A window into duration
The slider sets how many physical seconds pass during one display second. Every tile receives that same elapsed time. Click a tile to choose a rate at which its declared period or duration takes four display seconds.
Scenes stay active from 0.00035 to 320 cycles per display second, a further tenfold extension at both ends. Above 3.2 cycles/s, a short 1/1200-display-second exposure adds motion trails; above 320 cycles/s, the scene becomes a full-cycle average. Display sampling can make very fast motion appear discontinuous or aliased; “active” does not mean every cycle is individually resolvable. Physical phase advances at the chosen rate, including across slow rendering frames.
The spectrum button tunes to green light. The interface takes on a subtle hue when the wavelength whose cycle lasts four display seconds is within approximately 380–700 nm. This is the same four-second reference used by tile tuning: λ = c × 4 × simulation rate. The hue is illustrative, and does not change any physical rate. NASA: visible light.
Periods, frequencies, transit distances, speeds, half-lives, growth rates and phase are quantitative. Object sizes, camera position, colors and many spatial layouts are illustrative so radically different systems remain legible inside equal tiles.
Growth and transit scenes reset after the stated distance or interval. That reset is a presentation boundary, not a claim that the real process reverses or repeats. Radioactive tiles show expected populations; they do not assign predictable decay times to individual nuclei.
Human growth follows CDC median stature references for boys and girls aged 2–18, interpolated between annual samples; these describe populations, not an individual forecast. Hair uses an active-growth example of 0.35 mm/day. E. coli uses a favourable laboratory doubling time of 20 minutes; cell geometry and division-stage timing are schematic.
The massive-star example uses 5 million years of main-sequence evolution followed by a specified 0.5-million-year late stage for a 25-solar-mass star. Its size and colour illustrate the stages, rather than a calculated stellar track. ATPase shows a stationary α₃β₃ assembly and rotating γ shaft, with schematic dwell proportions. The EUV stream preserves droplet speed and spacing; static rays explain collection and the plasma glow represents a mean envelope, not an invented flash duration.
CO uses smaller oxygen spheres and the conventional C≡O formal bond order of 3 (one σ and two π bonds). Fractional bond indices depend on their computational definition, so 2.5 is not assigned here without a specific model. Benzene's inner ring represents six delocalized π electrons; it does not animate localized electrons orbiting the ring. UT Austin: CO bonding; OpenStax: aromaticity.
The Sun-like track uses approximately 10 Gyr of main-sequence evolution plus a specified 1 Gyr giant phase. White-dwarf cooling is outside its window. The low-mass red-dwarf model spans 14 trillion years of hydrogen burning. Both show schematic surfaces, core cutaways and fuel shading; those visuals are not calculated stellar profiles.
Aircraft, Parker and nerve motion share a 1-metre travel ruler: 4 ms at 250 m/s, 5.236 µs at 191 km/s, and 10 ms at 100 m/s. Aircraft and probe scenes show a marked surface point in an illustrative repeating 1 m close-up. A full aircraft or heat shield extends beyond this view; the small context icons above it are not to scale. Surface details are illustrative, and nerve diameter is enlarged.
The quasar tile times gas at one stated radius around a specified black-hole mass, not the rotation of a whole quasar. The caesium tile animates microwave field phase, not repeated atomic-state jumps. The glaciation tile illustrates a roughly 100 kyr interval, with specified 90 kyr growth and 10 kyr retreat; it is not a reconstructed ice-volume record.
Water and land carbon use mean residence/turnover times in ideal well-mixed reservoirs (τ ≈ 10 days and 23 years). Silicate weathering uses a published approximate 240 kyr feedback timescale. Their curves show exp(−t/τ) over five time constants, then restart a fresh illustrated cohort or perturbation. Diagram arrows explain pathways without assigning invented travel speeds. These are not universal periods for the complete water, carbon or rock cycles, and land-carbon turnover is not the removal time of emitted atmospheric CO₂.
The ISS, Earth-rotation and lunar-orbit panels use one shared sidereal rotation phase (86,164.09053 s per turn). Their orbital periods are independent. Earth remains animated when an orbital path is frozen, and its own motion blur is independent of orbital blur. Continents come from the unmodified Earth day map by Solar System Scope / INOVE, CC BY 4.0, projected onto the sphere with illustrative lighting.
Measured values come from observations or experiments. Defined and specified examples use an explicit wavelength, field, circuit or geometry. Model labels mark numerical or theoretical calculations. The far-future black-hole tiles use the ideal uncharged, nonrotating Schwarzschild-vacuum Hawking lifetime and should not be read as observations.
Drag the logarithmic rate slider or click any tile. The marked rail break jumps across the empty 10²²–10⁷⁰ second range. The − and + keys halve or double time speed; hold Alt for full-decade steps. A toggles continuous logarithmic acceleration. Space pauses, R resets elapsed simulation time, and Escape closes this window.
Each row gives the characteristic time used by the animation, its evidence class and the precise scope of the claim. Links open the supporting institution or paper.