Making Of 'This Little Pinkie'
Hi all. In the article below I will briefly describe the steps used to create "This Little Pinkie..."
The idea was to practise cartoon character modelling. Rather than design my own character I wanted to take a popular cartoon character and recreate it in 3D. I have always loved the Warner Bros. cartoons, and in particular the episodes featuring Tweety Pie and Sylvester the Cat.
I searched the net for some pictures and info on the pair. What I found was that Tweety had changed quite a bit from when he was originally drawn by creator Bob Clampett to how he was drawn in the 80's. I decided to keep him closer to the compact design of the original. Sylvester didn't really change much apart from originally having a black nose.
For this project I used Maya 6.0, but the same techniques can be applied to other 3D applications. Before I started modelling, I roughly drew front and side views of each character to be used as image planes and act as a guide for the modelling process.
For almost all of my modelling I use the poly-poly method. For a character I typically start with a single polygon face, and extrude the edges to form to the basic shape of the eye.
From there I continue to extrude the edge loops and move the vertices to give me the desired shape, and edge flow. I only model using quads as it gives you a nice clean mesh and results in a more predictable surface when subdividing.
I kept this stage of the process really basic. For Tweety I just used simple blinn and lambert shaders. Rather than worrying about UV mapping, I selected the faces I wanted for a particular shader and saved them as a Quick Select Set. That way I could easily select those faces again and assign a different shader if required.
For Sylvester I laid out the UV's into four separate maps, using planer and cylindrical mapping methods. I opened those maps in Photoshop and painted the black, white and red areas. From there I took the model into Deep Paint 3D to fix up the seams in some areas.
Rigging & Posing
This process was really streamlined by using the rigging tools available with the Maya 6.0 bonus tools. These tools basically automated the rigging process, and gave me a rig that was good enough for what I needed.
The eyes on both characters aren't perfectly round, so I used the texture projection coordinates to control the eye direction.
As far as posing goes, I wanted to introduce some squash & stretch as used so well in many of the Warner Bros. cartoons. This was done by just moving the joints in Sylvester's arm rather than going to the bother of setting up a stretchy arm rig.
Lighting & Rendering
For lighting the scene I used a basic skylight and area light approach. For the skylight I created a poly sphere, scaled it to surround the scene and then deleted the bottom half, leaving me with a dome to simulate the sky. I assigned a lambert material with a ramp connected to the incandescence channel. This will be used to emit Final Gather rays when rendering with Mental Ray. In the render attributes of the sphere I turned off the primary visibility so that it is not seen in the final render.
I then created a polygon plane and placed it below the characters to simulate the light bouncing off the ground. The plane was assigned a new lambert shader with the colour and incandescence channels set to a light brown colour.
For the sunlight I created a spotlight with intensity set to 0.850, the colour somewhere between pink and orange and with raytrace shadows checked on. I converted the spotlight to a sphere area light by checking the area light box under the Mental Ray section of the light's attributes. I set the sampling to 10 and the low sampling to 5, though I've later found that a sampling of 8 works just as well and renders a bit quicker. The light is scaled up a little, as the scale of the light affects the softness of the shadows.
Finally in Photoshop I painted the sky to place in the background of the render. For the sky I have a layer which is a gradient ranging from a pink to a light blue (image 1). On another layer I painted some simple clouds using a soft basic brush and smudge tool (image 2). I then placed the render on the next layer (image 3) and added a Hue & Saturation adjustment layer to boost the saturation of the render a little. One thing to note when rendering layers for compositing is to turn off premultiply (under framebuffer section in Mental Ray render globals) so that you don't have the background colour as a halo around the render. Finally I have a Brightness & Contrast adjustment layer to increase the contrast and darken the image a little. (image 4)
And that's it. Thanks for reading and hopefully some parts of this article may be of some use to you on future projects