cert-machine · playground

Nothing here is certified. That is the entire permission.

The rest of this repository proves things, and gates itself accordingly — a page there cannot ship if a number on it has drifted from the record underneath. This place is for the other half of the work: the instruments made touchable, the mathematics dressed so that you would want to look at it. Everything here is real and computed on this machine from data on disk. None of it is a proof, and none of it will pretend to be.

the projects one, so far
longest baseline 7.35 Gλ
This is the whole of what was measured. 7,317 samples from 26 telescope pairs on 21 April 2018, each pair sweeping an arc as the Earth turns, each arc mirrored because a real sky makes every measurement two. Everything between the arcs was never observed — and that empty space is the reason there is a set of pictures rather than a picture.
interferometer · M87 · 227 GHz

Not the picture. The set of pictures.

The famous black-hole image is one sky chosen, by an imaging prior, from the infinitely many the data allow. This draws the set instead: 18 skies fitted under deliberately different priors, agreement rendered as ink and disagreement as texture. One slider decides how much of the phase information the picture is allowed to use — and at zero the ring dissolves, because amplitudes alone cannot see where anything is.

visibilities838released rows, one calibration pipeline, one night
stations826 baselines, 544 closure triangles
skies drawn18 + 4fitted members, plus extremes that push the data’s limits
bracket82%worst of 4 radii — a prior-free, phase-free enclosure, best 67%
open the instrument →
how this place works

No gates.

This folder is built by its own thirty-line script. It is not wired into the repository’s site build and none of that build’s refusals apply to it. Nothing here can block anything there, and nothing there can slow anything here down.

But no fiction either.

Every number on every page comes out of a record in this folder, and every record comes out of running the code beside it. Freedom from ceremony is not freedom from arithmetic. Where a thing is a float and not a proof, the page says so in the same breath as the number.

The data is not ours.

The Event Horizon Telescope files here are the public release, byte for byte, under the collaboration’s own terms — the code is MIT and the data is not. They are carried so the pages can be re-run rather than believed.