The process of building the geometry was made easy with the use of a mock design in Photoshop, as I was able to build it straight onto a map of the words, removing the need for thought regarding placement or size of words in a 3D environment. I used an extrude generator for each separate text spline to add depth, and then added a background behind the words.
The approach I used for camera paths over the scene was to manually keyframe a camera by controlling it in the first person view and to do the whole scene in one take for continuity and the feel of motion and energy that I was aiming for – as opposed to drawing a spline/path for the camera to follow, and additionally perhaps setting a target for the camera. The reason I used this approach was for the freedom it offered with transition speed and angles between words and phrases, however I soon began to experience the obvious drawback of jerky movements caused by editing the transitions in between keyframes, by adding more keyframes where necessary.
Despite this, I have been editing the f-curve of the camera point by point in order to minimise the stuttering, the cause of which can be seen in the screenshot of the camera’s f-curve below, with each line representing an axis, zoom amount, or rotation that the camera is following. By removing points, and adjusting the curve with the handles, the curve can be interpolated between the points, smoothing the motion of the camera. A problem with this however is the interaction of the various axis, and getting them to match up properly.
As the project continues, I plan to incorporate some of the deformers and/or effectors available in CINEMA 4D to personify some of the more lively phrases to a degree further than just camera angles and motion.