a ton of them. Keep drawing and sketching until the idea gets fleshed out to a point that the design is near perfect. The more time you spend designing, the easier the model is to create.

    In terms of what you created modeling wise, keep at it. You seem to have a good understanding of the basics, it's just too hard to critic the model itself because I can't get past the design. Don't get discouraged, just keep at it.
Sign In or Register to comment.