Tuesday, January 31, 2012

Creature Design in Zbrush - part 1

Thought I would share some thoughts here on my blog. Nothing pretty to show unfortunately, but maybe some interesting ideas?

The idea behind this exercise was to come out of Zbrush in just a few minutes with some rough ideas, or prototypes of creature designs. Thumbnail sketching is always a good way to go. But what I am exploring here are new techniques in searching for that golden idea, and speeding up productivity.

Whether you are doing it for yourself or for a client, you want to get the idea out quick. Or, in this case you want to figure out what the idea even is! The best part about this method? It leads into your next step of production. If you are working for the movie industry or for video games, this will come in handy. You can easily take this prototype to the next stage of development, and finesse the idea even further. Nothing is ever thrown away. And you can even recycle parts of the prototype into another idea.

If you are familiar conceptualizing in two dimensions, then this could easily fit into your current workflow. It's just a matter of getting comfortable with Zbrush. But as you watch these videos, you will see how quickly you can come across shapes and ideas. What you do with them and where you take them is up to you. In my case, I will be taking these proxy meshes further in Zbrush and try to get more information, design, and detail out of the model. In some cases you may just want to take some good screenshots and paint
right over it. This method is for whatever you are in the mood for.

In this first part of the demo, I'm just getting warmed up. Cooking up some ideas. First attempt produces nothing special. But, the techniques here are the same throughout. Starting with any of the Dynamesh spheres in Lightbox, I first turn its resolution down to a reasonable number (about 32 or 64). Once at a smaller resolution, I begin pulling and pushing with Snake Hook brush, Move brush, sculpt with Clay buildup, and Dam-Standard.

The method that produced the best results, was the one that did not have any expectation beforehand. I had no idea what shapes represented what, it was a matter of exploring, and discovery with the sphere. Pulling odd shapes out, adding new ones, and just seeing where it went. Creating interesting negative spaces, and later explaining their function, or reason. The thing with creature design is, you can get very creative and imaginative. And as long as you obey the recognizable laws of nature (anatomy, evolution, an animals survival needs), you can make even a very wild idea convincing.

This is essentially, the character's story. But for our creature, we are telling the story of how it survives. Why it is the way that it is.

This is another demonstration, this time with a little more thought put into the design. I knew what I was going for based off of the first attempts, but still keeping it free and open. Adding, subtracting, pushing and pulling.

In this final video, I decided to create a creature that was more Earth-like in its structure. The previous creatures one could argue were very insect like. I didn't view them as such, with the right treatment and attention to details, you could make them look like whatever you want. But the point of this short clip, was to show that you can take one idea and continue to run with it. Until it becomes something else. I even used a head from another character, and Mesh Inserted it into the Dynamesh concept. Also, you can see I am putting a lot more into this character's concept sculpt. I felt with this creature it was leaning more towards style, and it also had a little more intelligence to it. So I wanted to get a better feel for what/who it was. This okay to do at an early stage, but when you are trying to find that perfect design, it might be best to move on and play with more big ideas. Ultimately, all the concept sculpts that I feel are worth anything will take this path and receive more time.

Analyzing The Ideas

Once the sculpting is done (for now) and you have poked around with a handful of ideas, it is important to sit back and really analyze them for their major shapes. To do this I simply apply either the Flat material (for the silhouette), or the Outline matcap (silhouette plus some contour).

Figure A
Figure B

In Figure A, you can see I have applied the Flat material to all the designs. This displays the mesh without any shadow information. Bringing them all together on one sheet allows me to see what is working and what is not, across all the designs. I also usually take these shots of the model in Orthographic view, to get a true representation of the shape from one simple angle. Makes easier to read, and more graphical. The top row is all the profile shots of the designs, while the bottom row is taken from the front.

My shots used in Figure B are taken with the Outline matcap. A matcap in Zbrush that is used some times for composite rendering. In this case I am using it to show off the design's silhouette like before, but this time in perspective view. You can see a slight turn to the model, and it already feels more dynamic. This matcap also features a rim light, perfect for revealing some of the shapes within the silhouette. Very important when we want to consider where the light will hit on our form.

There are many materials and matcaps to consider when taking these test shots. Think about harsh shadows, and where the dark areas will fall on the design. Also the high points. A material with a lot of specularity is perfect for showing the ridges and planes.

Typically, you will know if a design is working from a silhouette alone.

What Works, and What Doesn't ?

You can see in these images (Figure A and B) that some of these ideas might actually work and be interesting. From certain angles however. It is extremely important to design from all possible views. We are, after all, working in a three dimensional environment. Something I did not do very well here on all of these creatures. So how do you know what works and what doesn't? Well, there are plenty of books on design. And too much information for me to go over in one blog post. But I will point out what in these designs here, is working and why. And what you can avoid in your studies.

  1. Parallels rarely work. Things that are straight, stiff, at right angles, even. These are things to avoid. You can see in some of my designs I fell victim to a few of these symptoms. Legs are too straight from the front view, too evenly spaced between joints, etc.
  2. Patterns are everything. Working negative shapes against positive. Heavy to thin. Tall to short. Chose which you will have be more dominant in the design. You want to balance the larger shapes either on top, or the bottom. With smaller shapes opposing it. Never Large shape, under a large shape. That's splitting your character 50/50. Better example would be the designs in Fig A, the ones that work have a large body and small legs. Or the opposite, small body and large legs.
  3. Balancing your creature. It is important to make it feel like it can support itself naturally. Almost all of these designs have a wide stance, or an even distribution of weight. Keeping their central mass, centered between all points of contact to the ground is very very important. Otherwise it will want to fall over. Keeping the joint's and limbs lose, and comfortable looking. Don't design columns for the creature to walk on! Curves are everywhere in nature. Which leads me into the next point.
  4. Curves versus straights. When it comes to anatomy, it is important to always think of the muscle's origin points. Where is the tension coming from? Usually a bone of some kind. Building up the boney points first (i.e. the joints), and sculpting the muscle in between will save you a headache in the long run. countering a curve with a straight, opposing curves to add flow and interest.
What you should take away from this, my main point, is don't be afraid to be beat up your model. Save iterations, if it reaches a state that you like. Then continue to punch holes, and find those shapes. This was the first part to an analysis of creature design. Documenting some of my findings and personal techniques. In future posts, I will share some thoughts on finding these shapes, and just what shapes you should think about using. Always helps to have a library of parts at you finger tips.


  1. Thanks for sharing this!
    Great post

  2. Good post and very interesting approach to thumbnailing. I will give it a try :)