vivar कृष्ण विवर (source on GitHub)

1. The shadow. Light aimed inside a critical impact parameter b_c = (3√3/2) r_s ≈ 2.6 r_s spirals into the horizon. The dark disc is 2.6× wider than the horizon itself because gravity bends rays into it. The D-shaped shadow. A spinning hole drags spacetime around with it (frame dragging). Light orbiting with the spin can skim closer before it is lost, light orbiting against it is captured further out, so the shadow is flattened on the side turning towards you and shifted the other way.
2. The photon ring. Rays that just miss b_c loop around the hole once or more before escaping. Each loop squeezes another, thinner image of the whole sky and disk into this bright edge.
3. The far side of the disk, over the top. The disk is flat and thin. Light from the part behind the hole is bent up over it, so you see the back of the disk as an arch above the shadow.
4. The underside, below. Light from the far side that passes under the hole shows the bottom face of the disk as a thinner arc hugging the shadow from below. Rays that cross the disk plane more than once add up here.
5. Doppler beaming. Gas near the inner edge orbits at up to halfa large fraction of the speed of light. The side moving towards you is blueshifted and beamed (intensity ∝ g⁴), the receding side is dimmed and reddened. Toggle it to compare.
6. The ISCO. Below 3 r_s no stable circular orbit exists, so the disk ends there and gas plunges in. Light from the inner disk is also redshifted by gravity: it climbs out of a deep potential well. The ISCO moves in. For gas orbiting with the spin, the innermost stable orbit is at 3 r_s instead of 3 r_s, so the disk reaches deeper, runs hotter there, and its inner light is strongly redshifted.

Numbers are placed by tracing the same rays on the CPU. Turn on Guides to see the horizon, photon sphereequatorial photon orbits and ISCO circles, and the predicted shadow edge (dashed).