From the July/December 1999 issue of the Operating Cameraman

In early July of 1998 I received a telephone call from Jacinda Davis of Knight Scenes, Inc of Washington DC, inquiring what it would take to engineer the installation of two 35mm motion picture camera systems at two different camera positions in the vicinity of the Washington Monument.

Furthermore, how could such an installation be monitored and facilitated over a time period of almost two years. Their purpose was to document, in a time lapse motion picture film, the restoration of the Washington Monument in Washington DC.

The footage is to be used in the Washington Monument documentary It Stands For All, for the Discovery Channel. This documentary is the project of Knight Scenes, Inc in Washington DC, which is owned by Emmy Award nominee Producer/Director Brian Leonard, who conceived the idea of the time lapse sequence.

When I first heard about this idea, I had many questions requiring answers and clarification. "There are so many variables to consider" kept going through my mind at all hours. Finally I spoke to Director Brian Leonard, who informed me there would be a three to five month long scaffolding process; followed by a one year period during which the monument and scaffolding would be wrapped up in some kind of slightly translucent material, much like a cocoon, during interior and exterior restoration; followed by a three to five month de-scaffolding process; and finally the unveiling and revealing of the newly restored monument no later than July 4, 2000.

I was offered a variety of camera positions at all different angles and elevations, some on rooftops, some in towers; even the roof of the Lincoln Memorial was considered at one point. All these locations were discovered and scouted by Director Brian Leonard and his assistant, Jacinda Davis. They sent me a complete series of photographs taken from the various prospective camera positions. Many, many questions needed to be answered.

Questions like:

1. Where is the sun rising and setting in relation to the monument?
2. Where will the sun's exposures be drifting to over time?
3. During what hours of the day will work be taking place at the monument?
4. Where will the sun be during those critical work hours? How early will is rise? How early will it cross into the western sky? What is the earliest it will set?
5. How much on-screen time will need to be acquired each day in order to show progress at a good rate for screen presentation?
6. Should the recorded progress be incremental or gradual? In essence, should the camera acquire a lot of individual frames over a short period each work day, for instance between 10 AM and noon, or should the camera acquire fewer frames per hour spread out over the entire work day, providing a more gradual progression?
7. How will the cameras adjust for varying illumination and changes in weather conditions?
8. What are the maximum and minimum temperatures to which the cameras' environments could be subjected?
9. What are the humidity variables? Will the camera's booth, in which it has been installed, steam up and fog the optical glass through which the camera is filming? Will water droplets form on the outside of the optical glass due to some extreme temperature differential? Do I need to consult a thermal physicist?
10. How will I deal with the possibilities of rain, snow, smog, wind, dust, frost, insects and even bird droppings landing on my optical glass, and thus contaminating my acquired images?
11. How will the cameras be secure from theft or vandalism 24 hours a day for two years if placed in a compromising area of the park surrounding the Washington Monument?
12. How will I protect the cameras from power failures, electrical surges and brownouts?
13. How will I monitor and service the camera installations without having to sleep next to them for almost two years?

RESEARCH

The first thing I did was to make contact with a friend of mine, Dr. Ivan Linscott, who is a researcher in the field of astronomy, located at the Stanford University Star Lab. This was in order to understand and determine where and when the sun will be coming and going for the next two years. This alone was a very interesting series of lessons.

I was informed by Director Brian Leonard that the work would be taking place at the monument from approximately 7 AM to 3 PM, five days a week. This helped me to determine that the camera positions should be facing from east or southeast toward the west in order to gain early morning direct sunlight illumination.

So the Lincoln Memorial, which faces the Washington Monument from the west, was ruled out as a possibility, because it would at best give only three direct hours of sunlight per day, overlapping with the daily work times at the monument.

It was decided that one camera would be in a tower which faces the monument from the southeast at a height of seven stories; however, the tower camera position we selected has a three-story metal spiral staircase to the tower landing from the fourth floor of a government building. The stairs are about 18 inches wide, and there was severe doubt about just how and if we could fit all of this heavy equipment and camera mounting platforms up this spiral staircase without having to cut things down to fit.

The other camera would be located in a specially constructed camera booth measuring approximately 6 feet by 6 feet by 6 feet, built on the top of the Washington Monument Lodge snack bar at a vertical height of only 10 feet above the ground, and about even with the base of the Washington Monument, which is located about 1,000 feet due east, facing west, toward the monument. This is a particularly low angle wide shot looking up and to the west at a tilt of about 65 degrees. This would require a very wide-angle lens.

