Making Of 'Chieftain'
My name is Alander Alexei. I am 25 years old and currently live in Finland. I have recently graduated from the University of Hertfordshire where I have been doing a 3D digital animation course. This character was a part of my final year project.
Researching the Subject
Before any modeling or concepting, I usually try to find a lot of visual references for my character. I like to use Google, Flickr, Deviant Art or any other image-searching engines. I think this step is crucial if you are trying to make a good-looking 3D character. Another good thing is to be inspired. I look at other characters and I try to understand what makes them so good. When I feel I am familiar with the subject, I start doodling some small thumbnails.
These thumbnails show different combinations of the references I'd gathered (Fig.01). At this stage I was almost sure how my character would look when finished. I then chose some good elements from all the thumbnails and photos (Fig.02) and started the sculpting process.
Personally I like to make everything from scratch, but since we had a tight deadline and a bunch of other projects I decided to save some time and used a base model to start with. This model was just a low poly version of a human body. I then tweaked it and made it so that it was more appropriate for my character. Here is the base mesh after tweaking (Fig.03). There was no need to develop the anatomy any further because it would be hidden underneath the clothes. The only parts that needed detailing were the hands and the head.
When the base proportions were done I started on the face. I think the face is the most important feature of any human character and this is why I tried to spend a lot of time on its development. It is very important to have anatomy references before you while sculpting human a body. While developing the face I made lots of variations and then chose the best of them (Fig.04). This is the final sculpt of the face (Fig.05).
When I sculpt clothes I try to have as many references as possible, preferably from different angles. However, this is quite hard to find on the web and so I take pictures of myself wearing different types of clothes. I find it very helpful to see the same reference from different angles and it is also a great way to study the subject even better (Fig.06 - 07).
I used random cloth texture as a base and blurred it to get basic color variations. This blurred texture was placed on top of the occlusion layer in Multiply mode. The occlusion is in Normal mode. Then I used a cloth pattern, which was also used in ZBrush with Noise Maker, to make details for the normal map. Then I added some dirt textures as overlay to break the perfect and clean look. I added lots of different color corrections, and some brown-orange color to break the green look. The color was added in Normal blending mode, but with opacity of around 70%, and I used a soft, large brush with low opacity of around 30%. Finally I copied the AO, placed it on top and masked it with a black color. Then I painted white in the mask channel in the places where I wanted my highlights to be. The Highlight layer was set to around 70% opacity. That is pretty much it (Fig.08).
This is a low poly version of the tunic with the texture applied in the Marmoset engine (Fig. 09).
I did the retopology for the high poly head in Topogun. Then I unwrapped the head in Max and used Xnormal to bake my cavity, convexity and occlusion maps. These maps were used as a starting point for the face texture.
Next I opened Photoshop and loaded the occlusion map. I used it as a guide for making the first layer of the head texture. The first layer was the photo image, which was done by combining and stitching various photos from different face images from the 3D.sk website (Fig.10).
Now I hid the photo layer so that only the occlusion layer was visible. I then chose a soft brush and painted in some color on top of the occlusion layer. I merged the occlusion and the paint layer to get a colored occlusion map (Fig.11).
I set my colored occlusion layer's opacity to 40% and made a few hue/saturation adjustments (Fig.12).
The texture looked a bit blurred now because I had painted on top of the occlusion layer with a soft brush. To fix this I added a cavity map in Multiply mode on top of the current layers. I also added some color adjustments to the texture (Fig.13).
I constantly checked my textured model in the Marmoset Toolbag engine so that I knew how the texture looked in 3D. At this stage I noticed that it was too bright in the engine and it also looked too yellow to me. So I added a few more adjustment layers on top to make the skin darker and more bluish. I also added a bit of redness and highlights to some parts of the face (Fig.14).
Now I converted the texture into a grayscale image and fixed some errors, like black moles. They needed to be white because they had to push the surface. Then I opened ZBrush and loaded my grayscale image as an alpha image. I chose Mask by Alpha and used the Inflate option, which can be found in the Deformation palette. Make sure to do Morph Target before doing Inflate. The amount of Inflate was about 5 in my case.
Then I used the standard brush to add more depth to some wrinkles and I also used my morph brush to remove the inflate effect in some areas like the ears and eyelids, because they looked too fatty now (Fig.15).
Here you can see the final normal map of the face (Fig.16).
For the specular map I used the base photo layer with a cavity on top. Then I added a bit of a Noise filter in Photoshop and changed the overall color to blue in order to get a bluish tint to the skin. I also added some highlights to the nose, lips, cheeks and forehead (Fig.17).
Here's the final image (Fig.18).
I really hope you have enjoyed this Making Of and have picked up something new! If you have any questions, please contact me via email at any time of day and I will be glad to answer your questions. Thank you for reading!