The article and its title are translated; processing notes and some crop captions are still in German.
M 31
M31 — our neighbor galaxy, its two companions, and a ring of glowing hydrogen
Andromeda is the nearest large galaxy and, at magnitude 3.4, the brightest thing in the sky that does not belong to the Milky Way. Some 2.5 million light-years lie between us and M 31. The light in this image was already on its way before there were people on Earth. And still the galaxy is no distant speck: its faint disk spans about six times the width of the full Moon — the eye usually catches only the bright core.
A small counterpoint to the famous superlative: M 31 is not reliably the most distant object visible without optics. M 33 sits farther out at roughly 2.76 million light-years and can also be glimpsed with the naked eye under exceptionally dark skies. For both galaxies the problem is not how much light arrives, but how thinly it is spread.
New in this version: the hydrogen. On top of the August broadband image sits a dedicated Hα layer built from 26 dual-band exposures across three September nights — starless, continuum-subtracted, and laid over the color image in Photoshop with Screen. What used to be a pink tint now breaks into individual objects: the great star-forming ring turns out to be a chain of dozens of H II complexes, the southwestern arm carries a string of its own, and even at the very edge of the disk one knot still glows. The most prominent ones are marked in the overlay with their catalog names — mostly after Walter Baade and Halton Arp, who measured 688 emission nebulae in M 31 on Palomar plates in 1964 (BA 1-…), plus the catalog published by Pellet and colleagues in 1978. Three of the marked complexes contain or touch a supernova remnant.
The field. The image measures 3.330 × 2.212° astrometrically, at 2.2196 arcseconds per pixel. At the distance of M 31 that is roughly 145,000 × 96,000 light-years; one pixel covers about 27 light-years. At that scale dust lanes, blue stellar associations, and the red H II knots become structure rather than mere hue.
Why the spiral arms look like ellipses. We see the disk inclined by about 77°, so nearly edge-on. The spiral therefore appears not as an open pattern but as narrow, nested ellipses. It is also why the dust lanes are so sharply drawn: the line of sight runs obliquely through the disk and stacks a great deal of dust along the same path. The strongest ellipse is not simply one arm but the star-forming ring at roughly 10 kiloparsecs, or 33,000 light-years, in radius — the most active zone of the galaxy. That is exactly where the Hα chain sits.
The nucleus has three parts. At the center sits a black hole of about 140 million solar masses, some thirty times the mass of Sagittarius A* in the Milky Way. Hubble resolved its surroundings into the components P1, P2 and P3:
- P3 is a disk of more than 400 blue stars immediately around the black hole, only about one light-year across.
- P1 and P2 are not two clusters but the two visible density peaks of a single eccentric ring of old red stars.
The open problem: star formation that close to a black hole ought to be suppressed by tidal forces. Why a young blue stellar disk exists there anyway is still not fully explained. In this photograph the whole structure of course stays below the resolution limit. One light-year in M 31 spans roughly 0.08 arcseconds; even a hundred hours of exposure would not turn that bright point into a resolved P1/P2/P3 structure.
Two companions, two very different puzzles. M32 and M110 both sit in the field. M32 is unusually compact and dense; it may be the stripped core of a formerly larger galaxy. M110 is far more diffuse and contains, of all things, dust and young stars — hardly what one expects in a classical dwarf elliptical. Both were probably shaped by their interaction with M 31, but in opposite ways. Incidentally, both stay dark in the Hα layer: their reddish tone comes from old stars, not from ionized gas.
NGC 206 and the outer disk. Southwest of the nucleus lies NGC 206, one of the largest OB associations in the Local Group. Its blue stars sit in a cleared gap; the H II complexes of the southwestern arm — Pellet 162, BA 1-318, BA 1-457, and the large complex BA 1-363 — lie all around it. In the northeastern arm the overlay marks the chain from BA 1-577 to BA 1-245, and the supernova remnant beside BA 1-212: birthplaces and remains of the same short lives of massive stars.
Is the collision even coming? M 31 is moving toward us, and the merger with the Milky Way was long treated as near certain. A simulation study published in 2025, using Gaia data and revised galaxy masses, puts it at only about 50 % probability within the next ten billion years. In those models M 33 raises the chance of a collision while the Large Magellanic Cloud deflects the system sideways. Even for our most famous neighbor galaxy, the future is less settled than popular accounts have long claimed.
Where in the sky
The rectangle is this image's real field of view, with its measured rotation — not a symbol. Constellation figures and star positions come from SIMBAD.
3 entries sit within 27° of each other here — click a field to jump to it.
Crops
Nucleus & disk
Nucleus and disk — 77° of inclination, a triple nucleus below the resolution limit, and the inner ring
1.06° × 0.88°·00h 42m 46.58s +41° 15′ 19.9″
A field of roughly 41,800 × 32,400 light-years (57.5′ × 44.6′), centered almost exactly on the nucleus; M32 stays in frame at the left edge. Left is south here — north lies to the right in this image.
