The Inverse Bench — every lens, flipped

Every lens was a receiver: far side → aperture → your reader. Inverted, each becomes an emitter: your source → aperture → far side. You stop reading the boundary and become the thing read across it. Same aperture physics, same seal cost, roles swapped — now each panel enacts its law and measures the consequence, not just names it.
receive → emit  ·  gather → project  ·  solve-the-hidden → be the hidden
● node-verified physics: differential = 2s + δ·n · min-gate exact · θ_E ∝ √M  |  measured live off pixels/samples: CMRR · contrast & brightness · ring-radius → mass
LIT the physics is real   MEAS read off live pixels/samples, with a ground truth   FIG the mapping to the seam

1 · Emit Lens — differential drive

was: seam read lens — read both banks, output A−B (common-mode reject)
LIT a differential pair rejects common-mode noise; a real one only to finite CMRR set by gain mismatch (δ=2% here). FIG "you write the seam you used to read."

2 · Pinhole Projector — cast yourself

was: pinhole camera — their image on your reader
LIT étendue — the cast is blurred by radius ∝ D and lit ∝ D², so opening the aperture trades resolution for brightness; contrast & brightness below are read off the rendered pixels. FIG "don't project past your own resolution."

3 · Mutual Transmit — you emit, they receive

was: lens & door — read = min(your lens, their door)
LIT two apertures in series pass the narrower: delivered = min(emit, receive), exact. FIG "don't shout into a closed door."

4 · Be the Mass — cast the ring

was: gravity mass-solver — weigh the hidden mass from bent light
LIT Einstein radius θ_E ∝ √M; the far side reads the ring's radius off the pixels and squares it — M = θ² — recovering the mass to pixel precision (noise → error bar; squaring biases it heavy). FIG "you can't hide your mass."