417-megapixel Andromeda galaxy panorama took over a decade to make | Hacker News Reader

417-megapixel Andromeda galaxy panorama took over a decade to make

petapixel.com

297 pointswallflower98 comments

Mandatory recommendation of the Gigapixels of Andromeda [4K] [1] video/version. Especially with this particular song(!), as the 8K version [2] has a different one which doesn't really give the chills... Although, 60fps makes the image much better. Maybe combine the song from [1] with the video from [2]...

The source picture is the 1.5 gigapixels version (69.536 x 22.230 pixels).

Fun fact: watching the video on certain TV's makes them flicker wildly. Probably because they struggle with many dots in motion. On a monitor it works flawlessly.

[1] https://www.youtube.com/watch?v=udAL48P5NJU

[2] https://www.youtube.com/watch?v=D9bNqBeAtC8

It’s probably motion smoothing causing it to flicker on certain TVs. Why motion smoothing exists on TVs is another question. I guess some people want everything to look like a cheap soap opera.

It's not so simple. Background panning on modern TVs can look very juttery/flickery with motion-smoothing completely off. OLEDs can turn on and off very quickly, and 24 frames a second really isn't that many, so you end up seeing each frame rather distinctly instead of the more smoothed out and less instant frame updates you got on older TVs.

I've found the lowest motion-smoothing setting makes watching stuff like this far more enjoyable while avoiding the awful soap-opera effect you get from higher settings.

It felt awful to admit to myself since I hated on motion-smoothing for so long, but I simply cannot not see the 24 frames in pretty much all scenes where the camera is panning and background has to move a lot.

These kind of videos are great but what makes them even better is background music selection. Videos like this (as well as astronomy related channels) are the main reason why my music taste shifted from mainstream to indie music makers like Carbon Based Lifeforms, Sync24, Pete Namlook, Solar Fields, Stellardrone, Cell.. to name few.

Have loved some of those for years and never heard of others. Thanks for the recommendations and it sort of speaks to how special and unique they are (o;

If you can't reliably stream 4k or 8k 60fps, or even if you could and are stymied by Youtube compression (your 'smart TV' may be turning this incompressible video into meaningless snow), grab the source from the author as a download over Bittorent:

https://daveachuk.gumroad.com/

AT 10 trillion kms = 1 light year, 10 quadrillion km = 1000 light years, 10 quintillion kms = 1 million light years. Since Andromeda galaxy is 2.5 million light years away, you are looking at an object 25 quintillion kms away. If that doesnt ring a bill, that is 25000 quadrillion kms away, 25 million trillion kms away! , 25 billion billion kms away!!! Simply put if you travelled 1 billion kms that would be 0.000000004% of the way to reach Andromeda galaxy. Imagine that!

Recreational mathematics, fun.

And if we could see it with the naked eye it would appear six times bigger than a full moon.

It’s about the width of three finders on an outstretched hand.

It is visible to the naked eye in good locations and conditions, actually.

And yet, Andromeda is the closest major galaxy to the Milky Way: next door neighbors in the small patch of the Universe known as the Local Group.

the further ones such as the cartwheel galaxy = 500 million light years = 5000 quintillion kms and hoag's object 600 million light years at 6000 quintillion kms are unfathomable to like ever reach. Given a quintillion has 1 followed by 18 zeroes, to cover 0.1% of a quintillion, we gotta travel 1 quadrillion kms, how the hell do we ever do that

Wormholes!

> If that doesnt ring a bill

That'll be 23 Zettameters, please.

And even more: Every object you can see with the naked eye in the sky is within our milky way, except for the Andromeda galaxy and a handful of other galayies. If you know exactly where to look, you can see it as a little cloud with the naked eye.

https://en.wikipedia.org/wiki/List_of_galaxies#Naked-eye_gal...

That led me down a bit of a rabbit hole and I found this

https://en.wikipedia.org/wiki/List_of_galaxy_groups_and_clus...

So there are actually a few galaxies visible with the naked eye outside of our local group even. However, all are within our local supercluster.

Star Trek ships travel (very roughly) 1000x the speed of light at Warp 9. So it would take the Enterprise (or Voyager) 2500 years to reach Andromeda at max speed.

that is still unfathomably slow i am afraid, looking at farther galaxies like the cartwheel galaxy at 500 million light years or hoag s object at 612 million light years, this star trek ship would take 50000 years at the minimum to reach there

Funny how come the original post by NASA was seen be fewer people here on HN:

https://news.ycombinator.com/item?id=42731686

I guess it has to do with the time and day of posting something, on how much it will be upvoted and hopefully rise out of the new posts pit :-)

Thanks for this, couldn't access this posts link because of this:

> We value your privacy

> With your permission we and our partners may use precise geolocation data and identification through device scanning. You may click to consent to our and our 1464 partners’ processing as described above.