Furthermore, being that I use Mitchell film cameras, I had extreme considerations about whether I could get the camera's film magazine, located atop the camera, to clear and not hit the interior of the wall of the booth in order to get the lens close enough to the optical glass in order to squeeze the images' field of view into the frame without any side or top intrusions creeping into my frame.

How large an optical glass should I be filming through not to have any frame interference? Tests revealed that a 12mm lens would be just wide enough to squeeze the Monument vertically into a 1.85 to 1 frame aspect ratio. This would be optimum for later digital video 16 x 9 presentations.

Another similar situation existed at the other camera position in the tower. With the added glitch that the only windows in the tower were facing north-south and east-west, but the monument was exactly northwest, the windows needed to be first removed and then rotated by 45 degrees.

I drew plans for what I called "window angle correction assembly," a plan whereby existing windows are removed and stored, and are then replaced with a combination of rigid weatherproof materials and an optical quality glass frame assembly through which the cameras will view the subject, in this case the Monument. This assembly would have to be built in advance and would need to be totally weatherproof and be designed to repel rain and snow.

PREPARATION

I finally got a chance to actually use the American Cinematographers Manual to work out many details of this project. One of which was to determine exactly what my field of view was with various focal length lenses. As it turned out, a 20mm was perfect for the tower camera position. However, knowing that I could really only effectively get the lens' front element within, let's say, a few inches of the optical glass, led me to order (3) 11" x 14" glass filters as my optical flats.

These glass filters were actually neutral density 1.2 absorptive glass filters. The reasons why I used neutral density filters of this particular value, instead of clear optical glass, will be explained later in this article.

With the 11" x 14" openings I knew it could accommodate a 12mm lens, getting the entire image field of view into the opening without needing to place the lens right up to the optical glass. The third optical glass (neutral density 1.2 filter) was to be kept as a spare glass in case one of the camera's windows somehow became damaged or broken.

The next thing I needed to know was exactly what is the necessary tilt factor or vertical angle to the optical center of the Monument.

David Insley SOC, a Washington DC/Baltimore based director of photography was brought into the project at the request of the director, Brian Leonard. David was brought in to supervise the advance construction of the two camera environments under my direction, and eventually perform weekly reconnaissance and servicing the camera installations, reporting directly to me.

David brought in the project engineer, Lee Carrick, a Washington DC-based machinist electrician and carpenter. Lee Carrick and David Insley took measurements with a surveyor's transit and determined the vertical tilt factor to optical center of the Monument.

With the help of Lee Carrick's assistant, Rob Coughlin, they built the window angle correction assembly leaving me an 11 1/2" x 14 1/2" opening in which to retrofit into place, the custom-built optical glass holders or frames. These were made of 3/16" steel and were pre-tilted and slotted to accommodate the vertical angle needed to be exactly perpendicular to the light path or parallel to the film plane.

This made it possible to vertically slide the 11" x 14" optical glass into and out of the frame for regular cleanings and possible replacement, if required. The tower shot required the frames' filter slots tilted at approximately 80 degrees, and the booth shot required the frames' filter slots tilted at approximately 65 degrees.

Both frames were built in advance according to my plans by Jim Hole, a highly skilled prototype engineer of J & L Machining in Simi Valley, California, a usual stop for me prior to most time lapse experiments and the new time lapse endeavors I undertake. All factors involved in this rushed preparation had to be anticipated and remedied at once.

TARGET: WASHINGTON MONUMENT

Once all preparations were completed, I set out on a cross-country drive in my special camera vehicle named "The Dozer" (an immaculate 1985 Mercury Grand Marquis two-tone luxury camera vehicle with a trunk packed like a Rubick's Cube), and once again, as on many past projects, I brought along my trusted assistant cameraman Curt Walheim.

The trip took six days from Santa Barbara, California to Washington DC where we arrived on August 7, 1998, ready for action, with cameras on board, all other required apparatus in transit, on its way to Washington DC.

EXPOSURE CHANGES

One of the most important factors was how to adjust the exposure for changing illumination conditions.

Again the Norris LPC, or Light Priority Control, proved to be the pivotal answer. This device does not adjust the lens iris. It adjusts the motor rotation speed per frame, or what we call the camera's shutter speed. It does this rotation speed correction in 1/1000th-of-a-second increments, incorporating and communicating with a modified spot-meter which is aimed at an illumination reference target of choice. In this case, the actual Monument itself was that reference target.

As the light dims down, the Norris LPC slowly and inverse proportionally lengthens the time of exposure for each acquired film frame. As the light increases to the reference subject, it speeds up the rotation or shutter speed with exact and gradual proportion to the illumination of the selected reference target.

