Wednesday, October 24, 2012

Thesis: The Clothing

With the hair a decent place a week ago, this last work week I turned to the clothing of the character. I was having second thoughts about my original designs, but came to like them once again. A change I did end up making was the general size of the clothing. I want them to appear large and baggy on the character. I want him to roll up his sleeve only to have it slip down again over his hands. Where the model currently stands, I need to simply scale up the shirt a little more.


Yes, he is missing his shoes. I am slightly stuck on what style of shoes he should be wearing. Sneakers, boat shoes, white, black, brown, I am just uncertain of what will tie into his appearance (most likely just sneakers given his younger age.) As for clothing dynamics I am also at a bit of an uphill battle. I have been searching to no avail for anyone who has any hints for how to properly do clothing deforms on a character without adding hundreds of hours of cloth dynamic cache calculation. I am still looking into it and will get back to you when I get over that hurdle. I will probably be able to figure out the clothing dynamics thanks to the hair dynamics, which are a lot of fun to play with. 

The other task I have decided to tackle this week is Zbrush. With the clothing modeled, I wanted to start sculpting the details in the fabric. I have only used ZBrush once before and that was years ago. Traditionally I use Mudbox for all of my sculpting needs, but the bottom line is that Zbrush simply offers more. It has been a program I have always wanted to get into but never had the drive to devote my time to a completely different interface. First impressions: I hate it. Zbrush's user interface is increadibly illogical from a 3D movement standpoint to a generic program standpoint. Its layout is oddly set up, models are not models but tools, and its difficult to move around an object - a feature that is extremely important to a sculpting program. Second impressions: I really like it a lot. Its interface may still be odd, but with the addition of a plug-in called ZSwitcher my moving troubles have been solved. ZSwitcher is a third party plug-in to Zbrush that instantly changed the movement scheme to that of Maya or Unity, the programs I use the most. It downloads fast and installs incredibly easily. It does cost 30 bucks, but they have a 30 day free trial that I just started. I would highly recommend this route to anyone who uses Maya often and is interested in ZBrush. With that problem solved I could finally start some sculpting, and Zbrush does a phenomenal job at that.


This is just the start point of the shirt sculpting. It still needs a large amount of work, but the speed of that work is increasing rapidly as I learn the tools better. Once the larger wrinkles are in I will be adding stitching and seems. One particular tutorial I was watching showed an easy way to develop your own seamless brushes in Zbrush for designing your own stitch. I should be able to create a few high quality stitch brushes to use on the shirt and pants to give them the look I want.

Monday, October 15, 2012

Thesis: The Progress

It's been a long time, you monster. This post is Long overdue. It is currently the start of week 7, and  my last post was the start was way back at the end of week 4 so I have some updates for you readers. For week five we needed a design document and a presentation. The design doc is primarily a collection of pre-production work, what is our thesis project, what research have we done, what are our inspirations, what our initial designs look like, what our schedule is, and where are we so far. I focused on my research into what makes a good character in an animated short, what gives them appeal, and why we as an audience root for them. My document also went into my plans and my progress showing sketches and blocking models. Hopefully I can add a link to download the doc as a PDF on here, I will update that as soon as I look into how.

Since the middle of week 5 I have been able to get down to the fun parts, actually doing the project. I have been focusing on the head of my model, the shape, the finer details, the general refinement. I was slightly under the weather week 6 and a bit over worked at my job, putting in around 45 hours on top of classes and this project so I did not get in as much work as I would have liked, but I managed to get through the work (sans blogging it would seem). I have been doing a large amount of technical research, reading blogs and watching tutorials, and have been having a lot of fun doing it.

 I have been looking into hair that can be dynamic and attractive. I started with a hood like shape made of lofted surfaces to get the general concept of how I want the hair to sit on the characters head. From there, I drew CV curves onto the surfaces. Those curves were then made into paint effects shapes to increase the number of curves, add thickness, and add some more natural layers of hair. Next, those paint effects were made back into surfaces, and then back to curves once again, giving me a large amount of curves to work with. I created a hair system and applied it to my newly made curves and started to play with the settings of the hair.



The results were not too bad. This is definitely the more realistic style I am leaning towards. I need to introduce some additional messiness to it as it is currently too clean and neat (specifically on the right hand side ) as well as clean up the back some. The color is not bad, but I may need to make it slightly less red. The good part about the Maya hair system is how easily I will be able to make it dynamic, even though calculating those dynamics may take a while. Hopefully along with a HQ character render by the end of this quarter I can also get a short animation of his head moving to show the dynamics.

So along with the modeling (nicely modeled ear for example) and hair, I have been researching into skin rendering. The above image is a solid start for me to work off of. The subsurface scattering is currently a little too heavy as the algorithm scale conversion is at 8, what it should be around 10-12. This essentially means that too much light is shining through the characters body. His ears are too red because the scale is too small and too much light shines through. I foresee a problem with his cartoon style that may require two skin materials with different scale conversions as his hands will only be slightly thicker than his ear.

I got slightly off course with my schedule, working on hair and skin before finishing the cloths of the model. Because I was never completely confident in the clothing design of my character, I haven't been able to commit myself to finishing the clothing models. My quarrels with the design seem to have calmed a bit more so I will likely be moving forward with those this week.