Making Of 'The Orange'
The idea behind this character was actually very simple. I've always liked robots, so I thought I'd make one with a clean and simple design. I thought I would make it look a bit retro rather than similar to the modern robots that we see a lot of these days.
Shapes and Weights
I started by sketching in Photoshop using a few simple lines to create a nice, well-distributed shape. I then dedicated some time to making sure that the character looked balanced. Every image has a virtual weight that is perceived by the viewer's eye. This weight is demonstrated in many elements of the picture like the composition, light, shadows and shapes. When I had spent some time sketching and considering my design, I came up with what you can see in Fig.01.
The next step was to begin the model. Similar to the sketching phase I started with basic primitives and shapes. This step was when I was able to finalize the shape and form of the robot. In Fig.02 you can see the basic model, which I created in about five minutes.
With the box model in the background I started to model and detail each individual part. The model wasn't complicated to create because of the simple forms so I won't talk too much about the modeling process. The only important thing to talk about and explain further is the hard surfaces. The model was made from individual subpatches in LightWave. This is basically a process by which the basic geometry is smoothed to create simple forms. The only problem with using this approach is that you lose some of the hard edges in the corners of your model.
As far as I am aware there are two solutions for creating hard surfaces using this technique; you can either weight the suitable edges or add additional cuts. I excluded edge weight because it was not compatible if I decided to export the model and use it elsewhere, so the trick I chose was to just add extra edges loops near the borders that needed to be sharp (Fig.03).
The last step after completing the mesh was to add the rivets. To add them I cloned one using a plugin, which repeated them around the body. Rivets were an important detail because they broke the plain surface of the body of the robot and at the same time they added some nice, smaller details (Fig.04).
The model was now ready to be posed! You can see the wireframes of the individual parts in Fig.05.
Rigging and Pose
I decided to rig the character so I was free to pose him as I chose and also so I would be able to animate him in the future. In LightWave you can rig a character in two different ways. You can use bones, which you assign to each part of the robot and apply weight maps to, or you can use hierarchies and separate all the robot parts into layers and animate them using their original pivot points.
To describe all of the rigging process would require me to write a very long tutorial, but I would always suggest that you use the first method as I did because it gives you more control and causes fewer headaches (Fig.06).
Materials and Lighting
The materials I chose were basically simple and clean. Chrome was used for the joints and rivets, rubber for the hip junction and a basic orange, blurred metal for the rest of the body.
The layout of the scene was pretty simple too. Basically I recreated a white photographic studio set with rounder corners, a luminous polygon and a HDRI. The luminous polygon really makes the difference as it creates a nice diffuse light, and at the same time adds a nice reflection to the model that highlights the shape and adds more depth to it. The HDRI was very subtle, but did enough to add small reflections on the rivets and eyes too. I chose a vertical format for the image to match the shape of the robot (Fig.07).
Post Processing and Color Corrections
Once the final render was complete it was time to do some work in Photoshop. The final render actually looked very nice, but when I thought about the original concept it was clear that the render didn't have the vintage look I wanted (Fig.08).
I thought that this could be fixed with simple color adjustments so I researched vintage photos and found examples of the type of image I wanted to create. I noticed a color gradient that seemed similar in a lot of the photos, so I tried to apply this to my image. To do this I used Curve corrections as they give you more control than if you just use the Color Balance tool (Fig.09).
Due to the anti-aliasing the rendered image lost some sharpness in the final output. To fix this in Photoshop I used a High Pass filter. The High Pass filter finds the most contrasting areas of the image and masks the rest with a 50% gray. The higher the pixel radius is in the filter panel, the sharper the image will be. After creating the high pass mask I set the blending mode to Hard Light to enhance the general reflection on the image, especially on the chrome (Fig.10).
The final step was to create a nice gradient in the shadows. I used an orange/blue gradient over the image and set the blending mode to Soft Light. This mode is similar to Overlay, but more subtle and useful when working on shadows. These notes about blending modes are not absolute rules and cannot be applied to every image; the key is to test the modes and see what works for you. When this was done the colors looked nice and I was happy with the image (Fig.11).
It's not very easy to describe the creative process behind a 3D image as every step is made up of many more small steps, but I hope that in some way this tutorial is helpful to you and that you enjoyed reading it.
To see more by Riccardo Zema, check out Prime - The Definitive Digital Art Collection