gibtango-blog

[gibtango-blog]

MDU115.3 3D Asset Production

Straight off the bat when I was presented with this assignment I knew I wanted to create some sort of futuristic pistol, as retro-futurism has always been one of my favorite visual themes and I figured it would be a fun idea to create a model from. As I was doodling some quick sketches I came across an idea for a gun that really pleased me.

The design for the main weapon used in “Robocop” was one that really interested me, as it was essentially a standard police-grade pistol with very minimalist additions attached to it to replicate something futuristic. As I wasn’t allowed to directly copy on object from another piece of media, I decided to model a standard police-grade Glock 18 with additions to give it a similar visual style to the one used in Robocop.

To create the base model for my space glock, I traced around a side image of the pistol that was textured onto a flat plane. Striking a nice balance between accuracy and keeping a low polycount was difficult, as can be seen by the amount of polygons on the lower half of the handle. I split the model into different parts so it would be easier to animate, with the slide, trigger, clip, and base all being seperated into their own layers. Once I finished modelling I created UV maps for all the objects and started texturing them.

I used Adobe Photoshop to texture my weapon and chose to create the textures myself as opposed to using pre-made textures from the internet. This was mainly done by following Pirosan’s texturing tutorials on deviantart, which catered to the specific types of metals I would be using in my weapon design. The method seen here involved creating a base texture using filters in Photoshop, then using the dodge and burn tools to simulate depth on the different faces. When texturing was finished I applied the textured UVs to the model and began animating.

Animating my weapon was the easiest and most rewarding part of this whole process, as learning the techniques to make the objects move properly and figuring out how the object works in motion was satisfying beyond comprehension. Using the model viewer for Counter Strike : Global Offensive I was able to observe the weapon animations for the game’s pistols and figure out how they recoil when fired. To animate this I relied on the use of bones to move the different parts, as I could parent them to the handle of the weapon using the “skin” utility so that they move with it, but still be moved individually when performing their own functions.

The process of rendering my model was fairly straightforward, but I did run into an issue where the bones used on my weapon would make it show up crumpled in the final result, making the weapon resemble a crushed soda can.

I resolved this issue by disabling “displacements” in the render settings, which finally bore the fruits of my labor, the finished result.

[gibtango-blog]

Pirate’s Gold Textures, Animation and Rendering.

Texturing:

Once I had finished modelling my chest, I made some UV sheets for it and got to work on texturing. This was definitely the weakest aspect of my production process, as had poorly arranged the UVs, making it difficult to paint on the textures and identify what aspect of the chest I was working with. I acquired my textures online and applied a “cutout” filter to give them a more suitable aesthetic, and then added a vibrance filter to make it match the scene.

A UV used in the chest, before and after texturing.

The finalized rendition of the chest.

Scene Assembly and Animation:

Whilst assembling the scene for my production, I found it to be much less challenging than I first assumed, as most of the skills I learned whilst modelling were applicable in this case as well.

The fully assembled scene and animation timeline.

Whilst I had no experience in 3D animation before this project, I found that most of what I learned in 2D programs such as Flash and Toonboom could apply in 3Ds Max as well. Elements used in flash such as the virtual camera had their equivalent in 3Ds, and the principles of 2D animation were just as relevant here as they are in a 2D space.

Rendering

Before rendering I needed to make a few tweaks to the scene to ensure that it looked correct. The self-luminosity values on the textures were all set to max to give the production the vibrant “fullbright” aesthetic, and many objects and elements used for animation had to be hidden from view. I rendered it as a .MOV file with an h.264 codec in 720p to ensure good quality whilst retaining a small file size.

Texture Sources:

Textures.com,. WoodPlanksOverlapping0019. Retrieved from http://www.textures.com/download/woodplanksoverlapping0019/37736

Melnyk, O. Vector - Jewellery and a box isolated on blue background. Retrieved from https://kr.123rf.com/photo_25943921_stock-photo.html

Vonc, C. MetalBare0234. Retrieved from http://www.textures.com/download/metalbare0234/57525

[gibtango-blog]

Pirate’s Gold Concept and Early Development

Over the course of this development blog, i will be detailing the thought processes that went into the production of the “Pirates’ Gold” assessment, and the various animation and artistic choices i explored during the production. The assessment involved several creative decisions that i had to consider and reassess in order to create a cohesive and well executed production.