The bright patch at the center holds the entire triple nucleus: the black hole of 140 million solar masses, the blue stellar disk one light-year across, and the red ring five light-years wide. At 2.22″ per pixel, all of it sits far below a single image point.
What you can see clearly is the geometry of the disk: the 77° inclination turns the spiral arms into nested ellipses, and the dust lanes stand in front of them as sharp dark arcs. Above, the great star-forming ring passes through with its Hα complexes — BA 1-40 and BA 1-67 are marked — and closer in, the Hα layer reveals the inner ring: BA 1-4 at 11.7′ and BA 1-406 at 15′, that is 8,500 to 11,000 light-years from the center. At lower left, on the inner edge of the southwestern arm, sits the complex BA 1-292 with the young clusters KWE 45 through 48.
M110
NGC 205 (M110) — a dwarf elliptical with dust it ought not to have
33.3′ × 25.4′·00h 40m 21.22s +41° 41′ 24.0″
A field of roughly 22,100 × 15,200 light-years (30.4′ × 20.9′), centered on the catalog position of NGC 205. Projected distance from the nucleus of M31: 36.5′, or about 26,600 light-years.
M110 is a dwarf elliptical companion — and therefore, in principle, an object made of old stars with no gas left. That is not what we find: the center of NGC 205 holds dust clouds and young blue stars, pointing to star formation within the last few hundred million years. For a dwarf elliptical that is the exception, not the rule. In the Hα layer M110 nonetheless stays dark: whatever young stars exist there no longer ionize an appreciable amount of gas.
The plausible explanation for the dust lies in the neighborhood itself: tidal interaction with M31 can drive gas inward rather than only stripping it away. The diffuse, soft falloff toward the outside is typical of this galaxy type — no structure has been lost here, there simply is none.
M32
M32 — the most compact galaxy in the neighborhood
18.8′ × 19.2′·00h 42m 45.31s +40° 51′ 52.0″
A field of roughly 11,900 × 12,300 light-years (16.4′ × 16.9′), centered on M32. Projected distance from the nucleus of M31: 24.2′, or about 17,600 light-years — in projection M32 therefore sits almost within the disk.
M32 is the prototype of the compact elliptical: tiny compared with M31, but with a central stellar density that exceeds our Sun's neighborhood by many orders of magnitude. A black hole of a few million solar masses sits in its core.
Where such an object comes from is open. The strongest suggestion: M32 is the leftover core of a larger galaxy whose outer stars M31 stripped away during a close pass. A related hypothesis makes M32 the remnant of the massive galaxy that merged with M31 about two billion years ago. It is also debated whether a passage of M32 through M31's disk helped create the prominent star-forming ring. In the image M32 stays unremarkable — a smooth, highly concentrated ball of light with no internal structure and no Hα emission at all.
NGC 206
NGC 206 and the southwestern arm — a stellar workshop and its neighborhood
51.2′ × 32.7′·00h 40m 05.88s +40° 38′ 51.0″
A field of roughly 34,700 × 18,700 light-years (47.8′ × 25.7′). The crop follows the whole southwestern arm; NGC 206 sits above and right of center, 40.2′ (about 29,200 light-years) southwest of the nucleus, almost exactly on the major axis of the disk.
NGC 206 is neither a galaxy of its own nor a star cluster but an OB association: a loose gathering of several hundred very hot, blue, short-lived stars spanning some 4,000 light-years — the brightest and largest star-forming region in all of M31. Its brightest members are only a few tens of millions of years old. Where it sits is telling: in a gap between two dust arms. The association has used up most of the gas around it and swept the rest away with radiation and stellar winds — which is why it barely glows in Hα itself.
The gas it lacks lies all around: directly above it the complex Pellet 162, beside it BA 1-318 with the young cluster KWE 17, farther in BA 1-457 — and at lower left, where the arm is already thinning, the brightest H II complex of the entire southwestern arm, BA 1-363, around the OB association OB 139. The Hα layer turns the arm into a sequence: spent gas at NGC 206, active star formation a few thousand light-years farther on.
Northeastern arm
The northeastern arm — where the disk runs out, with a chain of star factories and one remnant
52.7′ × 56.5′·00h 45m 49.96s +42° 07′ 07.7″
A field of roughly 36,000 × 35,900 light-years (49.6′ × 49.4′). The crop follows the northeastern outer arm into the region where the bright disk fades out. The dust lanes break into individual filaments and the surface brightness drops quickly.
What was a pale blue arm before the Hα layer is now a chain: BA 1-577 and BA 1-578 at the bottom, then BA 1-212 with the OB association OB 51 and the young cluster BA 1-210 — 27″ away lies the supernova remnant [LL2014] 144, so this is a region where stars are being born and dying at once. Farther out follow BA 1-642, Pellet 930, and at the very edge BA 1-245, 71′ or about 52,000 light-years from the nucleus. The arm forms stars all the way out to where the starlight itself is barely separable from the sky background.
The outer arm is also the most honest part of any galaxy image: its signal sits only slightly above the sky, so every additional hour of exposure pays off here first.
