American Cinematographer - February 1993 - 68

intercom, and the pair give

the

tion control in order to create a new

of the actors, the actors follow the

model another minute nudge.
In addition to being

type of movement that allows him programmed movements of the
the total freedom. Because the actors camera. Location sound is not re¬
director, producer and scriptwriter are moving on rotating turntables, corded, so the actors lip-sync their
of Kafka, Zbig is also the director of
they are not bound by the physical pre-recorded dialogue. These dia¬
photography; he heads out onto the limitations of the studio and can go logue tracks also help to determine
set momentarily to adjust a kicker
anywhere. The same is true of the how long a move should last.
camera. In one scene, in fact, the
for the model. Eventually, the
This type of production
creates unusual
group is ready
Like pieces of a puzzle, the actors and the set's
to videotape the
demands on the
cathedral. The
turntable begins
to move as the

floors, ceilings and walls are photographed

separately and then matted together, creating
composite images made of up to 70 layers.

Lucas

Peter

explains,

"You have to

high rig, tracks toward the model.
The computer controlling the cam¬
era
rig is operated from the control
room
by motion-control engineer

camera

miniature, in scales ranging

used to
knowing where
to look [when] there's supposed to
be another character or a piece of
furniture. You're often acting while
listening to the director tell you

Paul Bachmann.
Weeks later, the scene is

6:1 to

where to look and when."

camera,

on a

mount¬
15-foot-

travels a full three-quarters
of a mile. To ensure the fullest ma¬

nipulation, the sets

all built in
from
128:1. Upon looking at a fin¬
are

completed as the actors place them¬

ished scene, one would never guess
that the window Kafka peers

selves

through is several times smaller

on a

turntable 24 feet in di¬

ameter and with

sion

move

painstaking preci¬

from mark A to mark B,

the giant turntable slowly mim¬
ics the movement of its four-foot

counterpart. After a take, the play¬
ers

rush into the control

room

watch the completed scene. Unfor¬

tunately, Lucas's Kafka walked just
a bit too slow and the frame
speeds
past him. "One more time," Zbig
intones for the umpteenth time.
The actors return to the

stage and attempt another take.
Throughout this try, Zbig gives di¬
over the intercom as he
watches monitors in the control
room. One monitor shows the black
silhouettes of the actors against a

than he is.

green background, another shows
them against the blue Ultimatte,
and yet another presents the

composited image of the actors in
the set. Finally, the take is a success
and the actors and crew break for a
late supper. There is still one more
element to layer into the scene-an
actor in the cathedral's pulpit-but
that will be added the next day.
Like pieces of a puzzle, the actors
and the set's floors, ceilings and
walls are photographed separately
and then matted together, creating

get

Birgit Bofarull agrees. "It's
difficult at first to act to playback
since you have nothing to 'play
with.' You must do your home¬

While the actors move
about in their make-believe envi¬

work, and it forces you to draw

ronment, the camera

get a sense of the rhythm of your

glides effort¬
lessly through the sets. This move¬
ment, like the seamless interaction
of the real people and the tiny sets,
is

essentially achieved by precisely

replicating the movements of the
turntables and the camera rig. The
rotation of the turntables is

cal element in

criti¬

creating Zbig's idea

of three-dimensional motion simu¬
lation.

rection

actors.

Zbig refers to the tech¬

from

yourself. After

while,

you

own voice and a sense of
spatial
relations. You start seeing the sets

in your

mind. It's good exercise. If

you can

do this,

you can

any¬

thing."
Although Zbig's multigenerational style is unusual, it isn't
ostentatious or overly obvious.
Many of his camera "moves" are
not going to cause the average
viewer to ponder how he did it. As

dolly movement. The high-defini-

his editor, John O'Connor, com¬
ments, "It's not an effect for an
effect's sake, it's an effect that

tion video

makes the scene. There's one scene,

nique
cause

as "motion simulation" be¬
there is only limited camera
camera

is mounted

on a

rig 15 feet high, allowing vertical

movement and movement perpen¬
dicular to the turntables on a 70-

furniture, in which the

foot-long track. The camera dolly
cannot track left or right. Instead,
the camera pans left to right, moves
forward or away and cranes up or
down (while the lens zooms in or
out) as the actors and sets rotate on
their respective turntables. The ro¬
tation of the turntables, coupled
with the movement of the camera

dolly, is what gives the illusion of

long shot of a room crowded with
camera ro¬

through the room. There's no
way you can shoot something like
that, even if you have a Steadicam
and a track going over everything
in the room; the way we do it, the
walls just aren't there. [Using tradi¬
tional filming methods,] you'd run
up against the wall."
tates

It took several months just
to build the motion simulation sys¬
tem. Paul Bachmann explains how

composite images made of up to 70
layers.
In photographing his sets

fluid motion.
In many ways, Zbig's mo¬
tion simulation reverses the usual

and actors separately, Zbig has suc¬

order of filmmaking.

trol) mainframe computer which
controls nine channels, each of

cessfully managed to modify

camera

which is connected to

American

Cinematographer

mo¬

Instead of the
following the movements

February 1993

the system works: "We have an
IMC 3565 (interactive motion-con¬

motor. In

American Cinematographer - February 1993

Table of Contents for the Digital Edition of American Cinematographer - February 1993