Long Term Time Lapse Motion Controller

I got asked about the possibility of making a 10 metre long timelapse track that would be running for 5 months in an orchard in Scotland.

For various reasons it didn't happen but not before I'd worked out a proof of concept. The idea is that the camera does the whole length of the shot in a single day and then the five months of passes are edited into a single timelapse shot.

Plywood and string demonstration of concept Proof of concept. The laptop represents an embedded computer,to the left of that is a cased computer that is used to store the images form the Canon 50d. The opened boxes contain stepper motor drivers and an intervalometer. This would all be enclosed in an IP65 box with only an LCD display and some buttons visiable. Below is a representation of the LCD display on a laptop screen(with an LCD display held in next to it).

Representation of the LCD display.

Mechanically the intention was to make 10M of bearing track and put a dolly similar to the KT dolly on it. This would be driven by a belt similar to the regular KT dolly but the drive shaft would run into an industrial IP65 housing (via a seal) and inside the housing would be a stepper motor and drive .. a computer would also be built into the same housing.

The idea is to shoot the same timelapse pan every day and then post produce those into a single shot. Here is a test shot with a pan and tilt head.In theory a difference matt is the ultimate "Proof" of a motion control shot. If the lighting were the same then this would have been a black screen. However you can see the changes in lighting between the passes and the movement of the trees.

The problem with a job like this is that the hardware to control the motion control rig is so expensive .. A copy of Kuper alone is £4000 and Kuper is not ideally suited to long term unattended jobs. I have developed an embedded computer that has overcome this problem.

In the method I developed the dolly would be connected to a Kuper rig first and the move demonstrated to the director (possibly not in real time) and then the Kuper rig would be removed and replaced with a single board computer with controls and only an LCD display. The move could then be loaded onto the embedded computer via a flash card.

The embedded computer would work like this ..

On boot it would go in a negative direction until it found the 0 end of the track (if it found the far end of the track then the positive direction flag would be changed and it would search for the 0 end of the track again.

Then it would go to the number one position as read from the CF card .. then it would wait until it sensed the flash sync connector closing. This is a separate circuit with a delay variable by a control on the box and can be adjusted at any time. So for instance the flash is triggered and then the output is held high for ten seconds to allow for the exposure. On the falling input the rig moves to the next position as determined by the CF card .. The LCD display shows the present frame and the present position ..

Head being used as a test rig for the proof of concept.
The board on the right triggers the digital still camera the board on the left "catches" the flash sync and then holds for a variable time. This time can be varied on site without a trained operator. The head moves to the next position. Then the Framecounter on the laptop screen (representing an LCD display) advances. Then the next move is triggered by the "flash catcher". No my income does not depend on my hand holding skills with a mobile phone fortuanatly.The setup showing the flash connection.

If the camera is triggered more times than the move takes then the rig will stop in the last position and hold there ..

On an external input (possibly a time switch set for midnight or an Light Dependent Resistor) then the rig will search for the far end of the track (and record any error) then return to zero and then to it's number one position. As it runs grease is pumped through the front two bearings to re grease the track to protect it for another day.

It will wait in it's number one position for the first shot of the new day.

The advantage of this is that come what may .. hell, high water long term power cuts the rig will reset itself automatically without any need for a technician to attend the site .. the worst that can happen is that the the rig stops or is inaccurate for part of a day before reseting itself at midnight (should there be a power cut and then it ends up trying to go over it's end stops then it will automatically reset .. ). All the settings (automatic reset, move hold time etc, manual hard reset) can be adjusted by a non technical person without having to reprogram the software.

© Justin Pentecost, 2011.