High speed photography

There's no denying how awesome it is to see things in super slo-mo. Particularly when the subject of the photography is ultra-crazy fast, and the result has an unusual beauty about it. A fine example below.



So how do they do it? What kind of camera setup could produce one million crisp images a second?

Thanks to the miracle of modern technology - this particular footage was taken with a digital camera: the Shimadzu HPV-1. According to the site, it takes 312 x 260 resolution images up to 1 million FPS. That probably takes a lot of light. You'll notice that the camera is black and white - that makes sense, given that it's probably hard enough capturing that volume of luminance information such a small window of time. Color data probably requires too great of resources.

However, decades ago - before the advent of amazing cameras such as the Shimadzu HPV-1, you still saw some amazing high speed photography. Even at "bullet time" speeds. How did they do it?

Conventional film cameras use a shutter. It regulates the precise duration of light exposing a single frame of film. Without a shutter, ambient light would quickly over-expose any ordinary photographic film - leaving a featureless white image.

However rapid a shutter may be, it still is a mechanism that has an upper limit of how quickly it can operate. (Even a single time... thousands of times a second is another story.)

While I am sure there were many different variations (or entirely different approaches), the following method was used to get around the limitation of the shutter:

Don't use one.

That's right - eliminate that limitation. So what then? How do you keep the film from being overexposed?

Keep the scene in total darkness.

Sure, you'd keep the "lens cap" on (or whatever the equivalent is) on while setting up the shot. But when it got time to roll, lights out.

Then comes the matter of film. If a 35 mm frame was *very* roughly one inch tall (to simply this example), and film strips are just a sequence of these - shooting 50,000 frames in a second would require 50,000 inches of film in a second. That's 4167 feet of film. About 3/4ths of a mile - in a second. That's movin'.

So they add a lot of leader - get the two spools going damn fast, and when the actual film is unspooled and entering the brief window of time it could be under the aperture -

Then-

They hit the strobe light. That's right, rather than blocking the light except for brief moments - they supply the light at brief moments. That's the trick. The strobe must be very bright and precisely timed. These flashes produce the images that expose the insanely sensitive film as it whizzes by.

Of course- they also need to trigger whatever event they were planning to photograph, and hope it all goes off as planned in the extremely small window of time available. Otherwise, it's time to load up another mile or two of film and try again.

The downward spiral of comic book hero franchises

I have a simple theory about why comic book franchises tend to start strong, and then by film #3 it's more marketing and visuals than actual filmmaking.

Character bloat. It starts off as an interesting look at a hero and an arch-villain. Then, which each successive installation - more heroes and villains get added until it's a ridiculous ensemble.

Take the classic example of the Batman franchise.

#1 - Batman (1989). 71% on RT. Hero: Batman. Villain: The Joker (total: 2)
#2 - Batman Returns (1992). 77% on RT. Hero: Batman. Villain: The Penguin. Wild card: Catwoman (total: 3)
#3 - Batman Forever (1995). 44% on RT. Heroes: Batman and Robin. Villains: Riddler and Two-Face (total: 4)
#4 - Batman & Robin (1997) 12% on RT. Heroes: Batman, Robin, Batgirl. Villains: Mr. Freeze and Posion Ivy. (total:5)

-then the Nolan reboot-

#5 - Batman Begins (2005) 85% on RT. Heroes: Batman. Villain: Scarecrow. (total: 2)
#6 - The Dark Knight (2008) 94% on RT. Heroes: Batman. Villain: The Joker. Wild card: Two-face. (total: 3)

Let's take another example: Spider-Man.

#1 - Spider-Man (2002). 90% on RT. Hero: Spider-Man. Villain: Green Goblin (total: 2)
#2 - Spider-Man 2 (2004) 94% on RT. Hero: Spider-Man. Villain: Doc Ock. Wild card: Harry Osborn (total: 3 )
#3 - Spider-Man 3 (2007) 63% on RT. Hero: Spider-Man. Villains: Venom, Sandman. Wild card: Harry Osborn / New Goblin (total: 4)

Another example: Superman.

#1 - Superman: The Movie (1978). 93% on RT. Hero: Superman. Villain: Lex Luthor. (total: 2)
#2 - Superman II (1980). 87% on RT. Hero: Superman. Villains: The trio of Ursa, Non, Zod. (total: 2 - yes, I know there are three names but effectively they act as a single unit. They arrive together and leave together.)
#3 - Superman 3 (1983). 23% on RT. Hero: Superman. Villains: Superman(!), Ross Webster, Badass Computer Thing(?) Wild card: Gus (total: 4-5? This is a strange one due to the Superman duality and the computer-thing.)
#4 - Superman 4: The Quest For Peace (1987). 11% on RT. Hero: Superman. Villains: Lex Luthor, Nuclear Man. (total: 3). [But it must be said: This film was a piece of crap on so many levels it's hard to pin the problem on anything a subtle as the # of heroes and villains... ]


And recently, Iron Man:

#1 - Iron Man (2008). 93% on RT. Hero: Iron Man. Villain: Obadiah Stone (total: 2)
#2 - Iron Man 2 (2010) 75% on RT. Heroes: Iron Man, War Machine. Villain: Ivan Vanko (total: 3 - BUT, it's easily argued that Hammer as another Villain, or Black Widow as another hero. And then there is Nick Fury... )

So what might we take from this mini-analysis? Perhaps the sweet spot for comic book heroes is film # 2 - with a hero, a villain, and a wild card. More than that spreads the screen time to thin, and doesn't permit the title protagonist to develop a true relationship with a villain (that the audience can see is evil, and must be stopped). And it could be that a "wild card" character keeps things from being the usual white hat vs. black hat buildup & showdown.