Six thousand light-years away in the constellation Vulpecula, a cluster of young stars is dismantling the cloud that made it.

The cluster is NGC 6823, and it's roughly two million years old, which in stellar terms means it's barely finished forming. Its brightest members are massive blue O-type stars, the kind that burn through their fuel in a few million years and pour out ultraviolet radiation the entire time. That radiation is what you're seeing in this image. It ionizes the surrounding hydrogen, making it glow, and it pushes the gas outward, hollowing out the cavity at the center of the frame where the cluster sits.

The wider glow is Sh2-86, a large, faint emission region catalogued by Stewart Sharpless in the 1950s. The cluster sits at its heart and serves as the core of the Vulpecula OB1 association, a loose grouping of young, massive stars that all formed from the same molecular complex. NGC 6823 spans about fifty light-years, and Sh2-86 extends well beyond the edges of this field.

The pillar

The dark column just right of center is the most interesting thing here. Catalogued as NGC 6820, it's a denser knot of the original molecular cloud that has held out against the erosion happening all around it.

The process is called photoevaporation. Ultraviolet light from the cluster strips gas off the surface of any cloud it touches, but denser regions resist longer. What survives is a column pointing back toward the radiation source, shielded behind its own dense head — the same mechanism that produced the famous pillars in the Eagle Nebula, running here on a smaller stage and a shorter clock. Look at the bright rim along the edge of the column: that's the ionization front, the boundary where the cluster's radiation is actively eating into the cloud.

These structures are temporary. Given a few hundred thousand more years, the pillar will be gone.

A faint visitor

Sh2-86 isn't the only thing in this frame. Two small patches of oxygen emission sit at opposite edges of it, and neither has anything to do with the cluster. Both are planetary nebulae — the glowing shells of gas thrown off by dying sun-like stars at the ends of their lives, seen by chance in the same direction and at completely unrelated distances.

The first is Ou 3, about a minute and a half of arc across. The catalogue prefix belongs to Nicolas Outters, a French amateur astronomer, and the same series includes Ou 4, the Giant Squid Nebula. Objects like these get found in deep amateur narrowband exposures rather than by professional surveys, because they emit almost entirely in a single oxygen line and are effectively invisible in ordinary imaging. Ou 3 now carries the formal designation PN G059.2+01.0, having been confirmed as a genuine planetary nebula after its amateur discovery.

The second, PN G059.7-01.0, is the more striking of the pair — a compact blue disc roughly 45 arcseconds across, sitting sharply against a curtain of red hydrogen filaments. It has no common name, only its catalogue number.

The contrast between them and their surroundings is worth sitting with. Sh2-86 is a nursery, lit by stars a couple of million years old that are still tearing apart the cloud that made them. These two are the opposite end of the same story: quiet shells of gas cast off by stars that have finished burning, drifting outward and slowly fading. Beginning and ending, in the same frame, separated only by which direction you happen to be looking.

Still forming

The cluster's center is two million years old, but the outer regions contain objects younger still. Star formation in Sh2-86 didn't happen all at once — it's ongoing, and the pillars are part of the reason why. Compression at an ionization front can trigger collapse inside the cloud even as the outside is being stripped away, so the same radiation that destroys the pillar may be triggering the formation of stars inside it.

There's also a supernova remnant nearby, G59.5+0.1, and some research suggests its expanding shock front may have helped trigger the burst of formation that produced NGC 6823 in the first place. That connection isn't settled, but it fits a pattern seen elsewhere: massive stars die, their shockwaves compress nearby clouds, and the next generation lights up.

How this image was made

This is 45.67 hours of data, gathered over fourteen nights between 22 June and 9 July 2026 from a remote observatory in Texas under Bortle 1 skies, about as dark as the continental United States gets.

Rather than photographing the nebula in ordinary color, it was imaged through three narrowband filters, each isolating light from a single element. Hydrogen-alpha at 656 nm traces the bulk of the ionized gas. Doubly-ionized oxygen at 501 nm requires more energy to excite, so it appears only close to the hottest stars, which is why the blue-white cavity in this image marks exactly where the cluster's massive stars are. Singly-ionized sulfur at 672 nm tends to sit in cooler, denser regions and along the shocked rims of pillars.

Each filter is 3 nanometers wide, narrow enough to reject essentially everything except the one emission line it's tuned to. Fifteen hours went into each.

Those three monochrome images are then mapped to red, green and blue to build the color you see. The palette is a choice rather than a photograph, but it isn't arbitrary: because each color channel represents a different element at a different energy, the color in the final image carries real physical information. Where the frame runs gold, sulfur and hydrogen dominate. Where it turns blue-white, oxygen is being excited by nearby massive stars. The transitions between them trace the structure of the ionization front.

The stars are the exception. Narrowband filters give stars strange, unnatural colors, so a short set of ordinary red, green and blue exposures was taken separately and used purely for the star field — colour-calibrated against a catalogue of known stellar spectra so the blues, whites, yellows and oranges you see are close to what your eye would register if it were sensitive enough.

The result sits somewhere between a photograph and a diagram. Every structure in it is real and every color means something — it just isn't the view you'd get through an eyepiece.