Hα ring
The star-forming ring — the Hα chain of the northeastern arc
1.06° × 0.80°·00h 44m 21.02s +41° 32′ 35.1″
A field of roughly 42,300 × 28,700 light-years (58.2′ × 39.5′); the nucleus sits in the lower left corner and the center of the crop lies 24.5′ northeast of it. This is the crop the Hα layer was made for.
The arc running from lower left to upper right is the star-forming ring at roughly 10 kiloparsecs — in projection the near half of an ellipse. Without narrowband it is a pink seam. With Hα it breaks into individual complexes, and the catalog names from Baade and Arp acquire a face: BA 1-117, the brightest Hα complex in the whole image, a good 2′ or 1,500 light-years across, with the OB association OB 42 beside it; BA 1-67, at whose center sits the supernova remnant candidate BW-61; BA 1-40, with the remnant [LL2014] 89 only 18″ away; BA 1-163 at the eastern end of the arc, 47″ from the remnant BA 1-160, which is also detected in X-rays. At lower right, on the northwestern side of the disk, BA 1-548 lies over a giant molecular cloud — the ring continues there behind the dust.
That supernova remnants sit so close to the H II regions is no coincidence but the principle: the stars that ionize the hydrogen live only a few million years. Where a region has been active long enough, the first ones are already exploding while the next are still shining. Closer to the nucleus, BA 1-4 and BA 1-577/578 mark the inner ring and the start of the northeastern arm respectively.
Southwestern end
The southwestern end — where the disk fades into the dark, one last knot, and a star in front of it
40.8′ × 26.3′·00h 38m 10.97s +40° 18′ 35.0″
A field of roughly 33,000 × 15,600 light-years (45.4′ × 21.4′); the center lies 81′, or about 59,000 light-years, from the nucleus — farther out than any other crop. At the upper right the bright disk ends; to the left a last faint veil of stars dissolves into the sky background.
The red double knot is BA 1-490 with its neighbors BA 1-487 through 1-492: the outermost H II complex at the southwestern end, 77′ from the nucleus. BA 1-490 is also DDB 1-1, an entry in the first optical catalog of supernova remnant candidates in M31. That stars are still forming this far out shows how much farther the disk's gas reaches than its visible starlight.
The bright blue-white star below does not belong to any of it: HD 3431 is an A0 star of the Milky Way about 640 light-years away — some 4,000 times closer than the galaxy it stands in front of. Its light had been traveling for less than a human lifetime, that of the galaxy for 2.5 million years.
Acquisition and processing
Broadband: 114 exposures of 120 s each from the nights of August 20/21 and 22/23, 2026, taken in Friedrichsdorf with a ZWO ASI2600MC Air on a William Optics RedCat 71 (349 mm focal length), gain 100, bin 1, no additional filter, at a steady −10 °C or so: 13,680 s, or 3 h 48 min, of color signal.
Hα: 26 exposures through the dual-band Hα/O III filter from the nights of Aug 31/Sep 1, Sep 1/2 and Sep 2/3, 2026 — three at 240 s and 23 at 300 s, together 7,620 s, or 2 h 07 min. Only the red channel of the calibrated dual-band master was carried forward, registered onto the broadband image, continuum-subtracted against the scaled broadband red, and computed starless. That makes 140 frames and 5 h 55 min in this image altogether. The O III content in the green and blue channels of the dual-band filter was not used.
The two August nights were calibrated with their own darks and flats. The second flat group matters most: it was taken after the camera had been rotated by about 90°. Flats are tied to sensor orientation, dust pattern and optical train, and must not simply be mixed across both orientations. LocalNormalization and PSF signal weighting evened out transparency and background; all lights were then registered onto the new 90°/270° geometry and integrated together. That keeps the field fully covered, without black wedges from an oversized union area.
In the linear master the green channel was first normalized multiplicatively and the large-scale residual gradient then removed subtractively via DBE. This order keeps DBE from having to model a channel-wide color cast as if it were spatial background. Then came the astrometric solution, Gaia-based color calibration, one corrective BlurXTerminator pass, and separate treatment of galaxy and stars. NoiseXTerminator was applied to the linear and starless layers respectively.
The combination was done in Photoshop: the starless Hα layer sits above the finished color image in Screen mode, opacity by eye. Screen brightens only where the Hα layer carries signal — star colors, dust lanes and the yellow core stay as they are in the broadband image, while the H II regions gain saturation and extent. It is a deliberate way of making the emission regions visible, not a photometrically calibrated Hα map.
Astrometry of the published AVIF (5401 × 3587 pixels): 4,388 reference stars, residual 0.423″, scale 2.2196″/pixel, center RA 00h 42m 45.6s, Dec +41° 15′ 30″, field 3.330 × 2.212°, rotation +89.98°. North therefore lies to the right and east is up. Overlays, crops and angular sizes are computed from exactly this solution; the positions of the marked H II complexes were measured as centroids of the Hα excess in the image itself and identified through SIMBAD.