This repository contains code for a Raspberry Pi Pico microcontroller running Kaluma.js. The code operates two servos to control zoom and focus for a camera lens, creating the 'dolly zoom' camera shot (example below).
Example of ‘dolly-zoom’ camera shot from Jaws (1975).
The 'dolly-zoom' shot is a notorious technique in cinema that is frequently employed to depict a character's sudden realisation or a significant turning point. In the example above from the film Jaws, it captures the moment when Martin Brody becomes aware of a deadly shark attack.
Executing a 'dolly-zoom' shot necessitates the involvement of at least three crew members: one to handle the zoom adjustment, another to control the focus, and a third to operate the camera dolly for forward movement.
Behind-the-scenes photo of the Jaws example above. Cinematographer Bill Butler is aided by 4 members of crew to achieve the iconic shot.
By utilising a microcontroller, we can automate the zoom and focus aspects of the 'dolly-zoom' shot, achieving professional results without the need for additional crew members. This is the idea behind the project.
The following equipment is used to run the servos and attach them to the camera:
- Raspberry Pi Pico
- 2x MG996R 180° servo
- 2x Servo bracket for 15mm rod
- 2x 56 tooth gear for servo
- Battery Holder 3xAA with MicroUSB (to power the Pico)
- Battery Holder 4xAA with jumper header wires (to power the servos)
- Lots and lots of jumper cables
You can see this equipment used in the set-up photo below:
The system uses servos that are mounted using 3D-printed brackets. These brackets securely attach to the 15mm rails located beneath the camera, which is a standard component in filmmaking equipment.
The code is currently calibrated for a Sigma 18-35mm lens moving forward on a 60cm track. To calibrate to a different lens and distance, the fromDegrees, toDegrees, and duration arguments cant be altered for the rotate function(s) in the index.js file.
You can see an example of the device in action in the GIF below, which is an excerpt from a short film I made!
The effect is far less pronounced on a 60cm rail but still adds emphasis to the shot.
This movement could be automated even further by utilising an ultrasonic distance sensor and doing the following:
- Measure the distance in real-time between the camera and the subject using the ultrasonic distance sensor.
- Integrate the distance measurements obtained from the ultrasonic sensor into the zoom and focus adjustment mechanism.
As a starting point, the US100.js file included in the repository to gather readings from the US100 ultrasonic distance sensor.