Get inside the mind of Spain-based CG artist as he reveals the creative process behind Rob the Robot.
I started working with 3D in 2004, mainly based in advertising, and I am now a lead artist in textures and shading for environments and props for Kandor Graphics. I find working on personal projects keeps my skills fresh and current.
As an artist, I am always inspired by other artists and seeing their work - people like Juan Siquier (with whom I have had the pleasure of working), Ian McQue, Michael Kutsche and Olivier Ponsonnet, among others.
From conception to completion, this image, titled Epic, took me a month to complete. I don’t mind admitting that I’m really pleased with the result. All the modelling and rendering is my own work except the face, which is a copyright-free emoticon.
01. Modelling
Beginning with a simple hand-drawn sketch, I add a proxy model to obtain the initial shape. After this, I use reference photographs in order to detail every part until I reach the hi-poly or final model. It is important to use the tools available to really do the job well. On occasion, it is best to do a quick ZBrush model, and then do a retopology. In this case, in parts such as the trousers, the gloves and the vests.
Domingo added a proxy model to obtain the robot's initial shape
02. UV mapping
With UVLayout, I open different and parts then group them into six models for more comfortable texturing. I personally do not organise or distribute the islands. What I do instead for texturing is to take a render ID for each object and use it as a mask. This is so that I differentiate between materials, metal, paintings and plastics.
Dividing up the model into different parts, dependent on the material that is being rendered, can make it easier to render the model
03. Texturing
First of all, I work on the base textures that I will need later on: metal, rust, leather, fabric, paintings, and so on. Afterwards, using render texture, I extract the ID maps, AO, displacement, normal map and UV. In the end, it is all a matter of painting and texture use. I always recommend using a tileable texture base. In this case, there are nine textures in 8K, three of which are for the terrain.
The face reflection level was created before the emoticon was added
04. Shading
For each part of the model, I extracted the polygons that would be the same material. Then, it was a matter of using the properties of each material, specular and glossy, according to the finish I was looking for.
The head monitor was tricky to map as it was built up of different textures
05. Illumination
For the exterior lighting, I have used a direct light with Sun and Sky in the environment and a HDRI for the GI, in order to pick up the areas in the shade. It is no more complicated than this. I always use a Physical Camera from V-Ray.
"The compositing of the model took a lot of tweaking to get it the way I wanted," says Domingo
06. Composition
One of the most important steps was adjusting the colour, saturation and contrast
This is one of the most important steps. Playing with colour, saturation and contrast is essential. It is also important to use all the maps available, using Render Elements from V-Ray - as is creating an atmosphere that will highlight the object and that will be in harmony with the story the illustration is telling. It is also useful to employ fog and noise.
07. Final
As a last step, I would like to point out that it is important to get feedback from someone who has some 3D knowledge, to avoid mistakes that may have gone unnoticed. In my case, I was helped by Victor Loba. I certainly owe him a dose of thanks for helping me improve
This article originally appeared in 3D World issue 173.
