Making of The Artist himself

So I try to reconstruct my old picture. To do that I used recent pictures of my face and later I changed geometry to make it similar to my self 6 years ago. Of course I had to upgrade my face to look, let’s say, better :)

Introduction

Most of my works are technical experiments. I’m applying new techniqes, new ideas, solutions. Similar situation was with my self portrait, but in this case I was trying to do something technical with some art in it.
Idea come from old picture of my face, it was the best picture in my life. By some accident with my hdd, I’ve lost it, and only copy whitch I found was my avatar on cgtalk. So I try to reconstruct my old picture. To do that I used recent pictures of my face and later I changed geometry to make it similar to my self 6 years ago.
Of course I had to upgrade my face to look, let’s say, better :)

Modelling

To model my face I’ve used old model, but I change topology to be more universal for future experiments. I spend lots of time to have my model accurate as much as possible. Then I’ve made UVs, to do that I’ve used standard maya UV editor. My goal was to do UVs optimized for the face, not to waste any space of the texture. That’s why I made its face to be as big as possible and back of the head didn’t use UVs space at all. I still have hairs there :) So texture is not needed :) Maybe in the future…

Lighting

When modelling part was finished, I took avatar and placed it on the background, and I started adjusting camera lens and angle. I was suprised ‘cause my face didn’t change so much :) But I had different face expression so I was trying to adjust this detail. It was quite hard because my avatar has 150×150 pixels, and there was also strange light. To make everything more complicated picture was changed in photoshop: contrast, brightness and curves. Maybe something else but I don’t remember after such long tme. For my old picture I was using 2000 watt cinema lamp so there was ultra strong light on my face. I chose area light because my lamp was 19 inch fresnel lens and there was a big lighting surface, not a point light. For the shadows I chose soft raytrace shadows. After 2 nights of playng with angle and distance of light finally I found the right position :) It was close enough to proceed with textures.

Texturing

I decided to use 4k textures, because my plan was to render it in 3k. I didn’t have my precious move lamp anymore so I had to do many pictures of my face from different angles to have sharp texture in every place.

Because of the camera depth of field when there is not enough light you have to go lower with iris to have still pictures. Using high power lamp you can set iris higher like 11 or 16 and still have quite good time like 1/60s or 1/125s than almost entireface is sharp. To make those textures I was using canon eos 20d.
I took those pictures to photoshop and i scale them down to 1500*1000, to have possibility to rotate camera in maya. And I placed them as the image plane. Later I set camera in maya to have similar lens length. Than I tried to place my 3d face exactly as it was on pictures and bake current piece of texture using full resolution images. When I baked all of them I’ve started puting them together in photoshop, tweeking color, brightnes, etc, to make them look as one image. Have to say that was hard work.

Than the time has come for misss fast skin shader. I’ve created about 11 textures which I’ve connected to the shader in those slots:

overal color – texture

diffuse weight – mask

epidermal scatter color – texture

epidermal scatter weight – mask

subdermal scatter color – texture

back scatter color – texture

primary specular color – texture

reflect weight – texture

bump – I used layered texture and I mixed there displacement, bump and painted in photoshop some additional bump details. I was trying to use displacement but there was no visibile difference compare to displacement as bump so I decided to combine those two and use as a bump.

For the eye I also used misss fast skin shader, but with less amount of textures. To have more interesting light on the iris I put bump and deformed a bit the geometry. Then I duplicated eyelid area and deformed it to fit the eye. I also created create transparent blin shader with mask on reflection and specularity to add it later as a separate render layer. Also I added abmient occlusion on the separate layer.

Hairs

When I’ve finished more or less texturing and shading I had to do some hairs.
for beard. It was fast and easy. I duplicated my face and cut everything except part where beard is gonna be. Than I did new UVs, because fur is taking UVs space to place fur. After that I painted fiew attributes and it was finished. Eyelashes were done in the same way, but a bit more painfull.

For the eye brows – the same technique but I used hair system to attract fur. So I started adding hair by hair and dress it on start position. Later I connected fur and tweek to have good result.

Hairs…. This was the way through pain. I was close to bite my hands off. I did like 3 setups, and all of them were looking fake. I also used hair as attractors for the fur. My final solution was to create hairs clump by clump. Later I took the same hair system and changed values to have some puffy dry hairs. I did a lot of painting to tweek values of those hairs. And hairs attractor strenght to have exact shape of clumped hairs.

Rendering

For render I used maya mental ray. Rendering those kind of resolution is a bit tricky. Because if everything look fine in 1k in 3k become very wrong. Bump become huge, every edge is sharp. Hairs look like wires. Because of this I was rendering all the time in full resolution. Thanks to render region it didnt take ages :)

Compositing

I rendered almost everything as separate layer. And I was checking how it looks on the finall composition, back and foward. It gives much more flexibility in animarion or still images, and saving alot of time. You can see the composition process on the pictures above, from maya render to the final image.

mental ray render with invisible hairs casting shadows. i use it to have credible shadows in misss fast skin shader

with beard added
eyelashes and brows additional shadows for the hairs, for that i just render hiden hairs casting shadows
I added hairs, eye liquid and additional eye reflection all layers together
I rendered Zdeph from maya to add depth of field in postprocess final image with dof and lens distortion filter to give impresion of photography

Something about me

Since I was a child I had strange need to create. I was bulding strange machines, airplanes models, electronic things, etc. One day my parents bought me a camera, since that moment I wanted to be a photographer. To make my dreams came true, I went to the school of photography and after it I was a press photographer for a 10 years. Press photography became boring to me, and I got a new idea! I want to do special effects for the movies :) So i was learning like minimum 18h per day 3ds max, reading books, magazines, everything where was something about 3d and movies. But at the same time I still was a photographer After 2 years learning slowly I’ve started working in 3d industry. Later I switched to maya. Now I’m traveling from country to country working on different projects. Right now I’m in london :)

Related links

3D Tutorials & Courses Online at 3DExport

About The Author

You might be interested in

3DExport Buy & Sell 3D Models