In this case, such a device is a lifesaver. I would go so far as to say the project quality hinged on this device.

An auto-iris would be vastly inferior for an application such as this. It would greatly affect the optical quality by varying the depth of field and would not have the vast amount of range that the Norris LPC enables the camera to have.

In this case, our exposures vary from 1/15th of a second (fastest shutter speed) to 60 seconds exposure time per frame (slowest required shutter speed at night), a variable equivalent to 10 f-stops of iris adjustment, or the equivalent range of T1.4 to T45; in other words, off the scale or down to a pinhole of light transmission.

FIELD OF VIEW

In advance, I also had to design exterior shadow boxes or matte boxes in order to block out extraneous sunlight or reflections of sunlight from hitting the gigantic 11" x 14" ND windows (optical glass/neutral density 1.2 filters), without cropping the frame in any way. Again, the American Cinematographers Manual was very useful.

By determining the angle of view of the given lenses, I drew plans to full scale and sent them to David Insley and Lee Carrick, who built these shadow boxes to exact specifications.

It was very scary knowing that once I arrived in Washington DC with the cameras, lenses and related apparatus, it would have to fit perfectly and function perfectly, without and gray areas of uncertainty. The matte tunnels or shadow boxes had to be optimum in effect, without encroaching on my frame. Furthermore, they had to repel rain and snow.

An additional aggravation was that an irremovable brick arch overhang threatened to encroach on the tower camera in the upper right corner of frame, as did a brick ledge about 3 feet out from the camera's window which threatened to encroach on the bottom of frame.

After drawing many sketches to scale, I determined that we could just squeeze the 20mm image in between the two brick frame invaders. It would not just be close, it would be exact, with no breathing room up or down, right or left, or back from the glass. As it turned out, the lens snooted right up to the glass from less than 2" back, and just cleared the possible frame obstructions in the lens' field of view.

In the booth camera position, with a 12m lens, our shadow box looked like some kind of art deco architectural statement, affixed to a rectangular booth. This was due to the extreme wide-angle lens with the camera tilted up at such a severe angle.

The result is that the shadow box is not actually a box, it is a shadow triangle, narrow at the bottom and flaring out severely as it gets to the top of the image viewing or transmission area outside the booth.

POWER AND TEMPERATURE

Next were the temperature factors, from freezing to 100°F. Needless to say, this is a wide range of thermal travel. After consulting an air conditioning specialist who thought he was a thermal physicist, we ordered special air conditioning units. These units were ordered and shipped to Washington DC from the Hawaiian Islands. (They also know a thing or two about humidity in Hawaii, "you betcha!")

These special 8,000 British Thermal Unit air conditioners were very small, portable, well-designed and extremely expensive, but they fit the bill, as they produce a very dry, very cold, air flow into the camera's environment, completely lacking any humidity or moisture.

Additionally, small heaters were installed for the eventuality of winter. Each camera position required the installation of two separate electrical circuits with back-up initial UPS power units, as well as special line conditioners to clean the electricity so as not to risk disturbing the sensitive electronic motors and intervalometers.

Additionally, a new Norris Film Products emergency power sensing switch was installed at each camera position. This switch senses when the camera's power supply voltage drops below 24 volts, and switches instantly and undetectably over to a standard 24 volt camera battery as the source of power for any period of time in which the camera's normal AC to DC power supply is failing to deliver a constant voltage that is greater than that of the back-up battery.

Thus, I set my power supply to deliver 28 volts to the camera, and if the power supply current drops below 24 volts, the battery is automatically connected and the power supply is disconnected for a period of time until its voltage level exceeds that of the camera battery, got it? Don't make me repeat that one.

This is all in addition to back-up in-line AC UPS power systems designed to last for many hours during a possible power failure. Furthermore, each camera position was equipped with a spare motor , spare intervalometer, spare power supply and spare cables.

One day about a month into the project, the power to our entire booth AC circuits was disconnected from the main breaker box at the Washington Monument Lodge by contractors remodeling the Lodge and Lodge restrooms. The UPS powered all the equipment for many hours and once the UPS-power storage cells were drained, the camera automatically switched to the 24 volt battery. The camera equipment kept working; however the air conditioner, lights, alarm system, video surveillance system, etc were all down for almost a day, but hey! The cameras just kept on ticking and no footage was lost.

