Making Of 'Madness'
Several softwares were used to create the "Madness" image. The base mesh was prepared in 3D Studio Max 8. I tried to capture the facial expression that accompanies screaming - to catch the grimace at the sight of something really frightening; something, that changes one's life forever. In order to show this fear, I decided to model not only a suitable facial expression, but also the peculiar defensive pose of the whole body. Arms were also arranged in a very specific way, with the right hand trying to protect the head, and the left unconsciously lifted up and pressed to the cheek (Fig.01).
Preparations in Max
I set the figure in the 'T' position. This will allow us to prepare it for further rigging. Taking care of the smallest details, I worked on the upper part of the body, i.e. from the waist up, ignoring the legs altogether, as originally the frame I had in mind was to be a half-view. Finally, I chose the close frame, with virtually only the head visible in the picture. In addition, I modelled the eyes, tongue, gums and teeth. Then, after resetting the mesh (Reset XForm), I proceeded to attach the virtual skeleton to my figure. I used the biped animation system of Max, scaling the bones to give them the approximate shape of the figure. In a moment this would facilitate the bone wagging. The next phase was ascribing the skin modifier, which allows you to deform the mesh by means of the bones. We tick the skin modifier in the stack, and in its options we add the bones ('add bones'). We select all the bones of the biped and proceed to the next phase, i.e. wagging. This is done with the help of the 'edit envelope' button. I use the method of painting the envelopes along the mesh. In the Painter Options I set the Max Strength at 0.15, and Max Size at 0.5. Painting along the mesh, I wag the appropriate bones. The blue areas mark the places of the smallest impact of the bones on the mesh (about 0.1), while the red colour shows the greatest (c. 1.0). All the intermediate colours, such as yellow or orange, proportionally indicate all the values from 0.1 to 1.0. Then I delete the lower part of the body. Wagged in this way, the figure is ready to be exported to ZBrush (Fig.02, Fig.03 and Fig.04).
Modelling in ZBrush
ZBrush software is a very flexible tool for graphic artists, allowing you to model objects easily and quickly. It also has suitable implements to create displaces and bump maps; by means of the 'projection' option it is also possible to paint textures in Photoshop. So, I import the figure ('import') and proceed to model the details. I thicken the figure, adding more and more wrinkles, until the fourth level is reached ('Tool>Geometry>Sdiv 4'). During this phase of mesh thickening, I start modelling the wrinkles, pores and other nuances of the facial skin, using the Projection Master, with only the 'Deformation' option activated. Deactivating the brush MRGB options, I tick only the 'Zadd' and 'Zsub' options, setting the Z intensity parameter to 11. Various alphas will now be used to paint the skin pores. I select the 'Simple Brush', load the alpha, and choose the 'DragRect' stroke. When the wrinkles are ready, I lift the figure from the canvas, thus exiting the Projection Master. I set the figure in a different view and repeat the whole process, painting the skin of the other profile. Now the mapping begins. First we create the texture for our figure, setting the width and height in the 'Texture' options at 4096, and click on 'new'. Then, in the 'Tool>Geometry> mark, we set the lowest possible subdivision level. Then we pass on to the 'Tool>Texture' mark and click on the 'Adaptive UVTiles' ('AUVTiles'), creating automatic mapping for the object. In order to make sure that the mapping is correct, we can also click on 'UV Check Overlapping' option. Now we can make the displace map. In the 'Tool' mark we move on to the 'Displacement' and, with the 'Adaptive' and 'Displace mode' buttons highlighted, we click on 'Create DispMap'. Now the ZBrush generates the map, depending on the speed of your computer. When this is done, the displace map is stored in the 'Alpha' mark. We export it and rotate vertically ('Flip Vertical'), and then proceed to paint the texture. Here I used several samples, pre-prepared in Photoshop - these are parts of photos of several faces. The texture creation, just as the painting of the facial irregularities was done, is done with the help of the Projection Master. The difference now is that we tick the 'Colour' and 'Fade' options, deactivating all the others. I use the 'directional brush' with any alpha and 'DragRect'. The parts of the photos are loaded to the 'Textures' slot. In the centre of the interface I deactivate the 'add' option and leave only the 'RGB' button highlighted, adapting its intensity to current needs. This is repeated several times, exiting the PM and rotating the object in order to obtain a different view. When the whole texture is ready we export it, obviously rotating it vertically ('FlipVert'). I make several sample renderings, and then export the object from the lowest thickness level to the '.*.obj.' format (Fig.05 to Fig.11).
Having imported the object, I set the figure in the desired pose, using the biped bones. The shirt of our figure remains to be modelled. It is created from the figure mesh, by copying appropriate polygons and modelling the collar. We map and set the shader. I copy the 'skin' modifier for the shirt, dragging it from the stack straight onto the object. In order to generate the hair, I copied several polygons from which the hair was to grow. In the 'properties' panel I switch off the 'renderable' option and the object becomes invisible during the rendering. Now only the lights and rendering remain to be set. I used three 'vray' lights and to gain better control over the picture, I split it into several elements while rendering ('diffuse', 'specular', 'colour', etc.). These layers will be used to create the composition in Photoshop (Fig.12-16).
The last phase of the work is composing the renderings in Photoshop. Where any errors appeared, I corrected them using the 'stamp clone' tools. I retouched some of the hair, using other photos as reference.
For the background, I chose a picture taken during a trip to Budapest. The focus depth was simulated by means of the 'lens blur' filter, and through creating a suitable mask. Finally, I applied the 'photo' filter to the whole picture, which made the colours seem colder (Fig.17, Fig.18).