Concept and Development

For this assessment we needed to create and animate a treasure chest with a detailed interior that can be used in an advertisement. In the reference stages, I made sure to find treasure chests that would visually match the provided set pieces. Initially I wanted to give mine a darker color scheme, substituting golden browns for deep blacks.

Ultimately I decided not to do this as it wasn’t necessarily thematically parallel to the rest of the project, leading me to pursue a brighter, more traditional design.

Sketches:

This is the design I chose to go with, as it would be simple and efficient to create in 3D, as well as providing flat, broad surfaces for UV Mapping. 

Modelling

The modelling process for my treasure chest was where I learned most of the skills that I used in my production. I extruded the perimeters of each face to create the metal portions, and I raised the chest’s inner flooring, divided the face into a tessellated pattern, and made the surface uneven to create a surface onto which I’d add a scattered coin texture.

Early image of the chest’s development.

Image Sources:

Kojadinovich, M. empty treasure chest 3d illustration. Retrieved from https://www.123rf.com/photo_12557985_empty-treasure-chest-3d-illustration.html

Century Novelty,. (2007). 12" Treasure Chest. Retrieved from https://www.amazon.com/Century-Novelty-12-Treasure-Chest/dp/B002POEQEK

abramsdesign,. (2014). Empty Treasure Chest. Retrieved from https://www.turbosquid.com/3d-models/3d-treasure-chest-model/799336

Melnyk, O. Vector - Jewellery and a box isolated on blue background. Retrieved from https://kr.123rf.com/photo_25943921_stock-photo.html

j4p4n,. (2015). Treasure Chest. Retrieved from https://openclipart.org/detail/211928/treasure-chest

Guidecraft,. Guidecraft Pirate Treasure Chest, Black. Retrieved from https://www.walmart.com/ip/Guidecraft-Personalized-Pirate-Treasure-Chest/23746039#about-item

[gibtango-blog]

Lighting, Rendering, and Compositing

Lighting

Lighting is one of the biggest aspects of putting together a well made environment in 3d. It is an essential backbone for whatever visual theme or tone you’re trying to achieve in your production, and can be what turns a decently put together shot into a visual masterpiece. Rules from real life photography such as the three point lighting technique are just as necessary in 3d, as they help by adding a layer of visual depth to your characters and environments. The advantage of lighting in a 3d space rather than real life is that you can adjust and move your light sources on a dime, allowing for unique visuals that couldn’t be recreated anywhere else. The main goal of the lighting specialist is to try and recreate the visuals of the stimulus material in a 3d space, meaning there isn’t always room for complexity. A project with a basic artstyle in mind might adhere more to the artist’s vision with simple lighting rather than something complex and realistic.

Rendering

Rendering can be one of the more time consuming portions of a production, based upon circumstances. Every frame of the production must be processed (not unlike a photograph) with the computer’s processor accounting for any complex lighting or geometry that may be included in the scene. The more detail that’s put into the characters or the environments, the longer each frame will take to render. For a still image or screenshot of your environment, this wouldn’t be much of a burden. But with big budget productions that need to be animated at 60 frames per second, the render process can sometimes take days, if not weeks at a time. The instantaneous preview of our scene that we see while editing our production uses real-time rendering, similar to what you see in video games. This utilizes the computer’s dedicated graphics card, which is made primarily for quickly rendering decent visuals at multiple frames per second.

Compositing

The final step for most film productions is compositing, which involves layering our 3d footage with with other forms of media to create the illusion that they are both part of one seamless clip. There are many methods of compositing, the most prevalent of which being node based compositing, layer based compositing, and deep compositing. Whilst deep compositing is still a fairly new technology with its set of shortcomings, it benefits overall with 3D productions by being able to simulate other forms of media in a 3d plane, effectively giving the final product levels of depth that can’t be achieved by its predecessors.

An example of 3D compositing in a Hollywood Movie

Sources:

Justin Slick’s “Guide to 3D Lighting Techniques for Digital Animation” https://www.lifewire.com/3d-lighting-techniques-2090

Atomic Fiction’s “Deadpool” Visual Effects Breakdown https://vimeo.com/159011768

Justin Slick’s “What is Rendering” information blog. https://www.lifewire.com/what-is-rendering-1954

AutoDesk Knowledge Network “Light Hotspot” tutorial https://knowledge.autodesk.com/support/3ds-max/learn-explore/caas/CloudHelp/cloudhelp/2015/ENU/3DSMax/files/GUID-8E53096F-3E86-4416-BB3A-671A341690B7-htm.html

[gibtango-blog]

Texturing and Animation

UV Mapping 

Once you’ve fully refined the shape of your mesh, you can now focus on applying textures to its faces. To do this, you need to create what’s called a UV map. The process for this involves “unfolding” the mesh so all of its faces are laid flat, which can then serve as the canvas for the texture artist.