REMOTE SYSTEM SURVEILLANCE

One way I was able to avoid sleeping with the cameras was installation of video surveillance cameras at both camera positions. These small, low resolution cameras were aimed and framed on the LED controlled read-outs on the Norris LPC intervalometer, the Norris motor, the camera's frame and footage counters and at a digital thermometer. The video camera's signal was routed into an encoder that converted it into an audio signal that can be carried by a standard telephone line connection.

Consequently, I can sit in my Santa Barbara home viewing the video surveillance cameras from a monitor that is connected to my decoder, which is being fed a signal via the telephone line. At any time I can simply dial-up the booth or tower camera positions and check the footage count, frame count, interval time, shutter speed and even the temperature.

Additionally, the system has an audio microphone and preamplifier connected to the encoder, allowing me to audibly hear the operation of the cameras including the motor rotation with each frame taken. I can also hear if the air conditioner is functioning (pretty Stone Age, I know)!

Best of all, I can see if the darn film camera is on. If I don't see the lights I know I'm in trouble and should telephone Batman, dial 911 or in this case, call David Insley (the Washington DC/Baltimore based director of photography brought into the project by the director), aka Mr. Time Lapse East, a nickname he has earned through his diligence and dedication to the project.

Fortunately, I have not had to make that emergency call to him, because I have always seen the lights "on" when checking on the video phone; however, I have had to call him for other minor emergencies that I will go into later in this article.

OPTICAL GLASS WINDOWS

Getting back to exposure times, the Norris LPC allows you to operate at a maximum shutter speed of 1/15 of a second. No faster speed is possible. Using simple math, it is easy to see that with the lowest speed film (ASA-50/Kodak 5245), I would need to be at about T32 to shoot on a bright, sunlit day and expect to not over-expose the film.

This is why I selected neutral density ND 1.2 filters (subtracting 4 T-stops of light) as my optical flats or viewing windows for the cameras. In this way, I can keep an iris setting of around T5.6 to T8 to transmit enough light without having a pinhole as an iris setting.

Furthermore, with the 4 stops of neutral density, I could operate the cameras with enough lack of apparent depth of field to somewhat blur or diffuse contaminants such as dust or particles accumulating on my "optical flats" or viewing windows. This also helps diffuse bugs, rain, snow, dew, reflections, etc.

I could have even gone with ND 1.5s, but I thought 5 stops neutral density might be stacking them up with only around 4% light transmission.

INTERVAL TIMES

In earlier discussions with Director Brian Leonard, we talked about the various possibilities for the interval times and amount of acquired footage per day. We decided to shoot one frame every 15 minutes, giving us at least a second a day of acquired images during each 8 hour work period, with the knowledge that one can always speed up the action in post production, but also realizing slowing the action down if necessary, undetectably, is technically impossible to achieve.

Another interesting photographic approach was that during the almost one year period in which the Monument would be in a cocoon or under wraps stage, would be a time when no apparent exterior changes would be taking place, and any progress made during this stage would be undetectable to humans or cameras.

Brian and I came up with the idea of shooting sound speed (24 fps), in addition to shooting time lapse images, once a month during this stage of work. The idea was to run off one or two 10-second long shots each month, for the period of time the Monument would be in cocoon. These obtained images would be the original elements for a series of dissolves. This would allow us to flow through the change of seasons: for instance a sunny, well-lit blue sky, then a cloudy day, then storm clouds, then rain, then snow, and then a clearing with blue sky and white puffy time lapse clouds drifting through the frame, around the Washington Monument.

This sequence of dissolved images would last about 20 to 30 seconds on screen, accompanied by a good musical background, incorporating two to three seconds of film per month, with the sound speed images depicting rain and snow, and the time lapse images depicting the leaves turning color, falling, cherry blossoms blooming, beautiful cloud movement, and the eventual regeneration of the green foliage surrounding the Washington Monument.

Next we would dissolve into the de-scaffolding stage with time lapse photography exactly like the images acquired during the scaffolding-up process, ending with several nice sound speed and time lapse images of the newly restored Washington Monument.

BONUS--"VIDEO TIME LAPSE"

Another interesting addition to our project came by introduction through a friend of mine, Sean Fairburn SOC, a highly skilled, highly experienced camera operator, who has been in charge of field testing and introducing to the film industry, a new image-gathering tool known as the Ikegami/Avid Edit-Cam, a highly complex broadcast quality video camera which records its taken images onto a hard drive computer disk instead of video tape.

This camera, in addition to many advanced features, is capable of shooting one frame at a time with any interval and can record up to 30 minutes of high resolution NTSC video on each disk. Sean Fairburn SOC was able to get Avid to loan us an Edit-Cam to test on this project, at least for the first three to five month stage of the scaffolding process.

