Making Of 'Sabretooth'
Sabretooth was my favorite villain from X-Men and on my list of characters I always wanted to make. The first thing I did was to collect as many references as I had time to and find a look I wanted to use. I plan to use this model in a Sabretooth vs. Wolverine scene when I get the time.
I always model in a T-pose to utilize the symmetry and so my models can be easily rigged. Modeling was started with a base mesh I already had from an older project. Before moving into ZBrush, I scaled my model to a realistic size in 3ds Max (Customize > Units Setup). If the scale isn't set to a real size, you will have problems with your SSS shader and other rendering settings.
I used ZBrush for sculpting the anatomy silhouette, using the Move and Move Topology brushes for the first few subdivisions. The muscles and leather folds/details were sculpted in the higher subdivisions, using the Clay Buildup, Dam Standard, Flatten and Move brushes. The Flatten brush was used in place of the Smooth brush to even out the surface. I find the Flatten brush gives more control and maintains more sculpted details.
TopoGun was used for retopology and then cleaned up in 3ds Max. I used the muscles as a guide for the edge flow on the body, so the mesh would deform correctly when the limbs were moving. Also this allowed the base mesh to capture more of the silhouette and form, resulting in less displacement (Fig.01).
The UVs were cut up (creating seams) using the new feature in 3ds Max 2012: Unwrap UVW modifier. Once the UVs were defined, I used GoZ to export the model into ZBrush and used the awesome plug-in 'UV Master' to unwrap (flatten out and relax UVs), keeping the Use Existing Seams button active to utilize the premade seams in 3ds Max. Once the UVs were flattened, I used GoZ to take the model back into 3ds Max to rearrange the UV islands for easy texture painting.
The first thing I did for the textures was to decimate the high poly model with Keep UVs checked before taking it into 3ds Max. In 3ds Max I used Render Surface Map in the Render menu, and rendered an Ambient Occlusion, Cavity and SSS map at 4096 x 4096px to use for my texture maps.
For the texture painting, use reference material. Try to perceive what is color and what is light shading, reflections and specular. Not many objects are only flat colors. Shadows, reflections and specular values will have colors of their own and change the look of an object/material.
I used ZBrush to polypaint the diffuse map (Fig.02) and the hair density masks. To paint the leather suit I painted the base flat orange and brown. Then I used Cavity masks and a variety of alphas, and the drag stroke with a Standard brush to add the worn leather effects. To make the texture look more believable I wanted to add as many defects as I could. Some things to look out for are sweat marks on clothing, dirt build ups, wear and aging of the material. Leather has a lighter brownish yellow when it cracks from excessive rubbing.
My model was seven subdivisions in ZBrush. The displacement was generated from subdivision 1-4 and the normal map was generated from 4 - 7. Displacement maps were generated using ZBrush as 32-bit EXR files (Fig.03).
Hair Farm was used to make and render the hair. I used black and white masks to specify where the hair would be placed on the model (Fig.04). The hair was styled and hair shaders were made in a separate 3ds Max file using Scanline Render Engine, because of the speed in rendering. Standard spot lights were placed in the same position and intensity as the original V-Ray light (also required was lighting the hair in the final V-Ray render) (Fig.05). When I was happy with the hair test renders I imported the standard lights and hair into my V-Ray scene.
At this point the V-Ray scene consisted of V-Ray lights, standard spot lights (for illuminating hair only) and a HDRI placed in the V-Ray: Environment slots of the Render Setup menu. It sometimes helps to lower the gamma of your HDRI so it has more contrast, but it can cause a spotty light effect. It depends on the quality of the HDRI being used. SIBL maps were utilized for textures in the V-Ray lights so they had a natural feel. I used a soft box image from www.hdrlabs.com (Fig.06).
The skin shader and SSS shader were very simple, as the model was going to be seen as a whole and not close up (Fig.07).
I used a linear workflow in V-Ray (just Google it, there are plenty of online tutorials) to give me more control over the final render passes. I rendered all passes as 32-bit Open EXR files so I could control the gamma in post (Fig.08).
If you look at Fig.08 you will notice that it acts in a similar way to a HDR file, allowing you to lighten or darken each render pass without blacks and whites blowing out and losing information in your image. It's very handy to have such control in post when a single render can take from 8-24 hours. This also works for rendering animations as an image sequence.
In the VFB (V-Ray Frame Buffer) I used the 'Display colors in SRGB space' to give me a preview of what the final render would look like in SRGB. It was a bit washed out, but the saturation could be adjusted in post.
This is my favorite part of the process as I get to see what all my hard work is going to look like. For this I opened all the render passes and adjusted the gamma on the passes I felt needed it. Some tips I used are in Fig.09.
I hope some of you have found this helpful and I appreciate being given the opportunity to give back to the 3D community that helped me so much to learn 3D.