Shooting Lincoln

MAKING IT NEW

Everyone knows President Abraham Lincoln was assassinated - that fact is beyond dispute. So how does one go about putting a new spin on a familiar incident, make it emotionally involving and as stylistically consistent as the rest of "Abraham and Mary Lincoln: A House Divided," a six hour PBS/American Experience documentary? Director David Grubin, senior producer Allyson Luchak, field producer Amanda Pollak, associate producer Sarah Colt and I discussed various approaches to this and other sequences. A principal, self imposed limitation throughout the production which applied equally to the assassination sequence, was that actors would not portray historically identifiable people - thus ruling out impersonations of Abraham Lincoln, Mary Lincoln and assassin John Wilkes Booth.

David Grubin believed that after watching this documentary the audience's emotional identification with the 16th President would be significant. The director's intent was to prolong the dramatic moment as a means of satisfying the viewer's desire that the murder not occur. We decided on ultra high speed photography and to see the bullet that ended Lincoln's life.

At first, I considered the 16mm NAC camera from Photosonics that can shoot at frame rates up to 10,000 fps. Since we are shooting Super 16 for finishing on HD (1080i) I had concern about grain. Using the NAC, we would be required to shoot with Kodak Vision 500T (7279). The inevitable blowup required to make the 4:3 format 16mm negative conform to the 16:9 HD format would result in a significant increase in grain that would make seamless integration with the rest of the sequence impossible. I sought other alternatives.

THE MILLISECOND CAMERA

I had read an article in this magazine (ICG magazine) about a new prototype 35mm camera called the Millisecond Camera designed by Nathan Nebeker which could shoot at 12,000 fps. In the Millisecond camera, film is wound around a drum that produces a shot 120 frames long (five seconds of screen time). Prior to the event, the drum is accelerated up to speed, thus saving a significant amount of film usually wasted as the camera ramps up. The savings that result from less stock, developing and transferring is very considerable.

With the help of International Photographer magazine staff, I contacted Nathan Nebeker and we contracted him to film a working reproduction of the assassin's derringer. We sent Nathan storyboards for five camera set-ups. Working with the limitation of the drum camera's 120 frame shot duration we wanted to designed the sequence to be around 25 seconds. When we saw the resultant shots, events such as the hammer striking took only 2 seconds of screen time and the bullet itself was on the screen for only 15 frames even at 12,000 frames per second. We used some of these shots, but David felt we needed more screen time on the bullet alone and even requested up to 24 seconds of screen time to cover voice-over he wanted to try at that moment. Nathan Nebeker suggested dropping the bullet rather than firing it from a gun. But due to the 120 frame shot length limitation of the Millisecond camera we opted to shoot it ourselves.

DEFINING THE EVENT DURATION

For the bullet to appear on the screen for 24 seconds when projected or transferred at 24 frames per second, it will be on the screen for 576 frames. To determined the frame rate of the camera , we divide the number of frames - 576 frames - by the event duration (.25 sec, for example) to get the result 2305 fps.

(Frame Rate = Screen Time*Transfer Rate/Event Duration = 24 fps*24 sec/.25 sec=2304 fps)

A longer event duration would result in the bullet falling a greater distance. A shorter event duration would would require a higher frame rate. To get a shot 24 seconds long, the Photosonics 4C running at its maximum speed of 2500 fps at 208v would have a maximum event duration of .23 sec.

MAKING THE FACTS WORK

Lincoln was shot from a distance of about six inches with the lead slug projectile was about 0.375" in diameter. Since this is a public television historical biography (used for information as well as entertainment), we always stick to the known facts. However, to have the bullet travel only six inches would require the camera to run at a frame rate in excess of 4600 fps at the very least. Since that is not possible with the Photosonics 4C camera, I could have the bullet travelling further - and being larger or overscale- to keep the factual relationship of bullet size and distance it travelled.

LAW OF GRAVITY

An overscale bullet falls at the same rate as a smaller bullet. Recall the experiments of Galileo (the late 1500s Italian astronomer/physicist) on the Leaning Tower of Pisa with falling balls of different sizes landing at the same time. Using an old physics textbook from my school days, I found the formula to determine the rate of acceleration due to gravity for falling objects. The admonition, "Don't throw away those old text books they may come in handy someday" finally proved itself true for me. The formula is distance fallen (Y) = 1/2 acceleration due to gravity(g) x time [or event duration] (t)2 or Y = 1/2 g t2 where g = 32ft/sec2 or 9.8m/sec2 at sea level. In order to get a better sense of the distance the bullet would have to travel and the event duration, I made a chart to plot the distance fallen to the thousandth of a second with the help of a spreadsheet program. One column charted increasing event durations (time) in .025 second intervals while the adjacent column displayed the distance fallen from the rest position in feet, inches, and meters. Inserting the maximum event duration for the Photosonics 4C camera, discussed earlier, the fall had to be at least one foot if it started at the rest position.

(Distance traveled Y =1/2 gt2 = 1/2 (32ft/sec2)*25sec2=1 ft.)

Event Duration (seconds)	Distance (feet & inches)	Distance (meters)
0.125	3"	0.076562
0.25	1'	0.30625
0.5	4'	1.225
.75	9'	2.75625

SIZE MATTERS

By making the slug overscale to three inches rather than actual .375" (some 8 times larger) the traveled distance could be eight times more that the actual six inches traveled by the real bullet, or up to four feet and still be in a ratio to the actual event.