As straightforward as this seems, there are a few precautions that need to be taken in order to make things easier for the other people in the pipeline. If the seams of the mesh aren’t well hidden, and proper scaling isn’t given to certain areas of the mesh to compensate for varying levels of detail, it could lead to the texture appearing “stretched” and unpolished. In most cases these problems are handled by the modelling software, but for precautionary reasons there’s usually a way for the creator to customize the map to suit his needs, as well as those of the texture artist.

Texturing and Shading

Once the UV map has been made and passed down the pipeline, the texture artist can work on painting the object. This is typically done by porting the map into an image editing program such as Photoshop or Gimp and drawing/imposing the textures onto the different sections. 

Once this is done, a texture map will be applied to the object to accurately portray how light would reflect off of its faces in real life. An object that shines such as metal will make use of secular maps to reflect light, and objects with complex textures (such as a brick wall) will generally have displacement maps applied to their textures, so the modelers won’t have to waste time needing to worry about modelling all of the tiny imperfections on their objects’ surface.

Rigging

Now that our model is fully textured and perfected, we can start rigging it for use in animation. Typically you’ll see skeletal rigging, which involves applying bones and joints to the character's limbs to allow it to move realistically. The rigger will also have to cover the facial joints for use in lip syncing, as well as areas of the model that rely on the physics engine for their movement.

Animation

Finally, we’ve modeled the character, applied textures and shaders to his faces, and rigged his face and limbs for complex movement, now we can get to animating him. Generally this can be done in most basic 3d modelling programs, and is done through either key-framing, tweening, or mo-cap based upon the animators’ preference. Things like lighting, camera position, and physics simulations can all be utilized to the animators’ whim. Whilst they both have vastly different interfaces, 3d and 2d animation both still rely on basic animation principles such as anticipation and follow through.

Sources & Image References:

“Introduction to UV mapping" from the Autodesk Maya Users’ Guide. https://lukewjarrett.wordpress.com/2015/06/21/3d-production-pipeline-uv-mapping-texturing-and-shaders-rigging-animation/

Brokenverts “Introduction to UV Mapping In 3D Studio Max” By Ben Tate. http://brokenverts.com/tutorial-uv-mapping/

“CS:GO - SpeedPainting skin Glock-18″ by “Bryaan.  https://youtu.be/RWidElGxxho

Cogswell College “Rigging for Animation Courses”http://www.cogswell.edu/courses/rigging-3d-models-animation-courses.php

“What is CGI Animation” article by Kirsten at Natural Record Studios http://naturalrecordsstudios.com/what-is-cgi-animation/

[gibtango-blog]

Creating a Model For a 3D Production

When modelling for a 3D production, you need the help of many other members in the pipeline to create as smooth a process as possible. There are very few cases in which the help of another isn’t required to create something as intricate as an animation or game, and the few that do operate that way tend to take much longer to make and meet much lower standards in terms of quality. The pipeline method is in place to make sure the creation process is as smooth and efficient as possible.

Pre-Production

The project’s aesthetics and style are something that need to be set in stone before 3D is brought into the picture. What art direction is the project going to use? Will there be a specific color scheme? How will these reflect upon the tone you’re trying to create? If these aren’t established, odds are the modeler isn’t gonna give you an accurate portrayal of what you want. A good way to help him see your vision is to provide him with concept art and style guides of what he should make. This will not only establish an art direction, but it will provide him with specifics regarding the objects’ proportions and anatomy.

Source: Half Life 2 Combine Overwiki

Creating the Model

Once the framework is laid, the modeler will generally start constructing their object in a program such as Blender or 3Ds Max. This is done by creating a wireframe of sorts out of basic geometry, and then working to refine smaller details, similar to how artists construct their drawings. Some assets may be created with external sculpting programs such as ZBrush or Mudbox to properly refine certain parts of the mesh. The modelers’ goal should be to create a blank canvas for the texture artist to use. 

A visualization of a typical construction process

Source: Unit 66′s 3D Modelling Blog