Original subtitles

[MUSIC PLAYING]

JAMES CAMERON: Generally speaking,

a lot of visual effects works against you

for a lot of reasons.

It can be numbing.

You can kind of debase the value of a given effect

by having too many of them.

They can blow your budget.

It's better to have a few memorable set pieces.

From a writing standpoint, I sort of treat it

as a standalone thing.

I'll figure out where it's going to start, what the kickoff is.

You know, what the kind of opening bell of the set piece

is and where the resolution is.

I'll figure out where my--

my out point is.

And then I'll work within that so that it's not too long,

it-- it stays focused and it stays of a piece.

And, I mean, I think there's--

just like there's an art to scene writing,

I think there's an art to creating a set piece.

It's kind of a film within a film,

and it kind of has its own-- own rules,

and it's got a beginning, a middle, and an end,

some kind of resolution.

That end may or may coincide with an act break

or it may not.

But, you know, a suspense sequence

or a scene, a dramatic scene in a film

might be a four or five-page dialogue

scene that's done in some stylized way,

maybe without any cuts.

Maybe some circulating Steadicam shot that moves sinuously

through some dramatic moment.

That would be a-- that would be a set piece.

So then you have to pick your battles carefully.

What's the narrative purpose of the set piece?

So to use the example from "Terminator"

of the first future flashback, it was actually

just to build character.

For Michael Biehn's character.

It was to show his traumatic world

and why he was basically a walking, physically-scarred,

and mentally-scarred example of post-traumatic stress syndrome.

He was the living embodiment in the present day

of how bad that future was and why you needed to prevent it

from occurring, right?

So it had a powerful narrative and character

purpose in that film, but it was also

just stuff I wanted to see.

And that's another thing.

Is there are a lot of kind of rules and advisories

about why you put things in movies, and--

and they all have to serve a purpose, except they don't.

Sometimes it should just be something

you want to see as a filmmaker.

And sometimes the only way to see it

is to make it and show it.

And I remember having this discussion

with one of the heads of 20th Century Fox

when "Avatar" was almost done and they were complaining

about the flying scenes.

Why does this flying scene go on so long?

And you've-- you've already proved the point.

It doesn't advance the narrative,

it doesn't advance the character.

I said, you're right on every count,

you've ticked every box like a good studio executive.

But guess what?

I want to see it.

And I extrapolate from that, and Stanley Kubrick

was a firm proponent of this principle.

If I want to see it, my cognitive leap

is, there are going to be lots of people that want to see it.

I want to see it for itself, not for its purpose, right?

Because there is a purpose to the things you want to see.

The purpose is to be present, to be in that world,

to be enjoying what's happening, to sustain a feeling.

So those flying scenes actually did have a big purpose.

That's when Jake and Neytiri fell in love.

And I got into a big fight with the studio over that.

And I'd like to say fortunately I prevailed, because I

love those flying scenes.

They're great.

But life is a whole lot easier if you can actually

dig deep and examine why you've made a decision

or why you have that instinct.

Because it's going to cost somebody else a bunch of money

to follow up on that instinct.

And so the more you can share it,

the more you can inspire others with it

and why that choice makes sense, the better life is going to be.

Set pieces often are good candidates for previs.

Chances are, you're not going to have a budget

to previs the entire picture, nor do you need to.

You should only have to previs the things that

are extraordinary and require really

tight, interdepartmental planning well before the fact.

So in terms of the previs process, you've got to--

you can't just start storyboarding

until you know what things are going

to look like, because what are you going to draw?

Is the character wearing a trench coat or a loincloth,

you know?

So you have to do some design before you can even

start storyboarding, otherwise you're

just going to do stick figures, which

isn't going to tell anybody anything, yourself included.

So I say start the design process,

start to evolve the storyboards.

Now some people work with storyboards

and some people work more with previs tools.

You can use a-- a suite of--

of keyframe previs.

You can do a little animation pieces.

Put a blank face, but at least it

gives you the-- the dynamics and the framing.

And you start to think about composition,

you start to think about cutting.

And now you'll-- you'll essentially do some kind

of a pre-cut of the scene with your-- with your previs.

It doesn't mean you're bound to shoot it that way.

You're always going to be open to inspiration of the moment,

but it's a good jumping-off place

so people know how much you have to build,

what does the director expect here in terms of the action?

How long am I going to be on something?

How much detail do we need?

What kind of action are we going to try to do/ How are we going

to move the camera?

How are we going to support the camera?

Is it on a crane, is it on a Steadicam,

is it being handheld?

You know, it starts to trigger a thousand questions.

And a good team around you will start to hit--

hit you with those questions, you know?

The DP's going to want to know, are we on a wide lens style

or are we on a longer lens style?

Are we on a backlight style or are

we on a diffused light style?

What's the style, you know?

Because there's a thousand different ways

to light a scene, there's a thousand different ways

to shoot a scene.

Presumably the filmmaker, if they don't come in with a style

initially, they're going to have to find

one that suits the piece, that suits the story.

That's-- that's a critical part of it.

And when those questions start to get answered,

then you can start saying, okay, I

need this shot, this shot, and this shot.

Just from my own experience, a good example of a set piece

that was done on a film that was made on a very low budget

is Reese's flashback of the future, of the traumatic future

that he comes from in the first act of "The Terminator."

And I say the first act because Reese hasn't met Sarah yet

and he hasn't met--

he hasn't met the Terminator yet.

So this is all sort of in the character setup,

and I wanted to see this future soldier in his world,

and I wanted to go into it through his fatigue-induced

dream state and come out of it where he's shocked

and disoriented and back in our present, which is his--

his past.

I knew it was going to involve visual effects models,

you know, miniatures, miniature explosions,

and some full-sized live action elements

that were done in a-- in a set of some-- some ruins,

some post-nuclear holocaust ruins.

So I thought, I'd a better draw this up.

So I approach it in a very disciplined fashion.

I got my roommate to put on some kind of military uniform.

I went and bought some guns.

We went-- we went out in a field someplace,

and I took pictures of him in all the different poses

that I wanted.

And I also did that for the car chases as well, which I also--

also storyboarded.

Had him, you know, hanging out the window,

I wanted to be able to see what kind of positions

you could get into with the weapons, what kind of lenses

I wanted to be on.

Did I want to be on telephoto lenses or close lenses?

Wanted to make the guns look big, so shorter--

shorter lenses.

And start to think, well, how am I going to shoot this?

How can I get the Insert car there?

So I started going through the whole--

whole exercise.

I got my long-suffering roommate to basically play

every character in the film except for Sarah.

I think he played Sarah a couple times.

And then I brought it all back.

I printed up all the photos.

This is back in the days of photochemistry.

So I had a bunch of prints all over my desk,

and I just started drawing it all up.

[MUSIC PLAYING]