raspberrypi.org/magpi The Official Raspberry Pi Projects Book 39
“Most were due to my inexperience
and ineptitude,” he tells us. “Or
bad programming techniques.”
The first proof-of-concept device
drove the projector and triggered
a capture as each frame advanced.
“It fooled me into thinking this
would be an easy project,” recalls
Joe. But frames were often under-
or over-exposed. Sometimes it
would miss frames altogether.
As Joe iterated through different
versions of the project, he added
a separate client computer: a PC
running Ubuntu Linux.
The client computer runs a
control panel that lets the user
adjust the camera, and displays
the captured frames as they come
off the Raspberry Pi. It also sends
some simple projector control
commands to the Pi.
The latest version of the Pi
Film Projector uses OpenCV
(opencv.org). This software adds
computer vision and multi-core
processing to the project.
“OpenCV provides functions
similar to the HDR mode on cell
phone cameras,” explains Joe.
Given several photos at different
exposures, it can blend them into a
single image that preserves detail
in both bright and dark areas.
“For some films, this can make
an enormous difference,” says Joe.“The downside is it slows down my
captures by nearly half as I take
more photos, so I try only to use
it when necessary.”
Joe advises prospective makers
to have patience with the build.
“It has a lot of pieces,” he says.
The captures take a lot of time, and
getting good results can take a lot
of practice.
The results are worth it. Joe
has converted over 100 of his
grandfather’s reels. “Bringing
some family history to life again
and sharing it has made this the
most rewarding tech project I’ve
ever worked on,” he tells us. “I
hope it inspires other people to
bring that old box of film cans out
of the closet and reconnect with
their history.”PI FILM CAPTURE Projects
Above As the
project has
progressed, the
image editing
software has
become more
advanced, creating
higher-quality
recordings>STEP-01
Direct capture
A Pi Camera Module is stripped back and
placed directly inside the lens housing.
This approach removes any interference
from outside light.>STEP-02
Client computing
Each still frame of the reel is captured,
one frame at a time. The Raspberry Pi
sends the images to a client computer
for storage and editing.>STEP-03
Stepper motor
After capturing the frame and sending
it to the client, the Raspberry Pi controls
a stepper motor that moves the film to
the next frame.BUILDING A CINE-CONVERTER
Joe has captured over 100 reels
of film recorded by his grandfather