Since the bullet would be dropped rather than fired from a gun, in order to maintain the correct orientation and screen direction, we had our key grip Neil O'Malley mount the camera on its side so it was ninety degrees off the vertical axis.

DROP IT!

I contacted George Apostol, a New York prop master and rigger, to design and build a bullet drop rig. For more visual interest, I asked George to spin or roll the bullet across the screen, and to add smoke that appeared to have come from a gun muzzle. George's solution was as elegant as it was versatile and, as always, exceeded our expectations. The bullet was mounted in a plexiglass tube resting on a nonskid bar triggered by a pneumatic cylinder which snapped open on cue - this allowed the bullet to drop while giving it an initial spin. For the gun muzzle smoke effect, George filled the plexiglass tube with smoke from a smoke machine prior to the shot. At the same moment as the bullet was released, a capping shutter at the opposite end of the tube opened to allow a pump to blow the smoke out. The drop rig and capped pump were cued from separate switches so they could be operated with great variety and subtlety.

We sent George an actual-size lead slug and he fabricated an oversized projectile three inches in diameter. He added facets on the slug to catch the light and subtle rifling and other effects of firing the bullet. The projectile model was made from a rubber ball covered in plaster and painted silver.

Trying to anticipate any possible problems, I called Conrad Kiel of Photosonics. We reviewed the shot's specifications, camera requirements, film-saving techniques (such as splitting camera rolls down to 500' loads and letting them roll through rather than ramp down), and lighting problems at this film speed. I'd liked to thank Conrad again for sharing his expertise.

EXPOSURE REQUIREMENTS

Twenty five hundred frames per second is 104.1 times normal speed with an exposure time of 1/12,500 of a second. The stop compensation with the beamsplitter and the 72 degree shutter on the 35mm Photosonics 4C camera at 2500fps is 8.5 stops plus a half stop for reciprocity failure, or 9 stops total which requires 512 times more light than normal. Shooting Kodak's Vision 500T (5279), we need 80 footcandles (fc) to get a T5.6 @ 24fps. To get the same stop @2500fps, we need 40,960. fc. We needed this intensity on both the bullet and the red fabric panels which simulated the walls of the Presidential Box at Ford's Theater in Washington DC.

WEIGHTING THE LIGHTING OPTIONS

We were not doing a test shot and a reshoot was not an option. So after reviewing the pitfalls of all the possible light sources, I opted for going with tungsten as the safest and least expensive. But a quick review of some of the problems of various sources might be helpful. Although I wanted to use arcs, the studio available was not properly ventilated to allow it. At this speed, Xenons sometimes display a problem known as - "filament wander" - this phenomenon introduces a slight movement into the shadows and give the possible appearance of a flicker. If the square wave decay falls below 6% of peak intensity, flicker-free HMI's have a potential flicker problem at this speed. Tungsten lights of 1k or less can have a flicker problem due to light decay on the filament from AC phasing. I felt that this problem was the most easily resolved by separating the units onto different phases of the AC power supply.

A 1k PAR (parabolic reflector) with a Very Narrow Spot (VNS) bulb @ ten feet provides 5040 footcandles with a beam diameter of two feet wide by one foot high. Therefore, it would require at least 8 1k PAR with VNS bulbs @ ten feet to get a T5.6 @2500fps on just half of the bullet's path.

From the camera's perspective, we used two 9 light maxi brutes (9 light PAR) with very narrow spot bulbs (VNS) very close to the background from the bottom with a metal meataxe (24" x 72") serving as a top chop. Four 2K open-face tungsten lights were used to fill in the fall off between the rows of lights on maxibrutes. We also had metal flags used to keep the spill off the bullet. This gave us well over the 40,000 foot candles. The background was about ten feet behind the drop rig to allow for working space and soft focus of the fabric's pattern.

At the suggestion of gaffer Glenn Miller, we lit the bullet's path with two Ruby 7's with Very Narrow Spot (VNS) bulbs, which had become commercially available in New York City only that very week. The Ruby 7's were focused to a spot and we added a couple extra single PAR's with VNS bulbs and hampshire frost to smooth out the bullet's path. They were positioned as 3/4 back lights from the camera right side and above the bullet's path so that the specular highlight from the Ruby 7's off the bullet were about 45 degrees off the horizontal. Two 4'x8' foam core were used at the top and bottom of the frame as bounce fill getting their light by kicking back intensity from the Ruby 7's and giving us a contrast ratio of 8:1. The center section of the bullet was allowed to go dark. A Teaser was used on the top card to cut spill off the upper portion of the graduated fabric background. This was to keep contrast between the highlight off the bullet and the darkened wall. Of course dimmers and heat shield was used all around to protect the drop rig, the fabric background, and the crew from the intense heat.

It should be noted that when using the free falling projectile technique, the bullet is constantly accelerating from the moment it enters the frame to when it exits the frame due to the effect of gravity, thus is will exit the frame faster than it enters. In our shot, the perception of this gravitational effect is greatly minimized by the long screen time and is not apparent to the viewer.

The resultant image has such a lyrical quality that all the voice overs that were intended to accompany it were removed and it plays with only sound effects. When viewers see the final film, we hope we've given them time to reflect, feel, and understand why this tragic instantaneous act had such a profound effect on the nation and Lincoln's family. This shot is only the climactic moment of a six hour biography of Abraham and Mary Lincoln. There are many other events in the story that we are sure will surprise our viewers and that provided us challenging opportunities.

James Callanan
August 5, 1999
New York City