How Perigee calculates apparent size

Nothing in Perigee is scaled by eye. Every object in the scene is placed at a fixed render distance and drawn at whatever radius reproduces its calculated angular diameter, so the size you see is a consequence of the geometry rather than a decision anybody made.

How is apparent size calculated?

Apparent size is the angle an object subtends at the eye. For an object of diameter d at distance r, the angular diameter θ is:

θ = 2 · arctan( d / 2r )

That is the whole calculation. It is applied identically to the Moon, to Saturn and to the Andromeda Galaxy, and it is the only thing that decides how much sky an object fills. If the fixed render distance in the scene ever changes, the rendered radius follows automatically — no object is ever scaled by hand to compensate.

How does Perigee compare sizes to the full Moon?

A figure in degrees is hard to picture, so Perigee restates it in full Moons: the object's angular diameter divided by the Moon's angular diameter at its familiar distance of 384,400 km. Saying something spans thirty-three Moons is the same statement as giving its angle, in units an eye can hold.

How does Perigee calculate light travel time?

The selected distance divided by the exact speed of light in vacuum. It is a statement about the distance, not about the rendering: the scene is not delayed by that amount.

Which constants does Perigee use?

What does Perigee simulate, and what does it only draw?

Perigee labels every claim it makes with one of three boundaries, and shows that label in plain language beside the claim.

How are motion and surface detail presented?

The Moon, stars and Andromeda keep a fixed visible face. The catalogue sky stays fixed while the camera pans. Motion pauses while the tab is hidden, and reduced motion freezes ambient animation.

Jupiter rotates around its own axis using an approximate 9.9-hour period; Saturn uses the Cassini ring-seismology estimate of 10h 33m 38s. Mars uses a 24.6229-hour sidereal period; Neptune uses the approximate 16.11-hour Voyager magnetic-coordinate period. At automatic 120× time-lapse, Jupiter and Saturn take about five minutes per turn, Mars about 12.3 minutes, and Neptune about 8.1 minutes. This speed changes only planetary rotation; the sky, weather and observer's landscape do not advance with it. Captures freeze one pose across every tile, and reduced motion pauses the rotation.

The Moon rotates once per orbit, keeping substantially the same hemisphere toward Earth. Its sidereal rotation takes about 27.32 days; the roughly 29.53-day phase cycle is different. Perigee holds an Earth-facing view with fixed sunlight, without automatic spin, orbital motion or libration. Terrain shading uses measured elevation; earthshine is approximate.

Planetary imagery and reconstruction

The Moon uses a globe with LROC reflectance imagery and LOLA elevation from NASA's CGI Moon Kit. Dark maria, pale highlands and rays come from the map; measured slopes provide terrain lighting at their source resolution. Colour and contrast are adjusted for display. Polar reflectance is coarser, and the source grids include interpolation and small filled gaps. The nearly full phase and near-side orientation are fixed. Apparent size follows the calculated diameter and distance. There is no atmospheric glow.

Mars uses an oblate, rotating 3D globe. Mars Express HRSC colour is combined with Viking imagery and Mars Global Surveyor TES brightness; MOLA elevations supply measured terrain. The dated mosaics retain some photographed shadows and interpolated gaps. Polar caps use Viking imagery. Colour balance, reference orientation and a thin atmospheric haze are authored; the sources do not describe simultaneous weather. No geographic features are generated, and terrain heights are not exaggerated. Apparent size remains calculated.

Jupiter uses an oblate 3D globe with Cassini's December 2000 global cloud map (NASA/JPL/Space Science Institute). The near-infrared/blue composite approximates natural colour; its polar regions are hazy and less resolved. Contrast, sunlight and pole orientation are authored choices, while apparent size follows the calculated diameter and distance. The map rotates as one surface; this is neither current weather nor a simulation of differential cloud motion.

Neptune uses an oblate globe combining NASA/JPL Voyager cloud structure from August 1989 with explicitly modelled fine bands and atmospheric haze. Full-disc PIA01492 and a small PIA00058 close-up provide dated cloud morphology; their registration and lighting removal are approximate. Unobserved regions, polar caps and fine band continuity are reconstruction, not observations. The pale palette is informed by the 2024 natural-colour study; the source's enhanced blue is not treated as calibrated colour. Sunlight, haze and pole orientation are authored, while apparent size remains calculated. One rotating map cannot reproduce differential winds or evolving weather.

Saturn uses an oblate globe with a multi-epoch reconstruction from Cassini's observed cloud structure, Hubble's broader haze bands and a Cassini polar view. Missing coverage is interpolated; colour and lighting are authored, not calibrated measurements. Separate equatorial rings combine Cassini colour with Voyager optical depth, including the main divisions and mutual planet/ring shadows. The globe spins beneath a stable ring plane. Indirect light is approximate, and weak perspective preserves the approved full-ring framing. This is a cinematic reconstruction, not simultaneous weather or a current spacecraft view. Scene and saved captures use the same frozen pose and material.

