A machine to save old film: Greenpoint’s Kinograph

Film is rapidly degrading. It’s physical stuff and for miles and miles of film, there is only one copy in the whole world, and it’s losing quality as we speak.
Matthew Epler is an NYU ITP grad, which means that he’s a technically-skilled person who has at least hung out with a lot of artists. (Sorry, it’s always hard to pin down how to define anyone from there.) In 2008, he was a part of the first class of faculty at the Red Sea Institute for Cinematic Arts. While there, he stumbled on a huge trove of film that had nearly been thrown away. Read more on that story here.
If you know some Arabic (and other languages, too), you can help him translate the labels on the canisters he found.
But Epler also has a new company: Kinograph, a hardware venture building an open source system for converting 8mm, 16mm or 35mm film to digital. It basically takes an image of each cell of film, then software extracts the sound and matches it all up to the images. It’s mechanical. It’s data. It’s magic.

We first met Epler at the New York’s Next Top Makers event in Downtown Brooklyn. He lives in Greenpoint and is a company of one.
He’s looking for advice on the mechanical side of his device, as well as optics and working with the image sensor. And Epler is happy to listen to anyone’s opinion about how to best turn it into a business. Epler wants to find a business model that works because he strongly feels the clock is ticking on film.
Every day, it loses a little quality. Machines do exist to make the conversion, Epler explains, but they are huge and expensive. Epler’s system is about the size of a big school desk and is built from fairly easy to find parts, with a couple exceptions we’ll mention below.

###

What technology helped the most with the mechanical side of your product so far?
3D printing and CAD software, without a doubt. Any machine that deals with transporting film or tape from one reel to the other is going to have rollers on it. These guide the film along its path — a simple job, but they’re deceptively expensive parts. If they’re plastic, they are injection molded and to do that you have to pay someone to make a master mold out of metal. If the rollers are metal, then you either have to lathe each one or go for injection molding again.

I stand on the shoulders of generous giants. Without open source software, Kinograph may not exist.

I used Rhino to create my own rollers and then printed them at NYU’s Advanced Media Services lab, which has a high definition 3D printer. In the end, I was able to get custom rollers with grooves on it for 35mm, 16mm, and 8mm films for less than $40 each. And that price will only drop as my designs improve and the price of printing comes down.
What software helped the most with the software side of Kinoscope?
OpenCV (“an open source computer vision and machine learning software library”). I stand on the shoulders of generous giants. Without open source software, Kinograph may not exist.
What was the skill you really had to buckle down and learn to get Kinoscope functional?
Before Kinograph, I had only done projects that did one thing well. Kinograph forced me to integrate everything I knew into a single product. Code, circuits, digital fabrication — these were all integral. Without one, the thing doesn’t work. When you have to connect seemingly disparate bodies of knowledge, it’s like learning them all over again.
Where’s your thinking at right now for putting a business model behind the machine?
Harder than building Kinograph is finding a business model for it. Harder still is figuring out how to start out — which version of Kinograph should be the first to come to market? Who is it for? How much should it cost?
There are two markets: people with old home movies (8mm), and institutions with lots of films (35mm, and 16mm). I also see an opportunity for creating a stock footage market, as well as a social platform for home movies.
I’ve talked to dozens of really smart and experienced people about this and I get all different kinds of answers. “Go for the consumers!” or “They money is in the institutions if you can get to it.” The only consistent advice I’ve gotten always comes at the end of the conversation: “It really comes down to what you want Kinograph to be.” That’s the worst because I want it to be all those things.
I invented this machine because I had a real problem to solve. Of course I’d like to see it solve other people’s problems too. But I’m more inclined as an inventor to solve the next immediate problem, not build a business structure. That’s going to be a hard learning curve for me but I’m looking forward to it.
Why the commitment to open source?
I built Kinograph on open source tools. I can’t turn my back on that. Also, we’re talking saving films which implies a ticking clock. The more people doing it, the better the chances are we can save our visual heritage. From a humanist perspective, it just seems like the right thing to do. It’s probably naive but it’s the backbone of the project’s philosophy.

###

Here’s one of the films saved by Kinograph: