Skip to content

Yaffoman/AutomaticFollowFocus

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 
 
 

Repository files navigation

AutomaticFollowFocus

Team

Dayyan Sisson - Software Lead

Ethan Lerner - Hardware Lead

Gabriel Beal - Software/Hardware Tech

Overview

Motivation

Professional-grade consumer cameras are used by more and more cinematographers for a variety of projects, ranging from amateur shoots, to mid-size productions. While these cameras are developing better digital autofocus, these systems only work with compatible lenses. Since most cinematographers in this demographic use a variety of lenses, reliable digital autofocus is rarely an option. Manual focusing requires skill and experience in order to maintain focus in a scene. A follow focus is a device which allows for more precise control of focus, but still requires highly skilled operators (which often necessitates a second operator) while capturing a scene.

Our solution

Our solution is an automatic follow focus system. It would operate much like an industry standard manual follow focus system, but would automatically keep a target subject in focus without the need of human intervention. The automatic follow focus, after minimal calibration, would give camera operators a smooth, reliable, and precise focusing mechanism, that would remove the need for highly skilled focus operators, reduce the amount of thrown away takes, and shorten the scene planning process.

Approach

To create an automatic follow-focus, our system will need to be broken down into three main components: the follow-focus-rig, the depth-sensor, and the computer. The follow-focus rig would provide the mechanism for focusing a manual camera lens, and will follow the same structure and components of industry standard manual follow-focuses. Ideally, our system could be attached to existing follow-focuses. This component also includes the motor components needed to automatically drive the follow-focus. The depth-sensor would provide the data needed to determine how far and/or where the subject of the shot is to be able to determine focus. The computer would do the necessary calculations to control the follow-focus motor after reading in the depth-sensor data.

For our MVP, we will adapt an existing manual follow-focus rig with a small, fine-grained, and silent stepper-motor to control focus. The depth-sensor will be a LiDar time-of-flight sensor to determine distance, and an Arduino UNO as the computer. Once attached and calibrated to the lens, the focus will dynamically adjust so that whatever is at the center of the frame always remains in focus up to a range of 8m.

Media

Original Mount

This mount was made using a flat piece of steel which we bent and removed sections using a dremel. This was to give us a basic idea of what we were going to model.

original mount

Original System

This is the original mount attached to the camera with some of the system components attached (stepper motor).

original system

3D Models and Iterations

This is the 3D model we designed based on the original mount. We included mounting points for all of the systems hardware components. This took several iterations to get the mounting locations in the ideal placements.

original system

original system

Final Iteration: Carbon Fiber mount with attached hardware

original system

original system

Complete System

Here is our complete system wired up and running. The wires are a bit messy and given more time we have several changes that we would like to implement but this system achieves our base goals.

original system

original system

original system

original system

Repo Organization

Images

This folder contains all the images needed for the README

Code

This folder contains our audino file that controls the system

Relevant Files

Project Specification

Milestone Report

Oral Update

Final Oral Update

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages