Poetic licence

A perfect photon rocket: all the fuel turns into light, emitted in a collimated beam to the rear, with no loss at all. No such engine exists; there are no tanks, radiators or reserves. Even matter–antimatter annihilation loses a good part of its energy to products that cannot be collimated, and any such loss multiplies the departure mass many times over.

The ship crosses the interstellar medium as if it did not touch it. At this speed, the gas between the stars reaches the bow as a beam of enormously energetic protons — at the turnaround, 3.4 TeV each, and the flux grows with the square of γ. The panel calculates what hits the shield and assumes a shield able to withstand it; nothing wears out, nobody gets sick. Semyonov (2009, 2020) estimates that the incoming gas is the most intense radiation of a relativistic trip and requires shielding that this model does not have.

The model uses a single density for the whole trip, 0.1 atom per cm³ (that of the Local Bubble, where the Sun is); near the Sun and inside the cavity of M16 the gas is denser.

One constant g for seventeen years of ship time, an instantaneous turnaround at the halfway point and, while braking, bow engines that keep the cockpit facing the destination. A real rocket would turn the whole ship around, with minutes without gravity. There is no life support, no food and no consequence of seventeen years on board.

The coastal base is fictional, with facilities and terrain created for the scene. The nearby coastline and river are procedural, not a survey of any real place; the latitude of the base is the declination of M16 (−13.9°) and the visual longitude is fixed, so that the base stays beneath the ship. The clouds combine a historical NASA image with generated detail and do not represent the weather on the day of departure. The sky darkens between 50 and 100 km as a visual choice.

The ship climbs vertically, already pointed at M16 — as if the nebula were at the base's zenith at the moment of departure. It was not: at noon on 19 October 2026, M16 stood about 30° above the horizon at that longitude.

The ship leaves the ground in a straight line, with 1.1 g of thrust — 0.1 g net in the first minutes — and switches to a constant 1 g at 311 km of altitude, with no atmosphere, drag, heating, Earth rotation, parking orbit or solar gravity. No real vehicle launches like this. Departing from a 400 km orbit would change 41 minutes of ship time and 0.008% of the fuel.

The route is a straight line from Earth to a chosen point in front of the Pillars. Nothing along the way deflects the ship or changes its speed: there are no gravity assists, and the planets of the flybys are scenery, not mass. The ship arrives at rest relative to Earth, and the model gives M16 no velocity at all — the Sun and the cluster orbit with the Galaxy and move relative to each other, but the target stays where the photographs placed it.

The simulation adopts 7,000 light-years to the center of M16, the classic textbook value, and every dimension of the nebula is a measured angle times that distance. Parallaxes from the Gaia mission (Kuhn et al. 2019) place the NGC 6611 cluster at 1,740 ± 130 parsecs, about 5,700 ± 400 light-years.

The plane of the Pillars lies 8 light-years short of the adopted center of M16 (a design value), and the ship stops 7 light-years short of that plane, with zero velocity and the fuel running out exactly. The point was chosen by design: from 6.9 light-years off the nearest face, Pillars 8.6 light-years tall do not fit whole in the windshield.

The surface is NASA's Blue Marble map, a January 2004 composite, with no weather, seasons or cities lighting up. The Earth shown does not rotate: meridian and terminator stay as they were at the moment of departure, to keep the base beneath the ship — in six hours, by the time it leaves the frame, the real Earth would have turned a quarter of the way around.

For signal delay, Earth stays fixed at the departure point: its orbit, eight light-minutes in radius, is left out of the calculation. There is no actual communication: nobody transmits, nobody answers, and the ship radio is fiction. The ship calendar uses years of 365.25 days and months of one twelfth of a year, in Brasília time.

The sky has 299,858 cataloged stars — 109,400 from HYG and 190,458 from Celestia's Gaia EDR3/Hipparcos/Tycho distribution. The real sky has billions; the faint and distant ones are missing, and near M16, in the dense disk, so are those that aberration would bring into the window. Each star shines as a black body at its own temperature, estimated from its color (HYG) or its spectral class (add-on): there are no spectral lines, and a cool star seen blueshifted looks slightly different from the real one. The 190,000 add-on stars stay still, with no proper motion. There is no twinkling, zodiacal light or gravitational lensing; Mercury and Venus exist only as markers.

The window behaves like a camera with a fixed exposure, chosen once for the whole trip and the same in the rear camera. At rest it shows stars down to magnitude 8, about five times more sensitive than the naked eye under a dark sky; it never adapts, never dazzles, never flares. A monitor cannot reproduce the range of intensities that would reach a real window either: the cosmic background at the turnaround would saturate any screen.

Dust extinction comes from a smooth galactic disk model (0.55 magnitude per thousand light-years in the plane, scale height of 330 light-years), not from an observed map of each cloud: queries to the Bayestar three-dimensional map went unanswered. The diffuse glow of the Milky Way is a disk model with a 5,000 K spectrum; the Magellanic Clouds, Andromeda and M33 are clouds of random particles with the approximate shape.

Of the cosmic microwave background, only the monopole and the kinematic dipole are included: no map of the primordial anisotropies (of order 10⁻⁵), no cosmological evolution during the trip and no scattering by the medium. The visible patch is only calculated once the peak temperature exceeds 300 K. On screen the same fixed exposure applies: at the turnaround, with 19,652 K at the bow and 217 W/m², the center saturates to white.

The shape of the Pillars and of the surrounding nebula comes from two public photographs — ESO0926a, a wide field, and Hubble heic1501a, of the Pillars — registered to the measured positions of the stars; the depth comes from a reference mesh for 3D printing (NASA/STScI), scaled to the silhouette, and from assumptions: the gas emission in a sheet 15 light-years behind the plane of the Pillars, the cluster stars in a common plane 3.2 light-years behind it, the outreach enhancement of the photos undone by a power law with exponent 2.4. Spectroscopy constrains velocities, densities and the relative geometry of the Pillars, but not the volume adopted here.

The lighting uses eight stellar systems of NGC 6611 — eight of the roughly five hundred stars in the cluster. The gas emission uses the ratios of fifteen optical lines observed by MUSE at one point in the nebula, not a map; the absolute brightness is pinned to a single emission measure, a plausible value on the high side. The brightness scale, the scattering and the depth of the layers are assumptions, and there is no ultraviolet or infrared.

The simulation shows the Pillars of the photographs — light that left them some 7,000 years before it was photographed. The ship arrives some 7,000 years after departure, in Earth time, and sees light emitted seven years earlier: between the light in the photographs and the light that reaches the ship lie some 14,000 years, and the real Pillars of that day are not those of the photographs. At the evaporation rate measured with MUSE (McLeod et al. 2015: some 70 solar masses of gas per million years, a lifetime of some three million years), fourteen thousand years change the overall shape little; the 2007 suggestion that a supernova had already toppled them remains unconfirmed.

The Moon, Mars, Ceres, Jupiter, Saturn, Uranus, Neptune and Pluto were placed where the ship crosses the radius of each orbit, on alternating sides, at 1.5 radii from the center, to create close flybys. No real alignment would do this: on a straight line to M16 the ship would not pass near any of them. The light comes from a fixed direction, not from the Sun, so the phase you see is not the real one, and there is no Doppler color shift on the planets. Some maps have fills for unobserved regions, Ceres uses a fictional texture, Jupiter's moons sit in fixed positions, Charon is about 20% closer to Pluto than in reality and the asteroid belt is twenty rocks.

Saturn's rings are synthetic material with the C, B and A bands, the Cassini Division and the Encke Gap close to their real radii. At Uranus, the seven bands, their widths and the gap the ship flies through are cinematic choices: the real rings are 1 to 10 km wide (ε, up to 100 km), and the bands in the scene are hundreds of kilometers wide.

A scenic stretch of 24 hours of ship time begins 2,000 AU from the Sun, on day 90: nine groups of three icy bodies 35 to 125 km across, with a lunar texture, passing as close as 4,500 km to the ship. Sizes, positions, lighting and frequency are choices made for the scene, and the bodies do not get the Doppler color and brightness that, at a quarter of the speed of light, would already nearly triple their brightness. The real cloud is inferred, never observed; its bodies larger than 1 km are tens of AU apart, and the chance of passing within 10,000 km of one of them in 24 hours is essentially zero. The "wall" of filaments at 300 light-years, at the edge of the Local Bubble, is another artistic representation: the real bubble is hot gas, invisible in the optical, with no sharp edge.

The cockpit is a single image, generated by an image tool, with computed light layers on top. A real cockpit is three-dimensional, vibrates, has sound and weight; this one has none of the four. The panels are instruments invented for this trip, with units that change scale so the needle never stops.

There is no eye adaptation and no individual visual response: the window is a fixed camera, and the colors of the starbow and of the cosmic background are those that camera would record, not necessarily what a person would see after minutes in the dark or with the panel lit. Reflections, glass transmission and the monitor's photometric calibration are left out.

The two sensor towers at the bow are drawing, not engineering: no structural calculation holds the trusses up at 1 g. The four-pointed star around each beacon imitates the diffraction of a camera with four struts and has the same size in all four colors; in real optics the spikes grow with wavelength, and the blue one would come out about 30% smaller than the red. The blue blinks 50% brighter than the other colors so that it looks equal to the eyes, which have few sensors for blue.

The Borealis Horizon ship, the builder's plate, the coastal base and the department that speaks on the radio are fiction. Borealis Dynamics is not: it is the company behind the experiment.

The news strip and the voices that take turns on it — Borealis, for mission management; Brenda, from the Compliance and Asset Integrity Office; Gary, from Operational Support and Contingencies — are characters, and each of the 2,415 lines was written for this trip. The texts follow what happens on the voyage, but they predict nothing about the real world and quote no real person.