Home 3D Art Showcase & Critiques

mars rover

Hi-

I'm starting to work on a Mars land rover robot. You may have seen the mausoleum I'm working on. That's not done at all, and I am in the process of watching and learning tutorials for working on it in Zbrush, which is a new experience for me. To keep going with actually creating work in tandem, I thought at the very least I could carefully start the hi-poly model of a new piece.

As this is a new piece, my first concern is whether or not the subject is appropriate. This isn't intended to be rigged or anything, it's just a static model.

Here are the wheels, which are all identical:

mars-rover.jpg
mars-rover-wire.jpg

As always, let me know what you think when you have a moment. Thanks!

Replies

  • ZacD
    Offline / Send Message
    ZacD ngon master
    You should block out the whole rover before going into detail.
  • MRico
    Offline / Send Message
    MRico polycounter lvl 10
    ZacD wrote: »
    You should block out the whole rover before going into detail.

    This.

    Also...there's this trend starting...where people keep making "high poly" models....but they literally model every poly...

    For example those holes on the wheel hub-cap thing...the wireframe looks insane...not necessarily in a good way. A good 8 sided cylinder hole turbosmoothed would give you that same effect with a more efficient/pro looking wireframe.

    I saw your mausoleum on game-artist.net, where you did the same thing. Those pillars were just crazy. Don't be scared of turbosmooth...or whatever it is that smooths your models in whatever software you're using.

    Edit: I made this image to help you understand what I mean.


    This is cool because you can have the high-poly off while you wok on other peices, not slowing down your machine much.

    The way you have it now, you're stuck with that dense model.

    Adding a turbosmooth or maya's equivelant will Sub Divide your model, meaning it will split each face into four, smoothing out your model for you....here's an image of what I mean.

    Help.jpg

    The part I actually modeled is on the left. When Turbosmooth is on, it SubD's it to the one the right. You keep a better looking and not as dense wireframe.

    I'm sorry if I'm not explaining it well, I just woke up.
  • kylix3d
    Hi-

    Thanks as always for your feedback. As is evident, up until this point I have had basically no experience with subdivision modeling. I think I understand the issue, but I'm not exactly sure. I know I'm slightly closer at least.

    Here is a quick, simple experiment I did:

    subd_gear.jpg

    I have Maya, and I tried using the Smooth Proxy tool. The cage on the left is what I was actually working on, while corresponding changes were made automatically to the subdivided model on the right. It seems that to get harder edges, extra edge loops have to be added to the cage. I had heard of subdivisional modeling in the past, but I wrongfully assumed that it was only meant for organic objects etc.

    Is this what people are referring to?
  • MRico
    Offline / Send Message
    MRico polycounter lvl 10
    kylix3d wrote: »
    It seems that to get harder edges, extra edge loops have to be added to the cage.

    That's the idea, although it's not always that easy.

    Practice on that model, try to figure out where the edges should go.
  • Clockwork
    Yeah, you do have to have supporting edges to make sure they stay hard. Or if your software has it, you can use edge weight, I'm using edge weighting religiously when doing subd models.

    Adding extra loops is pretty much the norm when subd modeling, especially hard surface.

    Did you do any booleans for your holes, in the first example?
  • OrganizedChaos
    Offline / Send Message
    OrganizedChaos polycounter lvl 17
    I'd like to reccomend that you go through this thread if you haven't already

    http://boards.polycount.net/showthread.php?t=56014

    I read through it one day and found it to be an excellent resource for sub-d modeling
  • kylix3d
    I created the holes myself without using booleans. I'm really used to modeling with polygons, but I can now definitely see how subdivisions smooth the model while giving control simultaneously (as opposed to just smoothing it). I was asked about my process earlier. After gathering reference images, I prefer to start on the high poly model and then degenerate it to create the low poly model. Incorporating the subdivisional method, it seems that I would delete geometry from the subdivided model instead of the high poly cage. Otherwise, I would be basically working with three models. Is this right? Thanks!
  • dv8ix
    WHERE IS THE ROVER?
Sign In or Register to comment.