--- title: "He paid $20 for a C-mount lens. He got a Super 8 projector lens." locale: "en" url: "https://irz.fr/en/articles/kmart-digicam-projector-lens-en" markdown_url: "https://irz.fr/en/articles/kmart-digicam-projector-lens-en.md" category: "craft" tags: ["photography", "optics", "modding", "3d-printing", "camera", "tinkering"] published_at: "2026-09-05T02:30:00.000Z" author: "Léa Perrin" translation: "https://irz.fr/fr/articles/kmart-digicam-projector-lens-fr.md" --- # He paid $20 for a C-mount lens. He got a Super 8 projector lens. A 'C-mount lens' bought for $20 on eBay turned out to be a Super 8 projector lens. Mason Mirabito redesigned his 3D-printed adapter around it, and learned why a camera's geometry has to be discovered with a file. The lens cost $20 on eBay and was supposed to be a C-mount. When it arrived, Mason Mirabito found 32mm of thread and a lens designed for a **Super 8 projector**. His freshly printed adapter no longer fit: he redesigned the whole thing, wider.[1](https://mason.bearblog.dev/kmart-digicam-mod-part-2) Mirabito tells the story on his blog, between film photography and programming projects.[5](https://masondoesthings.com/) It starts with a simple urge: mount C-mount lenses on the cheap camcorder he bought at Kmart — an Anko JLR-80051, the chain's house brand. He knows from the start that the sensor won't improve — he calls it "a shitty 1/6" 720p mess," a chip that will never be good. It's not the image he's after; it's the build.[2](https://mason.bearblog.dev/modding-a-kmart-digicam-to-take-c-mount-lenses-part-1/) ## Measure what's there The camera opens easily. The stock lens is a small threaded module, roughly M5-sized, buried in epoxy: it comes off after a few strikes at the joint. The sensor connects through a ribbon cable straight to the main board, with no separate video-processing module.[2](https://mason.bearblog.dev/modding-a-kmart-digicam-to-take-c-mount-lenses-part-1/) On the front, a removable cover with an offset pinhole lets the light in. Mirabito takes it off with a screwdriver, gets out his calipers, and records the screw-hole positions and the lip.[2](https://mason.bearblog.dev/modding-a-kmart-digicam-to-take-c-mount-lenses-part-1/) > Illustration: The opened Kmart camcorder, showing the sensor connected by a ribbon cable to the main board. The camera opened: stock lens removed, the sensor sits at the end of its ribbon cable, straight onto the board. Credit: [Mason Mirabito](https://mason.bearblog.dev/modding-a-kmart-digicam-to-take-c-mount-lenses-part-1/). ## Too long, on purpose Then comes the clever part. Instead of defining the exact adapter length on paper, he prints one **deliberately too long** — a 15mm tube — with an estimated flange distance between 20 and 23mm, against the C-mount standard of 17.526mm.[2](https://mason.bearblog.dev/modding-a-kmart-digicam-to-take-c-mount-lenses-part-1/) The plan is explicit: print long, then file and cut in small passes, testing each time, until focus returns. The CAD file, published on Onshape, is updated as the measurements come in.[6](https://cad.onshape.com/documents/e254b5aff45950c222657cbe/w/d2bff3c5693d8ed55c12a4b5/e/f86c7baa9d19c08f166a861c) For anyone printing their own, he adds two tips: print at the highest resolution, and upside down, so support material doesn't grip the lip that must sit against the camera.[2](https://mason.bearblog.dev/modding-a-kmart-digicam-to-take-c-mount-lenses-part-1/) The method has a name in machine shops: qualify instead of calculate. You build the part long enough, approach the right setting by subtraction, and the real part becomes its own gauge. It won't replace proper metrology, but it teaches the system faster than a spreadsheet. > Illustration: Calipers and the measured specs of the Kmart camcorder's front cover. The calipers come out in part one: screw-hole and lip measurements, taken before drawing the first adapter. Credit: [Mason Mirabito](https://mason.bearblog.dev/modding-a-kmart-digicam-to-take-c-mount-lenses-part-1/). ## The wrong lens Then the parcel arrives. The thread is not a standard C-mount: 32mm in diameter, with a pitch he calls "really weird." The redesigned adapter gains a lot of diameter, and the printed thread only fits because he filed it down, patiently, until it was right.[1](https://mason.bearblog.dev/kmart-digicam-mod-part-2) > Illustration: Second view of the printed adapter, its thread worn down with a file to fit. The new adapter's thread, filed to mate with the unusual pitch of the Super 8 projector lens. Credit: [Mason Mirabito](https://mason.bearblog.dev/kmart-digicam-mod-part-2). The lens mounts, it makes light, but it won't reach infinity focus. What it hits is revealing. A taking lens follows a standardized flange distance: the gap between the mounting face and the focal plane. The C-mount sets it at 17.526mm. A projection lens has no flange and no standard: only the distance between the rear element and the film gate matters, a figure unique to each design and usually very short, since the film passes a few millimetres from the rear element.[3](https://forum.mflenses.com/projector-lenses-t69484.html) So the focal plane of this Super 8 lens should sit almost against its back. The Kmart's sensor sits several millimetres further away, behind the adapter and the chassis. The image forms too far from the sensor plane to ever reach infinity. To fix it, the sensor has to move closer to the lens.[1](https://mason.bearblog.dev/kmart-digicam-mod-part-2) ## The slack is on the sensor Except the adapter is already pressed as far as it can go against the chassis. There's no length left to grind: the only part that can still move is the sensor.[1](https://mason.bearblog.dev/kmart-digicam-mod-part-2) Mirabito tried paper shims under the sensor board, with no clear change in focus. His to-do list for next time is refreshingly honest: find a way to physically move the sensor closer to the lens.[1](https://mason.bearblog.dev/kmart-digicam-mod-part-2) This is where a small project stops being just tinkering. The mod was designed so the geometry reveals itself by touch — print long, file, test — until the slack shifts sides: it's no longer the adapter that has to shrink, it's the sensor position that has to be rethought. A general lesson for closed optical systems, where published specs never suffice because every real part adds its own tolerance: the epoxy layer, the slop in the printed thread, the height of the chassis. And when you buy the "wrong" lens, the standard disappears: you have to measure its geometry instead. The episode, self-posted on Hacker News, gathered about twenty points.[4](https://news.ycombinator.com/item?id=49417662) Nothing spectacular. But for a cheap camcorder whose sensor "will always be bad," that's exactly the right outcome: a logbook where dead ends matter as much as wins — and proof that you can learn a lot about a system by filing it down. ## References 1. [Mason Mirabito, « Kmart Digicam mod part 2 », August 24, 2026](https://mason.bearblog.dev/kmart-digicam-mod-part-2) 2. [Mason Mirabito, « Modding a Kmart Digicam to take C mount lenses Part 1 », August 6, 2026](https://mason.bearblog.dev/modding-a-kmart-digicam-to-take-c-mount-lenses-part-1/) 3. [mflenses forum, « Projector lenses »](https://forum.mflenses.com/projector-lenses-t69484.html) 4. [Hacker News, discussion « Kmart Digicam Mod Part 2 »](https://news.ycombinator.com/item?id=49417662) 5. [Mason Mirabito, portfolio](https://masondoesthings.com/) 6. [Public Onshape CAD of the adapter](https://cad.onshape.com/documents/e254b5aff45950c222657cbe/w/d2bff3c5693d8ed55c12a4b5/e/f86c7baa9d19c08f166a861c)