There's no reason to believe that [computer animated] characters will ever live on the screen as the characters do in the best hand-drawn films; given the way that computer-animated films must be made, the vital connection between artist and character simply can't be strong enough.Working off of the above quote, I'd like to talk a little about "the vital connection." Mainly, I want to talk about the technical side of how animators work in various media. There's no question that different forms of animation have different strengths and weaknesses, but, if anything, computer animators have a level of control over characters that easily rivals other forms and in some ways exceeds them.
In stop motion, the animator is limited by the puppet itself. If the puppet's movement is physically restricted by its construction, the animator must adapt to that. There are also limitations imposed by the recording technique. Ray Harryhausen's animation tends to be jittery due to his technology. Because his work was being photographed onto film, he was stuck waiting for it to be developed and wasn't able to relate his current frame to previous ones. On more recent stop motion projects, such as The Corpse Bride, the frames were digitally captured, allowing for playback of previous frames on the set. As a result, modern stop motion animation is generally smoother.
Even with digital recording, though, a stop motion shot still needs to be thoroughly visualized before animation begins. The animation is still being done straight ahead, so timing and paths of action must be worked out in advance and they're not easily changed without re-animating a character.
In drawn animation, an animators drawing ability is roughly equivalent to the limitations of a puppet. With drawings, it is definitely easier to revise shapes and the overall timing of a character than it is in stop motion. Visualization doesn't need to be as thorough as the animator can add or subtract drawings at any time. While it is easier to revise timing or the path of the overall motion, it remains difficult to revise timing on only a portion of a character. Assuming that all parts of a character are drawn on a single level, altering timing for an arm or a leg requires erasing and redrawing before a test can be shot.
In cgi, the limitations of the rig are equivalent to the limitations of a puppet. While I'm sure that cgi animators all have their pet peeves about the flexibility and controls of rigs, the rigging at studios doing high budget features is very impressive. There is quite a bit of flexibility of a character's shapes, though not as much as pencil animators whose work is heavily graphic, like Eric Goldberg or Fred Moore.
Timing in cgi is far more flexible than in stop motion or drawn animation. In cgi, it is trivial to alter the timing on the arms of a walking character. It literally takes seconds to select the relevant arm controls in the dope sheet and slide them forwards or backwards in time. Timing can also be globally or locally compressed or stretched in the dope sheet. This makes trying variations more practical than they are in other forms of animation. Paths of action for an entire character or just a part can also be altered with far less effort. If anything, from a technical standpoint, the level of animator control in cgi is equal to or greater than stop motion or drawn animation.
Yet Michael Barrier and others somehow feel that cgi character animation is lacking. Why? One possible answer is the need for pre-visualization of a character's actions before starting to animate. A stop motion animator must do this more than a pencil animator and a pencil animator must do it more than a cgi animator. If this was what was bothering people, then stop motion animation would be the gold standard and that doesn't seem to be the case.
Perhaps it is the animator's interface for creating motion. Stop motion animators put their hands on the puppet to manipulate it. That makes for an intimate relationship. Drawn animation is done with a pencil, something animators have used for 15 years before entering the industry, giving them a greater familiarity with that tool than with a computer mouse. A pencil certainly expresses individuality better than a mouse does. An artist's line is a form of a signature, though in drawn animation the animator's lines are often homogenized by assistants for the sake of consistency. A cgi character will automatically look consistent, though nothing stops cgi animators from having as individual a sense of posing and timing as any other type of animator.
Another possible answer is that the ease of revising cgi leads to over refinement. It's sort of the difference between whole wheat and white bread or molasses and white sugar. In both cases, the refinement leads to blandness. While cgi animators can revise more quickly, the footage quota on cgi features is not higher than in drawn features of a similar budget. The time saved goes towards refining the surface. There are few imperfections in the movement, which may lead to a kind of sterility.
While cgi lends itself to this level of refinement, it is not a necessity. As I've said, artists make decisions and some of them are bad ones. This is why I think that blaming a form of animation for the weaknesses in a film is wrong. The bigger problem is not the technique, but how the characters are conceived. I'll take up this issue in a future entry.