Monday, January 19, 2009

You can't grab my shirt!

It's easy to blow up a city in 3D, but it's hard for a character to grab another character's shirt.
Brad Bird, director of The Incredibles and Ratatouille
In the January version of Imagine magazine, there's a great article by Saint John Walker of FDMX which touches on one of the biggest issues facing machinima. There are some things that you just can't do well using low-end real-time animation and a home computer, no matter how hard you try. And, frustratingly, some of these are the sort of things that are trivial with a camera and actors.

Going right back to the dawn of film, one of the first Lumiere shorts, L'Arroseur Arrose (Lumiere number 99, often regarded as the first comedy, made way back in 1895), features a gardener watering the garden. A boy enters the shot, steps on the hose and the water stops. As the bemused gardener looks into the hose to see what's wrong, the boy takes his foot off, and the gardener gets a jet of water in his face as the boy sniggers. (Ho, ho ho!) Then the gardener, now annoyed, grabs the young scamp and pulls him around, then squirts the water at him. All very simple, and it was made with the most rudimentary equipment.

Now try doing that in machinima. Getting the facial expressions is hard enough, but we're just about cracking that now. Getting the water to look right is just about do-able in modern game engines, but getting the splashing isn't there yet. But pulling the shirt is just not possible in machinima yet. It's right on the edge of what top-end 3D animation can do. More from Brad Bird:
One character touching another character's hair? "Aaah! No! Isn't there anything else you could do?" I mean, I had to budget shirt-grabs.
And if it's hard for Pixar to do shirt grabs, then, realistically, can you expect machinima to do it well? Nope. You can get a hand pretty near a shirt, but don't expect to see cloth being pulled about. And when you run your hand through someone's hair, expect to see hair poking through the hand and not moving right. Taking clothes on and off is a 3D animator's nightmare. You just don't do it, not unless you have Pixar's budget and a lot of patience.

There are, of course, the three usual film-maker's solutions:
  1. Write a story that doesn't need shirt grabs, hair ruffles, or the like.
  2. Shoot it off-screen and edit the sequence so that the audience thinks they've seen something they haven't.
  3. Shoot something close enough, don't worry about the visual glitches, and assume your audience will forgive you because it's machinima.
This isn't to say that machinima is crap because it can't do something that simple. Machinima is still a very powerful tool which enables to you to make all sorts of movies. And, as it says in the opening quote, some things are unbelievably easy in comparison to any other method of making films. What's important is that you have to accept the limitations of the medium, and learn to work within them. Machinima just can't do some of the things that other media can do, so live with it. If shirt-grabbing and hair-pulling are central to your film, and you absolutely need to show them on screen, looking realistic, then don't expect machinima to work for you. Use a real camera, or learn to animate and do it all by hand in 2D.

The original Lumiere cine camera can still create images that 21st century computers find difficult.

But, as Mike Joyce is so fond of pointing out every time we speak, limitations are what provide artistic challenges.

Actually - there's a challenge. How close can you get to that movie in machinima?


snorkel said...

It took about 4 billion years of evolution to produce hands that can grab shirts; not sure if we have that amount of dev time? :D

But yes, I tend to agree with Mr Joyce: limitation spawns creativity. Cinema itself pushes its own limits, things that can't be done (and bizarrely it often looks to CG to solve them!).

I guess I could come up with something trite such as "it's the story you want to tell that matters". And there is some truth in there; visual glitches are a secondary concern. But how are we to communicate (say) tenderness and subtlety? And is not style - correctly used - part of the story? In Matt Madden's book "99 ways to tell a story" we see the importance of style, and hence detail, on the same plot (a guy who forgets why he's looking in his fridge). On the other hand, if one avenue is blocked, there are 98 other ways to go! So: detail can matter. Style can matter. Glitches can matter. It all depends.

Example (thinking of Pixar): Wall-E. Exquisite CG for 30 mins, all sepia-HDR, combined with emotionally captivating content. Yeah so they use the "big eyes=cuteness" response that's part of our genome. There's still more to the robot than that. But then, the remaining hour is largely disappointing. Visually ok, but the story loses poignancy, and it's that which tarnishes the film. And it wouldn't have mattered how much tech they'd thrown at it; a weak story is a weak story. Moral: to reap the benefits of subtlety and detail, you have to have a good story to start with.

One can look at it top-down or bottom-up. In the latter, we look at what we can currently do and figure out what sort of a story we can make with it. In the former, we start with an idea and figure out how to make it happen. Realistically, we usually do both, and no matter what we do, we always make compromises. As they say, "perfection is the enemy of the good".

Dulci said...

I think we do have to understand the limitations of machinima when crafting our movies and base our scripting decisions with those limitations soundly in mind.

We need to decide how much time we're willing to invest in each portion of a project. Some people have the patience to spend one weeks on one 10-second moment while others won't spend the same amount of time on a 10 minute project. Obviously, the detail-oriented individual is the one who will be able to tease the software and achieve the results they are looking for, whereas the shrug of the shoulders, hey, I'm just trying to have some fun here approach will only draw attention to the limitations if they try to find a way around them.

Anyway! Where there's a will there's a way, right?

Kate Fosk and Michael R. Joyce said...

I am also certain that Machinima is another method for telling a story or conveying an idea. To limit Machinima to only a tool that can be used to "duplicate" a film is an unfair use of the medium. I like to think that these Machinima tools can be a means to create a visual product. We need to look beyond just mirroring live action film.

Matt Kelland said...

Kate - spot on. It's easy to think of machinima as a cheap way to make the sort of thing you'd do in live action if only you had the budget. That's largely, I think, because machinima is mostly people-based (like live-action) and does the expensive shots (i.e. the big action set-pieces) with relative ease.

But when it comes down to it, they're not the same medium at all, just as photography and painting (or digital art) are different, even though the end result is a picture in both cases.

sisch said...

Well, Animé (Manga) is people driven, too, most of the time, and yet no one would accuse the artists of drawing images because they can't afford a camera?
Okay, so the comparison might be a bit crude, I admit. But I really do hope that machinima will be one day fully accepted as an artform all by itself.

As for limitations - yes, you can't show certain things, but isn't finding a way to show actions implying these things - or even totally different images - just the essence of creativity?

We don't need to copy "real film". We make our own visions.