ESCAPE - The Laboratory

My last thread didn't contain the "ESCAPE" prefix, so I'm starting a new one.

This will be my submission for the Escape challenge. In the world of computers, there exists a horrifying laboratory in which individuals are imprisoned and subjected to terrible, inhuman experiments. This is the 3D Modeling Laboratory - a land where, for unknown reasons, test subjects have their body parts rearranged, their faces unwrapped from their skulls, and their skin torn dissolved away into a wireframe cage, among other atrocities.

Amidst such injustice, one hero has broken free of his cell. As the result of numerous experiments, his body is now an amalgom of various types of 3D modeling projects. The alarms have sounded and the unseen agents controlling the world have been dispatched to re-capture our hero, but in spite of the desperate situation, he finds it in himself to free some of his fellow captives...

Replies

  • ChosenOne
    Catching up on some updates.

    My inspiration/concept sheet.

    conceptsheet.jpg
  • ChosenOne
    Week 2 (Belated Post): Block-Out

    For my beauty render, I want to find an angle that is able to feature my main character and a good sampling of the various 3D modeling experiments happening in the lab, but still demonstrates a sense of this virtual world's space stretching out to infinity. This is the main angle I'll be working with:

    escape01%20copy.jpg

    There are still some modifications in progress to limit some clutter, but I like the general flow of this angle. I feel everything leads back towards the main action in the center, then takes you away into the infinite height and depth beyond.

    I'm also still working on populating the rest of the scene to make a full 3D space, but here are a few more angles on the initial block-out:

    escape02%20copy.jpg

    escape03%20copy.jpg
  • ChosenOne
    Week 4/5 Update: Main Character High-Poly Mesh

    Unfortunately, I spent most of my time trying to make a reasonably clean smoothed mesh in Maya, rather than learning to effectively use Mudbox, which probably would have been more efficient. As a result, I didn't get to make any serious surface textures or small details, but I did get a fair amount in place to prepare for baking my normal maps, including some folds and seams in the clothing and a more detailed hand.

    This model was designed to be a more low-polygon character, so creating a high-res version has been a bit unusual. This character is designed to have four different parts - the organic, making up the lower-left, the mechanical and metallic, making up the upper right, the cartoon, making up the lower right, and the wireframe, making up the upper left (indistinguishable in the base model).

    I call him "Prelim".

    hires2.jpg

    ...
    hires3.jpg
    ...
    hires5.jpg
    ...
    hires1.jpg
  • GarfSnacks
    Offline / Send Message
    GarfSnacks polycounter lvl 5
    Wow man, you're going all out for this project, environment and multiple characters! I definitely like your idea, especially the face(uv's) being peeled of the head. Are you going for a realistic or stylized look?
    My only concern is that it looks like there is still alot of texture work to be done, plus populating the scene with more assets. If you have the time to do it, then great :), otherwise is does look like you might be overwhelmed with work towards the end.

    Definitely looking forward to the finished piece!
  • ChosenOne
    So, skipping a couple of steps, here's the final low-poly character model. Retopologized with diffuse, normal, specular, and emissive maps.

    headclose.jpg

    front.jpg

    back.jpg

    side2.jpg
  • ChosenOne
    Unfortunately, it seems I neglected to post anything over the past couple of weeks. I pressed straight through to the final submission. I had meant to post some assets as I created them, but I never managed to find the time. As a result, this post will be somewhat retroactive, covering the last few weeks of the challenge.

    Following the completion of the main character, I set about generating some general world assets, starting with base materials for the floors and walls. We'll take a look at those a little later in the final scene.

    As I knew fairly early on that I would be working in UDK, I tried to make my assets somewhat modular and repeatable so that I could build a few assets, copy them accordingly, and lay out the scene based on my initial blocking. Here are some of the assets appearing in the final scene:

    VariedAssets.jpg

    Around the same time, I started the process of rigging and posing the main character. Since I wanted the "robot" components of the main character to be individual objects that moved independently based on joint positions, I couldn't really combine the whole character into a single mesh. As a result, he's built from a variety of different skeletal components parented together.

    PrePose.jpg

    The above image shows the posed character standing next to the posed placeholder character from the initial block-out. Naturally, this was the next step - setting up the appropriate pose. The structure of the main character, mainly the cartoon hand and foot, made the pose a little more complicated to generate than it could have been, but all in all, it was approximated well enough. Since I was only working with a static image for this challenge, I took the liberty of separating the mesh from the skeleton to simplify the export and import.

    run.jpg

    From there, it was just a matter of brute force. I churned out the rest of my assets and went to work in Unreal. It was in the engine that I added the glow effect to some of my materials.

    UDKOverhead.jpg

    The additional resource I utilized in the engine stage was UDK's real-time texture-rendering tools. I had intended, right from the start, for each computer window in the scene to act like a window to another dimension - the scene viewed inside a window was larger than the window itself. With real-time rendering, I could take a model of the room inside the window, populate it, place a camera on it, and then render that image to the surface of the window's exterior. Admittedly, the effect would have been more useful in a non-static image, and I would have worked to increase the variance in each room's camera angle in that case (to make it look more like real space and less like a projection). Nonetheless, it does provide some level of depth to each window.

    Shot1.jpg

    I've already submitted my final entry, so what's done is done.

    BeautyShot1.jpg

    BeautyShot2.jpg

    This has been my first-ever entrance in a Polycount challenge, or indeed, any 3D art contest of any kind. It's been a fascinating experience, and hopefully, I'll be able to contribute more of my time to my next effort.

    BeautyShot3.jpg

    Thank you. That is all.
Sign In or Register to comment.