9.5.2

1002 • CHAPTER 19
TIMELINE • 1003
 Play back the animation and watch how the sphere’s motion has been
changed. Now the sphere reaches X=300 by about frame 45, then it ‘reverses’
all the way back to its starting point by the end of the animation — even though
the Position.X track has just two keys. The Time track controls how the sphere’s
motion is interpolated between the keys.
Why Time curves are useful
Suppose you’ve created a complex camera ight that has 40 keyframes dening the main positions
the camera passes through. You then decide you want to slow down or speed up the camera in
particular areas — adjusting the keys accordingly would take a great deal of effort. By using a Time
curve, you can achieve the desired control over the camera’s speed without having to change the
actual keyframes.
Or suppose you’ve animated a walk cycle for a character consisting of 253 keyframes. You want the
character to take two steps forward and three steps back. You could either copy the keyframes several
times and adjust them or simply use a Time curve with six keyframes to get the same result in less
time.
The accuracy of time curve points is limited, because they are stored internally
as oating point numbers. In extreme cases, rounding errors can come into
play and a time curve point could, for example, end up being positioned at
frame 2340.38 instead of at frame 2340. The problem is magnied if you are
rendering extremely long animations. In such cases, you may nd it helpful to
render the scene in several parts instead, with each part saved as a separate
scene le.