Realistic Pharah Fanart
Serguei Krikalev talked about his realistic Pharah fanart made within CGMA course Character Creation for Film/Cinematics led by Peter Zoppi.
I’m Serguei Krikalev from Brazil, Rio de Janeiro, and I’m 30 years old. I work as a Character, Texture, and LookDev Artist. My main projects are in the advertisement segment in Brazil at IMGTV, 3Dar from Argentina, Ephere Inc., plus, I love to do personal projects.
Character modeling has always been my passion, so I needed to learn something more professional, see how everything works in the environment production and how great studios work in that area. Peter Zoppi is a great instructor, and he prepared me for what I needed, such as building characters for an efficient deformation, UV mapping and its nuances, texturing, practical rigging for character presentation, and others. It was simply sensational.
From the beginning, I wanted to do an appealing character. I started to do researches where I found the Pharah interpretation by the amazing artist Yi Sui, and it was love at first sight. Her look caught my attention. I built a simple reference table (RefTable), as shown in the picture below.
In this work, I used a base mesh to gain time and only made the needed adjustments. The concept itself has already nice facial features so I only needed to put my touches based on a mix of other references. In the end, I always make a ZBrush render to see what I need to fix to improve and to finish my blockout step.
The eyes construction is relatively simple, I used two geometries: one for iris/pupil and the other for the sclera, just like most artists do.
The main difference is perhaps in the textures and shader process. All my textures are made in Mari, at least for the organic pieces (I’m still integrating Substance into my pipeline), and with the big help of the Texturing XYZ maps, I got amazing details. I made my Albedo/SSS color, displacement, and normal maps and exported them to Maya to make the LookDev. An important thing: Renderman is your SSS shader. It’s a bit complex to use at first, but sometime later you will figure it out and never stop using it. A little tip: please, consider using the Non-Exponential mode at PxrSurface. For the blending between sclera and cornea, I used the PxrLayer Shader with a ramp to get the desired shape of the cornea. Everything has a simple approach and a lot of try and error.
Hair is always a challenge, but Ornatrix allows me to create it very easily. Actually, I’m an Ornatrix contributor and Beta tester from Ephere. The non-destructive and intuitive system driven by layers allows people with a minimum of knowledge in Photoshop to get awesome results in a very short time. Ephere has many new things under the hood. Worth a try.
Every artwork needs some planning and the hair part is not an exception. I start doing some annotations about the CG Hair I want to bring to life, put down things like frizz, clumps, hair shapes I’ll need and others observations related to it.
With the initial step done, I create scalp geo with UV ready to use as a base for my hair.
After that, I positioned the hair guides to fit the concept.
Here is one of my first render tests after adjusting my guides a bit.
As you can see hair takes time, so references and patience are the key elements here. Generally, I take a few days to finish the hair, and the results worth the time. Just remember, everything begins with good planning and collecting some references.
In 2018, before I took the course with Peter Zoppi, I finished another class at CGMA, Texturing for Film/Cinematics with Chris Nichols where I learned amazing techniques for texturing. So, when I came to the character course, I had a good background in this area and Peter only pushed up my skill.
My skin textures are made in Mari. In general, my first channel is the fine displacement because with it I can generate maps like cavity, specular, and use it as a channel mask for the albedo texture. I make simple projections for that and do some render tests to find possible problems and fix them.
When I am ok with the fine displacement, I go to my albedo maps using the same approach: projections and render tests. In this case, I did a mix of a simple color map done in ZBrush and my base projection in Mari to see how everything is going.
More adjustments for skin tone and some maps that do not depend on albedo.
This is the part of the project where you need a little bit of patience and observation. After some feedback, I saw that there was a certain need to bring the model a little closer to the original character of the game without losing the features of the concept. So again, I collected more references for the original Pharah to compare some of the features and obtain the following result.
Marvelous Designer was a big help as always. The workflow is well known: first, we get the mesh in Marvelous, export it to ZBrush or Maya for retopology, and transfer the attributes into Maya. The main trick here was in UVs thinking. Keep in mind the directions of the UV. This turns your life easier when handling the textures.
Here is my first pass after the retopo and some render tests for the material visualizer. Take a look at the borders and note that I need some work there.
For texturing this piece, I used Substance Painter. Here are my first texture pass and the render test. I made some annotations for the render test, too.
And below is my final piece with the modeling fixed and texturing finished. You can see I inserted some wrinkles made in ZBrush to give a more realistic sense to the clothes.
Renderman is my renderer of choice. In version 22, Renderman was optimized, so it is now extremely comfortable to use. Each version is more and more user-friendly.
Generally, all my render tests are made with low quality. This way, I can improve some settings to see better what is happening with samples and antialiasing. Most of my shaders are layered, and this is really helpful. Some lobes like specular and glossiness can be controlled with more freedom. Of course, good maps like showed above in the skin part help a lot, too. There is no secret in Renderman. Everything is well documented.
The skin is composed basically of three shaders layered by PxrLayeredSurface: one for the skin itself, one for the makeup (this allowed me to control the color and glossiness), and the other is for the tattoo. I always try to keep everything as simple as possible. Simple but effective. In one of the workshops from Disney, I learned: “Work smarter, not harder.”
Finally, here’s my lighting test:
Guys, I hope this was useful for some of you. See you next time!
Feel free to contact me for more details at:
- Instagram: S_Krikalev