Due to Sean's reputation and accomplishments, Mr. Joe Torelli, an Avid executive and designer, agreed to provide and deliver an Edit-Cam to me in Washington DC, in August '98. I once again had David Insley and Lee Carrick build an additional window angle correction assembly to the north-facing window in the tower, thus rotating it 45 degrees to the west.

So we installed, programmed and started the Edit-Cam rolling at similar intervals to the film cameras, with almost identical framing to the 35mm Mitchell film camera shooting out the west-facing window (which was rotated 45 degrees to shoot northwesterly) of the tower. Fairburn, who is very familiar with this video camera, was a constant source of information and expertise during this initial installation and test period.

The Edit-Cam has an auto-iris, and cannot be used at night unless I select a filter that will open the iris to T1.8 at night, and give me an iris opening of T22 during the day. Other than the range of limitations of an auto-iris, the Edit-Cam is stiff competition for my Mitchell 35mm film cameras. It works great, looks great and is a definite contender in the future arena of time-lapse image-gathering equipment.

We were also able to connect the Edit-Cam's viewfinder video output into the telephone video encoder/switcher, as this encoder can accommodate up to 4 separate video inputs, with a variable dwell timer for each camera connected. The Edit-Cam's viewfinder display tells us how many frames it has taken and what the present frame address number is and the display also reads "time lapse" to let you know you are not rolling at 30 frames per second. Observing the image of the Washington Monument in the background is also of considerable interest to us. My thanks to Ikegami/Avid Joe Torelli, and especially B. Sean Fairburn SOC.

A COUPLE OF UNEXPECTED PROBLEMS WE ENCOUNTERED

On 11/10/98, David Insley telephoned to inform me that the scaffolding was now encroaching on the area of the Norris LPC's reference target (the southeast sides of the Monument) and was putting bright scaffolding reflections, that reflected high level sunlight on varying points of the scaffolding, into the Norris LPC's spot reference meter, possibly playing havoc with our automatic exposure control.

Possible Solution A: Reposition the Norris LPC's reference beam to another similar reflective surface in the vicinity of, and subject to, similar lighting exposure as the Washington Monument itself.

Answer A: NOT POSSIBLE--there was no reflective surface with similar height as the Washington Monument, thus enabling light conditions and times to be synchronous.

Possible Solution B: Telescope the light reference beam in tighter between the scaffolding in order to avoid confusing and unwanted reflections.

Answer B: NOT POSSIBLE--shadows from scaffolding would cross into reference sample area of Monument, thus confusing light meter and causing unwanted exposure variations.

Possible Solution C: Diffuse reference beam with proper value diffuser in order to "cut down" highlights or direct metallic reflections, without severely altering light gray or dark gray objects or areas of the frame.

This solution worked great--the diffuser (now know as the LPC diffusion area extender), took one or two T-stops off the highlights, but kept the mid and low-range subjects of reference within a half of a T-stop of normal reflective value. It smoothed out and absorbed those occasional high level bright pinpoints of light on the scaffolding. In essence, it worked!

Problem 2: Exposure system for tower camera position is being confused because in the morning, only east side of Monument is receiving direct sunlight. Later in the morning, both sides are getting direct sunlight, and then later in the afternoon only the south side is getting direct sunlight.

Question: Which side of the Monument should we aim the reference beam at, to avoid radical illumination changes, as the shadows shift from one side of the Monument to the other side?

Answer: Split east and south side down the middle with reference beam; however, reset and synchronize the Norris LPC's reference beam at a time of day when only one side of the Monument is under direct sunlight. Set the camera's intervalometer so that it is at its maximum shutter speed (1/15 of a second, with only one half of the reference target in direct sunlight and the other half in shade).

This way, when both sides become lit under direct sunlight, the exposure time cannot increase, and it remains the same at its maximum shutter speed (1/25 of a second). This enables a smooth exposure control with the same shutter speed for one or both sides of the Monument lit by direct sunlight.

Though the camera wants to go faster when it realizes both sides are lit (seeing almost 50% more light on reference target), it cannot go to a faster shutter speed, being that it is already at its fastest rotation or shutter speed. The result is a smooth exposure transition, or lack of transition from 1 to 2 to 1 sides sunlit without exposure variation or stop-down occurring on the shutter speed when both sides become illuminated.

CLOSING

I feel I have given you, the reader, enough information about this project for the time being. Though at this time the project is far from completion, this should give you an idea of just what is involved in such a long-term time-lapse project. As progress continues, I will inform the SOC of the developments in a later article following project completion in June 2000.