Morning talk with
Patrick Lin about his choice of camera lens, angles, movement, etc in Disney/Pixar's Up.
The movie is a dichotomy of static and dynamic, in his eyes, defined by a square and a circle respectively. (Square house, round balloons...) Using an outline of the script, he plotted on a scale of 0 to 10, static to dynamic, how each scene should be. The actual value wasn't as important as its relationship to the scene that preceded it. Static scenes were things like after the death of the wife until he launches the balloons. or after the kid is kidnapped. Most dynamic is the big rescue.
Then he explained how different camera variables were static or dynamic, and how using those in relation to the point on the graph was a good indicator.
StaticDynamiclong lens/telephoto w/narrow depth of fieldshort/wide angle lens with wide depth of fieldno barrel distortionexaggerated barrel distortionlittle variation between shots and compositiona lot of variation between shots and composition from shot to shotLocked down camerahand-held cameradolly track or boom movementpan and tiltpush shotszoom shotscamera movement on a 2d axiscamera movement through 3d spaceframing on a down angleframing on an up anglesqueezing, frame in frame, or frame divisionopen frameslevel shotsdutch anglesstaging and blocking in flat spacestaging and blocking in deep spacesingle shotover the shoulder or 2 shotBad guys were always introduced from the left of screen, good guys from the right. All positive movement in the film, including the trip toward the falls, is to the up and right.
He cited
Yasujiro Ozu as someone who used discipline to shoot very static movies... always very flat and level with a 15mm lens, locked camera. When the camera had to move, a person stayed centered in the shot and the background moved.
He also cited Akira Kurosawa's
Ikiru in which almost all shots of one character who feels isolated from the rest of the world because he is dying of stomach cancer always has some division in the frame between him and other characters... a door frame, a beam holding the roof, or some other vertical or diagonal line keeping him apart. This was used in some scenes in Up to show Carl's isolation.
Randy Thom's keynote was about Designing a Movie for Sound, and gave examples from Apocalypse Now and Wall-E. Sound design should be considered and experimented with early in the development process so sound & visual exploration can play together. He noted that a character who is an observer is also a listener, so the audience should hear what they hear. Almost all sounds in sound design come from recordings of found noises that are then modified. Entirely synthesized sound in live action film is still rare. There's more chance for play between sound design and animation than live action, but there's no reason for that to be the case. And lastly that one should not try to mix music, sound and dialog all happening at the same time. Let each take their place in turn.
The more interesting Visual Music talks of the afternoon were by
Fred Callopy,
Bret Battey and
Brian Evans. Assorted notes:
- Most color models use a 3 coordinate system, RGB, HSY, etc.
- Movement is 5d: scale (x,y), translation (x,y) and rotation.
- That gives us 12-15 dimensions in visual music. Instruments become the limiting factor.
- Don't map notes to hues.
- In form and movement, measure weight & quality (Klee)
- Presets are bad, because you can rarely move from one to another smoothly
- Color is good for fades allowing a continuity of motion
- Reset and Clear functions are a must in any software controller
- Look up Willard Huntington Wright's The Future of Painting
- Interface defines the character of the instrument, and also the music
- Embed a piece of the existing culture into your new interface to make it familiar to people/users. Dance?
- Departures from the theoretical makes it interesting.
- "It's not more options you want, it's better options." - Eno
- Realism is like lyrics, a strong tie to the audience to pull them through a narrative. Abstract can be less narrative and purely emotive. People will always interpret.
- Continuous unfolding (no edits)
- be a sculptor in time. dynamic shaping.
- Visual music is the unfolding of tensions in time for dramatic effect
- Music concrete theory: music of sound, not notes. Pitch and rhythm get obliterated but a timeline moves on.
- Compositional gestures through time evoke response
- The similarity of structure in disparate media is appealing.
- As a performer, the spontaneity of experience is immediate, not analytical
- There is a structural kinship between sound and image - the trick is in the mapping between the two
- The struggle of an artist is in knowing what to throw out.
- In Kandinsky's book, he quotes Ravel as saying "The laws of harmony are the same for painting and music."
- Art for art's sake (non representational) is fairly new. Sound has always been abstract. Striking a piano key means nothing beyond itself.
- The foundations of a musical pattern are repetition (aa) (provides structure), contrast (ab) (provides interest), and variation (aa') (something repeats while something else contrasts)
- Any expression is an utterance in many dimensions.
- Any of these dimensions can remain constant or change.
- Building blocks of music include motif, phrase, cadence, and tension/resolution (desire) which is a kind of narrative
- Visual music building blocks are formal, abstract and temporal.
- The elements of cinematic space are rectangular form, moving eye, and montage.
- There is such thing as visual tension, not unlike leading notes in music. "Wrong" visual experiences can be resolved to "right"
- Digital is discrete and numerical.
- If A=B and C=B then A=C. If a digital image is numbers, and a digital sound is numbers, then any digital image is a digital sound and vice versa.
- Maps are metaphors
- Consider the roles of diegetic and non-diegetic sound in visual music
- Brian uses Metasynth to convert still "scores" into music
We'll see if tomorrow is as informative.