Making Of 'Jaymz'

Hello, my name is Deniz Özemre. Ten years ago I was guitar/vocal in a death thrash band called "Torture". We played lots of cover songs and of course lots of Metallica. Time passed and now it's a good time to remember those days again in 3d. So within this tutorial, I will give you a general insight into this characters creation.

Modelling: The Face

The first and most difficult aspect to this sort of modelling is to nail the resemblence of a character. The character that I have chosen to model is James Hetfield of Metallica. I had to make loads of google searchs to gather as many close up photo of James Hetfield to get the expression that I was looking for, I ended up with these images to use as reference.

In this stage it's impossible to find side/top/front views belonging to the same age so we have to make an average detailing in the evident parts and to focus on them. This photo shows the areas that I've focused on.

I dont really have any development stages of my characters head mesh, but you can be sure it was made from a box. Here is the mesh, however, the mesh developed as the project went on .

The modelling has been done in a polygon style. I've just used the cut, loop and ring tools. There was no need to use any other tools because, after you build the mesh it then all depends on you nailing the details in the correct places .
With the head mesh completed, all that was left was the final pose and here is the base mesh view.

Modelling: The Equipment

In this part, the shirt , microphone, picks, hanger, teeths and the gums were all done using polygonal modelling. Now I prefer to make the shirt in Mudbox or Zbrush cause, in polygonal modeling, it makes too much time and the results are not aesthetic to what I want .


After unfolding the head I exported it into Bodypaint and then started to paint using basic brushes. Meanwhile there are lots of layers made such as wrinkles , pores , hair shadows , beards , skin toning and so on...
Then I have extracted bump and specular maps from the color map with some channel executions. And later some color corrections made in PS .

Other maps are generic with maxwell shaders. And also there was quite basic painting for the shirt .

Texturing: Hair, Eyebrows, Beard

Hair process completed with 3dsmax hair . I wanted it more realistic and made thickness , length , specular , frizz and multistrand maps . After extracting a hair shell from my base mesh , the half side was redesigned , unfolded , mirrored and combed as seen below .

After many render and map checkings , parameters occured as seen below .

Specific hair combing tips:

  1. CTRL+D deselect all hair strands
  2. X+LMB select the strands you want to eliminate
  3. Lock those strand with lock key

Now you can work with the rest

The eyebrows and beard were done using the same techniques .

Preparing the shaders

In this project, Maxwell render was used because it has a brand new material, understanding which is the same in the physical world. There are a few slots in the editor and those are calculated with the real world light parameters and the results, is a physically correct render. Before, I have tried Mental Ray but had lots of problems with the shaders and depth of the view. The Maxwell render engine has lots of fascinating issues in its body, such as auto DOF. It gives physically accurate results and you are always free form caustic, GI and Ä°ndirect Illimunation checks. Those are done automatically by the engine and are really fast compared to the other classic render engines. Now i'll show you two material settings, the other material has been prepared with the same workflow.

Below you can see the Maxwell render editor. When you right click on the "material layers" space there's a list that appears. For this tutorial we will just use "basic layer" and "subsurface" layers .

Now, in the skin layer you can see that I've replaced the colour map in reflectance 0 slot, transmittance map to transmittance slot (for light travel under the skin ), bump map to bump map slot and roughness is full to the top .
Reflection 90 is the side colour level of the IOR which remains white and this does a good job.
This skin layer is for the colour execution of the process and will only show the colour and SSS effect. Then, adding a subsurface layer in the index of the skin layer. Roughness is checked. If it's checked that means there's no specularity or reflection on the surface, so it's like paper. The weight of this layer is 85 so keep this it in mind for later.

Secondly we need to create a specular / reflection / oil surface. We add a second BSDF layer and call it "specular". Set the colour map to the reflectance 0 slot and roughness to 37. In the maxwell material reflection and specularity work in the same channel so that this channel defines by roughness. The less roughness you have, the more reflection/specularity you get. I use this layer only for reflection/specularity. Now, we have two layers, one for the base color and one for sthe pecularity/reflection. In maxwell's render engine, it will calculate 85 and 25 (layer weights) and process those according to their ratios. For example 85 + 25 = 110, this will be summed as 80 + 20 = 100 and the total weights will be exceeded to 100.

For The Teeths

The base (bone shader) colour is white, but at the same time white also means fully reflective. So, I will have to pump up the "roughness" to 99 and silence the reflection/specularity and additional Subsurface layer for directing the light inside. The second layer is to create reflection and specularity. This time Lambertian is off and its much more reflective with white, but with a 30 unit layer weight, I have to prevent it to compress the base layer.

So the other materials were prepared using these methods ...


In maxwell render, it's dont have any light gizmos but planes. We attain a light material to those plains and they become the light emitters. Again we have lots of light presets in material editor which are true with real world physics .

In my scene there are 4 plains and each of them have a strong watt and efficacy parameters. The plains are kept big as the bigger the planes, the easier caustic/reflection/shadow calculations become due to light spectrum formulas from the engine.Testing position/color/intensity of ltheight planes, took me some time but indeed the depth of the light made me happy, finally.


Well I'm not a compositing artist at all, but due to some constraints in maxwell render, I've need to render the hair/beard/eyebrows seperately in a targa file with an alpha channel and combine them in After FX.
In the begining I didin't know how to do this. I searched for a tutorial but couldn't find one so I created my own that I would like to share with you.

For the hair pass I've taken the hair/beard/eyebrows meshes, which interact with the hair in another scene with the camera. Then I rig the light setup for the best results for the hair. After that, I apply Matte/Shadow material to the meshes and uncheck receive shadows. That way some parts of the hair will not be showen where it intersected with the mesh and those areas will be added to the aplha channel as a mask .

For the base pass, I render the whole scene and throw it into After FX with the hair pass. In After FX, I accept the hair pass with the "Premultiplied-Matted with Color" check and paste in on the base render. After adding a background with spots and making some color corrections, our image is ready ...

An alternative Black-White ...

And finally with some additional work ...

That was a general tutorial about how a digital character is created. It was impossible for me to go deep into the detail. I hope you have gained some insight into how the work is done.
Good luck friends!

Fetching comments...

Post a comment