Anthony Caccese built the open source 240 MP after HD remasters of older shows revealed production seams 4:3 CRTs had hidden, including gaffer tape on Star Trek doors.
When Anthony Caccese rewatched Star Trek: The Next Generation in HD, the picture that came through was sharper than the one he remembered, and worse for it. The upgrade exposed production seams that 4:3 cathode-ray tubes had quietly hidden for decades: gaffer tape on the Enterprise's set doors, covering stenciled names from earlier episodes that were never fully painted out. The higher resolution did not preserve more of the show. It preserved different parts of the show, the parts that were never meant to be seen at the size of a living-room wall.
Caccese, whom The Register identifies as a principal product lead for enterprise platforms at Oak Ridge National Laboratory by day, is not the first viewer to notice this. He is, however, the one who built a response. The project is called 240-MP, and it is published on GitHub as open source. At the surface it looks like a lark: a text-based, on-screen menu styled like the interface of a 1990s VCR, sitting in front of files a Raspberry Pi is already playing. Underneath the menu is MPV, the command-line media player. The point of the project, as Caccese told The Register, is to make the choice of how a file is watched a deliberate one again, rather than a default that streaming services and modern displays impose.
The premise is simple enough to test against a concrete example. Older television was shot and finished for a 4:3 image on a CRT, where soft phosphors, limited scan lines, and modest brightness hid things that the production team had not fully cleaned up. The set dressing in The Next Generation was a known offender: doors reused between episodes carried identifying names from earlier scenes, with tape layered over them. On the original broadcast, the tape was invisible. On a 4K remaster streamed to a modern television, the tape is not only visible, it is obvious. The "improvement" did not recover information that had been lost. It recovered information that had been deliberately, if not consciously, suppressed by the limits of the original medium.
240-MP is what Caccese built to put that choice back in the viewer's hands. The interface is text menus. The remote is whatever the user pairs with the Pi. The library is whatever the Pi can see: local storage, a USB drive, an external disk, a network share, or a Plex library through one of the project's dedicated modules. Files play through MPV, which is a competent, modern player. The only thing 240-MP adds is the wrapper that makes navigation feel like choosing a tape from a shelf, and the option to send the output to a real CRT through a Pi-compatible composite cable, or to a modern display over HDMI. There is a configuration step that has to happen before install: the Pi's config.txt has to be set for whichever output the user is targeting. That detail matters, because it is the difference between getting a VCR-styled menu on the screen the user wants to watch, and getting a VCR-styled menu on the screen the Pi assumed.
The honest reading of the project is that it is small. Caccese is one developer, working in his own time, and the project is tested on a Raspberry Pi 3B, 3B+, and 4B, per The Register's writeup. The 240-MP repository on GitHub is the only place to see what it currently supports, what it does not, and what its install footprint actually is. It is not a Plex replacement, an MPV fork, or a Raspberry Pi roadmap item. It is a single maker's argument that the modern viewing stack has stopped asking the viewer what they want, and a tool that puts the question back.
That argument lands because it is falsifiable. The TNG example is not a quirk. The same logic applies, in different ways, to any visual artifact whose appearance depended on the limits of the original medium: film grain that was never meant to be inspected frame by frame, color grading that was tuned for a dim CRT, video compression that hid more than it removed. HD upscaling and modern displays do not reveal "what was really there." They reveal what the original production was hiding, intentionally or not, by virtue of the medium it was made for. Treating that as a flaw, and choosing to watch the older picture on its own terms, is a defensible position. It is also one that the major streaming services have no product surface for.
What 240-MP offers, in other words, is not a better picture. It is a refusal to accept the default picture, and a small, working menu for someone who wants to choose. The Register covered the project in its offbeat wire on June 10. The more useful frame is that 240-MP is what a constructive response to autoplay, algorithmic recommendations, and lossy upscaling looks like when it is built by one person, on a $35 computer, in their own time.