The test
A source and a knife near the centre of curvature: the knife cuts the image, and each zone of the Couder mask darkens at once as it passes its focus. How those positions differ from zone to zone gives each zone’s slope, then the glass’s profile against the nearest paraboloid, by Jean Texereau’s method, the one Foucault2 uses.
The readings
For each zone, the knife’s position in millimetres (or inches) from any origin: only the differences matter. Two zones read are enough for a quick check between two sessions, all of them for a result. Each zone is read at its window’s radius, √(z² + h²/2) for a middle z and a half-width h: a little past the middle, the outer part letting more light through.
The results
On the wavefront, at λ = 560 nm, against the best paraboloid: the peak-to-valley error (PV), its RMS, the Strehl ratio, and lf/ρ, each zone’s transverse aberration in Airy disc radii. The Rayleigh criterion, λ/4 peak to valley, is the usual limit of a good mirror. The centre hidden by the secondary can be left out of the figures, as in Foucault2.
The profile
The surface’s departure from the best paraboloid, in nanometres, the lowest point at 0; the vertical lines are the zones’ edges, the orange line the Rayleigh criterion (70 nm on the glass). A centre left out of the figures is shaded and dashed. The history’s thumbnails are drawn the same way, with their highest value.
The zones
Their number keeps each window evenly lit, with at most 37 nm of path difference across it: n = D²/16√(R³k), from 3 to 15, fewer on a small fast mirror. They are of nearly equal area, the first a little smaller; the 70 % zone, where a paraboloid departs most from the sphere, always falls in the middle of a zone, never on an edge.
The mask
Two windows a zone, symmetric about the centre, each zone on the other side of the diameter from its neighbours, so that no window touches another or overlaps it in the knife’s direction (2 mm at least). The first, round the centre, is a sector away from the vertical diameter, where the knife’s shadow is least clean. A home-made mask goes in by its windows’ radii.
Cover and selector
The cover to print: a 1.2 mm plate, a 10 mm wall 0.75 mm from the glass, a 45° seat for the bevel, two notches for the stand’s supports. From 6 zones on, a selector turns in front of the mask on a pin and shows only two zones at a time, the first three together for an odd number: the number at the top says which.
History and comparison
The saved readings, kept in this browser with each telescope; Export makes a file of them, and a Foucault2 file imports as it is. “Compare with…” shows the glass removed between two measurements: their difference against one same paraboloid, its focus term taken out, the least worn place at 0, glass having only been taken off.