Sources and credits: NASA/GSFC/Arizona State University, the near side of the Moon; NASA's Scientific Visualization Studio, LROC and LOLA teams; ESA/DLR/FU Berlin, G. G. Michael and the HRSC team; NASA/USGS/ASU Viking merged colour; NASA/JPL/ASU, Philip Christensen and the TES team; NASA MOLA and the PDS Geosciences Node; USGS Astrogeology and NASA Ames; NASA, ESA, Amy Simon and the Hubble OPAL team; NASA Voyager UVS and the PDS Ring-Moon Systems Node; Irwin and colleagues, 2024. Ring colour comes from Cassini PIA11142, NASA/JPL/Space Science Institute. Map processing, hashes and reconstruction limits are recorded in the repository's asset provenance.

Andromeda imagery and reconstruction

Andromeda uses an AI-generated artistic portrait informed by NASA Hubble PHAT/PHAST and visible-light references. Its core, dust lanes and companions are interpreted artwork, not telescope measurements or calibrated naked-eye colour. The same colour-preserving compositor used for the planetary and stellar portraits retains its original lighting in the live scene and captures.

Distance changes scale the optical major axis. The image retains its authored inclination and orientation; it does not reconstruct internal parallax, stellar orbits, a collision or a freely explorable galaxy volume. Small points within the portrait are part of the artwork; the surrounding sky is rendered separately from catalogue stars.

Visual references: NASA Hubble PHAT+PHAST mosaic and NASA SVS optical view (NOAO, AURA/NSF). The portrait is original Perigee artwork; these reference photographs are not redistributed in it.

The background sky uses data from the European Space Agency (ESA) mission Gaia, processed by the Gaia Data Processing and Analysis Consortium (DPAC). Funding for DPAC has been provided by national institutions, in particular the institutions participating in the Gaia Multilateral Agreement.

Stars, sky and exposure

Stellar surfaces are stylized interpretations, not photographic maps. Betelgeuse uses broad orange regions and bright patches informed by convection research; Sirius has pale granulation and Rigel has finer blue-white mottling. Surface contrast and colour differences are deliberately enhanced for visual detail. Radio and infrared false colours are not treated as visible surface colours. Betelgeuse, Sirius A and Rigel use approved synthetic globes with no claimed bulk rotation period or measured surface geography. Rigel's original blue-white mottling is artistic reconstruction; it is not solar granulation, a Betelgeuse convection model or a current photospheric pattern. Spectra and interferometric wind diagnostics do not provide a visible global surface map. Automatic Rigel motion remains disabled. The Sun uses synthetic photospheric granulation and invented spot groups, informed by visible-light HMI research and near-infrared DKIST granulation. Neither magnetic-field colours nor ultraviolet corona imagery are surface texture. Its warm ivory colour and optical glow are authored choices, and its spot pattern does not represent current solar activity. Automatic solar motion remains disabled.

The Yale bright-star catalogue and Gaia DR3 supply positions and colours. Diffuse Milky Way light comes from the CDS Gaia G-flux survey, with the Gaia counterparts of drawn points subtracted before downsampling. It is a neutral broad-band reconstruction, not a colour photograph. The selected star replaces its catalogue point. Gaia positions refer to J2016.0; Yale retains J2000.0.

Solar System reference directions use 1 January 2016 at 00:00 UTC. The local sky is rotated to the selected target at a representative latitude of 38.78° N; the landscape heading and time of night are staged. This is not a live sky chart. Horizon extinction and irregular point-star scintillation share one model, with stronger faint-light suppression above the city. Resolved bodies receive atmospheric attenuation without point-star twinkling.

Catalogue stars use normalized compact cores, with faint optical wings around brighter sources. Pixel-area integration keeps their light stable as resolution changes. Selected stellar objects retain a continuous point-to-disc flux budget. Extreme brightness ratios are deliberately compressed for display. One AgX transform follows linear-light composition; photographic landscapes receive no extra object-coloured wash. These exposure choices do not reproduce unaided-eye visibility or physically illuminate the photographed ground.

Occasional meteors are illustrative atmospheric events. Most are brief, narrow streaks; rare brighter events leave a fading train. Their paths stay fixed in the sky while you look around. Timing, brightness and train drift are authored, not a meteor-shower forecast. Landscape artwork also contains static stars.

Sky survey: T. Boch / CDS, CNRS / Université de Strasbourg, ESA / Gaia / DPAC. The Gaia-derived catalogue and diffuse map are available under the Open Database License 1.0. Stellar reconstruction reference: López Ariste et al. (2022).

Rendering and capture quality

Rendering automatically requests native display detail within device limits and adapts if it becomes slow. Capture automatically chooses a resolution suited to your device. High-resolution PNGs render the sky afresh, preserving the current frame's aspect ratio without interface text or captions. At 16:9, 4K and 8K captures are 3840 × 2160 and 7680 × 4320 pixels.

A larger export can reveal more available source detail, but cannot create observations beyond the original maps. Broad optical bloom retains the full-frame response of the live view. Export freezes motion, combines four samples per pixel and omits animated grain. Devices with limited memory may automatically use a smaller export or the current rendered resolution.

What Perigee is not

Perigee is not a planetarium catalogue and not a physics simulator. It does not model heat, radiation, gravity or tides, it does not place objects on real orbits, and it does not claim that any of these skies could occur. Surface lighting, atmospheric treatment and the impossible-proximity effects are authored visualizations sitting on top of geometry that is exact.

Sources

Every figure on this site traces to one of these, each opened and dated at review.

Explore the full Perigee sky

Created by Aliaksei Mazheika