Weekly Summary: 01-02

Switching courses has been a tough challenge. It has required me to learn new workflows and techniques very quickly. I'm at the stage now where I feel that I have almost caught up and can start producing a strong amount of quality work on a daily basis.

Due to having to catch up, the first few weeks have been based around skill-building.

Maya: 

Having to switch from 3DS Max to Maya hasn't been a challenge, it seems as soon as you know one 3d package it's very easy to use another. My main source of learning material for this was a small course bought on a website called Udemy which directly showed me how to do things that I did in 3DS max in Maya. Very helpful for a starter tutorial. 

After I got the hang of maya I set myself three tasks. The first task was to create an environment prop so that I could learn the UV process in Maya. This was my outcome:

This turned out pretty great, it allowed me to understand how to set up maps such as Normal and emissive so that they work in maya. I researched into lanterns to find that this was a 1920's Circa lantern, battery powered with an oak finish.

Next I decided to create a low poly app-game asset based on sci-fi space. I chose this because I don't really have much practise in designing that area of models so thought I'd give it a try, this was what I came up with: 

The space vessels from Dragon Ball Z were a large influence in the design of this model. It taught me not how to be efficient but how to rebuild as you go along and not stick to the design you did on the first try, I mist have changed parts of the model a few times while working through it, bear in mind I spent no more than 2 hours on this piece.

The final task was to ask someone what I should make and just wing it, luckily the first response was 'make k-9 from Doctor Who' which was pretty fun to make, here's what I made: 

Throughout the creation of the robot dog I was able to learn a few more things about editing meshes, such as tools to add more vertices and so on.

My next steps with Maya are to figure out the baking workflow and to learn some high poly techniques, this shouldn't be hard as I am quite used to it in other 3D software.

Zbrush:

Before this course I had never used ZBrush, so this meant that I needed to delve into a multitude of tutorials. I opened up the introduction course on Lynda and watched all of it, over the course of a few days I was able to learn the interface and understand a few concepts. Currently I am creating my first asset in Zbrush. 

I decided to create an animal, an incredibly tiny animal. While watching 'the Cosmos' a show about space, I was inspired to research a remarkable creature called a 'Tardigrade'. This animal is no bigger than a millimetre long, yet when zoomed in, the detail of the animal is breath taking, so I decided to sculpt it, but then Zbrush crashed and I lost it, which forced me to learn how to properly save in the program.

I decided to try something a bit different, I learned how to import and export between Zbrush and Maya to create a simple pillar. This would teach me how to sculpt broken geometry, teaching me a lot about different brushes and how they work. this is was I created:

(Textured in photoshop & dDo, rendered in marmoset toolbag)

The workflow was incredibly fast, I managed to make the whole thing in less than 2 hours and make it look pretty good, while keeping an efficient poly count (>200 faces)

Now I am going to focus on learning how to retopologise in Zbrush and Maya, and then bake the sculpt onto the low poly model.

Research: 

PBR (Physically based rendering):

I've started looking at PBR, seeing as it is the standard in the industry currently. Most current engines, even the new Unity uses PBR to the users advantage! Having fiddled with nDo and dDo the past few weeks I understand pretty much how to manipulate PBR.

Obviously everyone in the 3D industry understands the basics of Diffuse and Specular (Diffusion and reflection). The official PBR theory marmoset post explains their physical distinctions. 

Here's a little bit of basic render theory:

Let's compare light bouncing off a surface to a ball bouncing off a wall, if the wall is smooth you're going to get a smooth angled bounce, so smooth that you could pretty accurately assume where it will go, this would be a mirror-like reflection (Specular is derived from the latin for mirror). As the post says, 'mirrorness' is a bit more awkward to say than specular. Different objects will reflect and absorb different amounts of light, if an object is more diffuse, subsurface scattering will occur, to explain this imagine throwing a ball at a very damaged, jagged wall, it is a lot more uncertain where the ball will bounce and the velocity of the ball.

PBR takes this theory and expands on it, creating an engine with more reasoning behind the detail and light of surfaces, allowing us to create more specific parameters for different materials. As the article write Jeff Russell states, it's hard to actually give an explanation for PBR, because it is a loosely defined term, but I'm going to be doing more research to come to a personal conclusion.

(Theory and image from https://www.marmoset.co/toolbag/learn/pbr-theory)

References:

www.udemy.com

https://www.udemy.com/3dsmax-to-maya/learn/
http://www.onlinegalleries.com/art-and-antiques/search/hall-lanterns-wood-scandinavia-gustaf-iii - reference for lanterns.
http://bfy.tw/29Rn - images of dragon ball space stuff.

https://www.marmoset.co/toolbag/learn/pbr-theory - PBR Theory