Datasheet

10■n Chapter 1: Learning the Basics of Lip Sync
Open Mouth Sounds
Many sounds have no real shape to them, so they’re out as visemes. Another group of
sounds have no shape in the sense that the lips aren’t contorting in a particular way, but
they have the common characteristic that the mouth must be open. These sounds are
listed in Table 1.2. I don’t consider these visemes but instead refer to them as open or
jaw sounds. Visemes as we identify and animate them are really aspects of lip positions,
not whole mouth positions. Because the jaw, and therefore the mouth, is open in many
shapes, I’ve just kicked those shapes out of the viseme club, which makes things simpler.
For example, an OH sound (which should be read
as a very short OH, not like the word oh, which would
be OH-OO) is just a degree of Narrow and some
Open—which is really the same as an OO sound but
with different amounts of Narrow and Open. Instead
of referring to sounds as their phonetic spellings,
such as OH or AW, I like to break them down further to their components. OH and OO
have the same ingredients, but they’re mixed in different amounts. By separating things
out into some basic elements like that, you can animate faster and better and more pre-
cisely tailor your shape to the sound you hear. Again, this isn’t saying to break down OH
in time by opening it first and then making it narrow, as in OH-OO; it’s saying to figure
out the recipe for OH using Wide, Narrow, Open, and Closed.
When we identify visemes, we really are ignoring the open-mouth portion of open-
mouth sounds. After we finish quickly keying and identifying the visemes, we go back to
the start and add in the jaw motions. By treating these separately, we can move through
animations very quickly. If your only goal is visemes, you can burn through a long ani-
mation extremely quickly. It doesn’t look like much at this point, but you are left with a
simple version of the lip sync that you can then build on simply by going back and identi-
fying where the jaw must be open.
This approach is much faster than meticulously trying to get every sound right as you
move through your animation one frame at a time. This way, you end up at a jumping-off
point for finessing very quickly. The time you spend animating sync and expression will
be more heavily weighted toward the quality.
Disclaimer: The choices of what is and is not important are based on my own experience.
This is not torn from another book, university study, website, or anything else. The way I
break down words isn’t even a real phonetic representation; words are presented this way
here because if you’re like me, those phonetic alphabet symbols with joined letters and little
lines and marks all over them in dictionaries don’t mean much.
SouNd example SouNdS
UH fun, some, thunder
AH blast, bat, Vancouver
OH snow, foe
AW oxford, golly, lawn
Table 1.2
Example open
mouth sounds
609903c01.indd 10 6/2/11 1:52:17 AM