Black hole · Frequently asked questions

A black hole is a region of spacetime where gravity is so strong that nothing, not even light, can escape once it crosses the event horizon. It forms when a large mass is compressed into a tiny volume. This simulation shows its real effect: light passing nearby is bent (gravitational lensing) and a photon ring appears around the central shadow.

Seen from far away, you would appear to slow down and redden until you «freeze» at the horizon (gravitational time dilation). For you, locally, crossing the horizon feels unremarkable — but the difference in gravity between your head and feet would stretch you: «spaghettification», which you can watch applied to stars in the Playground.

Not directly — it's black — but you can see its shadow silhouetted against the hot gas orbiting it (the accretion disk) and against background stars warped by lensing. That is how the Event Horizon Telescope photographed M87* and Sgr A*: the simulation's «EHT» mode mimics that image.

It is a real simulation. For every pixel we numerically solve the metric's exact null geodesic (Schwarzschild, and Kerr with spin): gravitational lensing, the photon sphere, the shadow and the photon ring all emerge from the general-relativity calculation, not from tricks. We trace rays from the camera backwards, but light paths in GR are reversible, so the image is exactly the one real light would produce.

They are physically correct, not artifacts. Every background star actually has infinitely many distinct images: the primary (rays that do not orbit), the secondary (rays that orbit once), the tertiary, and so on. The boundary between successive orders is a precise mathematical caustic — a ray just outside does not orbit, a ray just inside orbits exactly once — and it generates a sharp ring in the star field. The rings close in toward the shadow by a geometric factor of e^{−π} ≈ 1/23 per order: the first ring is visible, the second barely so. They appear in every correct GR ray tracer, including the DNGR system that produced Gargantua for Interstellar.

No, and it correctly shouldn't. The geometry, measured in Schwarzschild radii rₛ, is identical for every mass: black holes are scale-invariant. Mass only changes the absolute scale (rₛ being 30 km or an astronomical unit) and the disk colour. What does deform the grid is spin: frame-dragging twists it.

It is the exact Kerr metric. With the Spin slider we integrate the exact Kerr null geodesics in real time (in Cartesian Kerr–Schild form, with no coordinate singularity). The asymmetric shadow, the displaced and flattened photon ring and the dragging of inertial frames all emerge — the same metric as Interstellar's Gargantua.

Yes. The inner edge is the prograde Kerr ISCO (Bardeen's formula): it shrinks toward the hole as spin increases, so the hot ring tightens. The Doppler and redshift follow the exact Kerr metric, and the radial flux is the relativistic thin-disk (Novikov–Thorne) profile with that spin-dependent inner edge.

No: in the playground they are compressed so both are visible. The real ratio depends entirely on mass: around a stellar-mass hole (~10 M☉, rₛ≈30 km) a star is thousands of times larger than the horizon; around a supermassive one (Gargantua) the horizon dwarfs any star.

No. The disk emits as a blackbody at its temperature, with procedural turbulence for the gas structure; magnetized plasma, synchrotron emission and polarization are entirely absent — they would need a GRMHD solution, feasible only offline. The thermal side and the relativistic radiative transfer (redshift, beaming) are correct.

It is the same Kerr metric. The difference is the computation: Interstellar's Gargantua was ray-traced offline (hours per frame); this runs in real time in the browser. Interstellar's disk is an artistic model too (no plasma), like ours.

Yes for the lensing and the orbits: integrating the geodesics (null and timelike) is exact and reproduces the photon sphere, Einstein ring, shadow, ISCO and periastron precession. Everything approximated — disk profile, procedural turbulence, stylized jets, un-lensed playground bodies — is stated openly.

The «Orbits» demo integrates the exact Schwarzschild timelike geodesic for a single body (exact precession and ISCO, with a conservation diagnostic). The Playground uses the Paczyński–Wiita pseudo-Newtonian potential, which reproduces strong-field effects but allows mutual N-body gravity and tidal disruption — an exactness/interactivity trade-off.

The «Ringdown» button simulates a perturbation of the black hole: the photon ring (the returning radiation — rays that have completed at least one full orbit around the hole) pulses with the damped sinusoid A(t) = exp(−ω_I·t)·cos(ω_R·t). Frequencies are calibrated to the tabulated Kerr l=2 quasi-normal mode values (Leaver 1985). The quality factor Q = ω_R/(2ω_I) = 2.1 + 11.3·a³ is physically correct: at zero spin the hole rings ~2 times; near a≈0.99 it rings ~13 times before settling. This is the same «ring» LIGO measures after every black-hole merger — but here you can vary the spin and watch how the number of oscillations and the damping timescale change.

It is Olbers' paradox: the deep sky is nearly black and only discrete stars glow. We keep a near-zero floor for exactly this reason — and the background light you see around the hole is genuinely lensed by the curvature.

Only the background sky changes, not the physics: lensing, disk, shadow and photon ring are identical. The normal view is «cinematic»: a near-black sky (Olbers' paradox) with a few discrete stars, so attention stays on the black hole. «Real sky» uses a genuine all-sky astronomical photo — NASA's «Deep Star Maps 2020» (Goddard SVS, public domain), built from real star catalogs (Gaia/Tycho) with the Milky Way's diffuse band — projected onto the celestial sphere and sampled with the ALREADY-lensed ray direction: so the real Milky Way is genuinely smeared and curved around the shadow. If the photo isn't available it falls back to a structured procedural sky (dust-laned band, bulge, nebulae). In return it loads a few-MB image.

Ray-tracing Kerr geodesics is heavy: each pixel integrates the photon's path, and the fragment shader runs on the phone's GPU. If it stutters, lower the Quality preset (Medium or Low): it reduces the integration steps and the resolution.

They are not invented. Each ring of the disk has a temperature (from the relativistic thin-disk flux via Stefan–Boltzmann) and we show the true blackbody colour of that temperature (Planckian locus → sRGB), then shifted by gravitational redshift and relativistic Doppler.

No. Only stars much heavier than the Sun (more than ~20 times its mass) collapse into black holes. In about 5 billion years the Sun will become a red giant and then a white dwarf — never a black hole.

Yes. The closer you get to the horizon, the slower time runs relative to a distant observer: gravitational time dilation, the effect made famous by Interstellar. The «Orbits» demo shows the measured speed and redshift tied to this, and at the horizon an object appears to «freeze».

General relativity predicts a central singularity, where curvature becomes infinite and the theory itself breaks down. What actually happens there would need a quantum theory of gravity we do not yet have. This simulation models only the outside of the horizon, where the physics is well-defined and testable.

Stephen Hawking predicted that black holes are not entirely black: they emit a very faint thermal radiation and slowly evaporate. For real black holes it is tiny (they are colder than the cosmic background), but over immense timescales it would make them disappear. The «Real scale» panel computes the Hawking temperature and evaporation time for the mass you choose.