No reject all button.

Weird, I only have to click:

1. Reject all

2. Legitimate Interest

3. Object all

4. Save & Exit

But then I'm in EU

I don't have to click anything. uBlock nuked 26 elements, EFF PrivacyBadger reports 14 trackers blocked, and Ghostery NeverConsent blocked the InMobi cookie popup entirely.

It's largely random, but yeah that was a miss. Probably too late to fix now. Sorry :(

I've never seen Andromeda, even when I was in a deep dark sky area and could remember where to look. This [1] NASA picture of the day shows, enhanced to be visible, just how big it actually is in the night sky.

[1] https://apod.nasa.gov/apod/ap201125.html

I'm still having a hard time: It's difficult to tell the camera FOV.

I looked it up: it's 1/4 of a degree. The sun and moon are each 1/2 a degree.

So, Andromeda appears half (diameter) of the sun or moon.

That is not accurate -- the Andromeda Galaxy is over 3 degrees wide. About 6 full moons.

Thanks for the correction!

And getting bigger, as it is moving toward us on a collision course.

Did you ask an AI?

The first time I imaged Andromeda was a total accident. I was shooting a timelapse with a wide angle, and I kept seeing this fuzzy patch in the images. It was one of those “oh neat” moments.

> Since Andromeda is so large and relatively close, although still 2.5 million light-years away

Considering those photons are 2.5 millions old, I'd say it took significantly more than a decade

(I'll see myself out)

From the photons' perspective it took no time at all!

No luck catching them electrons, then? It’s just the one electron, actually.

https://en.wikipedia.org/wiki/One-electron_universe

Is this a Hot Fuzz reference? Love that movie

So from the photon's PoV, it is carrying state instantaneously across the entire universe. Weird.

If the photon had a mind, this would make it go crazy.

At first I thought that was camera noise when I zoomed in, and was wondering why it's so noisy... then realised that's all the stars. Insane.

Same. Every few years it seems I have to refresh certain astronomy facts in my mind. The obvious looking stars are in foreground, in our galaxy. The "noise" is what a trillion stars looks like at 2.5 million light years. Dear brain, remember this please.

The article mentions that the galaxy is a big target for Hubble to image but doesn’t specify exactly how large: Andromeda stretches 3° across our sky compared to our Moon’s apparent diameter of 0.5°. It would be quite a sight to behold if it were bright enough to see by naked eye.

Is it not (assume no light pollution)

It is visible if you know where to look, but still not exactly the easiest thing to discern since it is just a fuzzy patch to the naked eye.

I've seen it very clearly with my naked eyes in Argentina. One prerequisite though - it was 5500m high on Aconcagua camp (we camped also in 6000m but sky was the last thing I was concerned about there). Any decent mountain ie in European alps will give you massively starry night if sky is clear, but that place was a notch above.

Those few unfortunates who died on 8000m peaks by ie getting lost and had the chance to see a starry night must have seen quite a spectacle.

It's not so much the altitude that gives you a clear view (although that also helps), but the lack of nearby light pollution and lack of clouds. The European Alps are lit up like a Christmas tree and in most places will not give a spectacular view of the night sky. Western Europe is just a bit too highly populated for good easy astronomy.

Rural Queensland Australia was amazing and my best views of the milky way. It is something else. There is a lot out there!

The southern hemisphere has more bright interesting stuff (including the centre of the milky way) than the northern hemisphere.

The easiest way to see it is with averted vision (looking slightly away from Andromeda, so that it falls on the most sensitive part of your retina). Unfortunately, though, with the naked eye, it just looks like a faint smudge.

Zoomable: https://esahubble.org/images/heic1502a/zoomable/

Why is it incomplete? I can't find an explanation on this or the NASA site linked from there; its an awfully big chunk missing nearly to the center

Speculating based on the sections:

https://archive.stsci.edu/hlsp/phat

https://archive.stsci.edu/hlsp/phast

It took a decade to get that much. Getting the rest, assuming they aren't able to shrink the chunks, would require a project equal in duration and scope. The JWST can probably capture it with similar resolution in a fraction of the time. If the JWST didn't exist, they'd probably go for another project to fill in the gaps, but it doesn't make sense when a much better telescope is available.

The JWST and Hubble are two totally different telescopes in that Hubble is mainly visible light spectrum where JWST is totally IR spectrum. They can both take an image of the exact same object and the images will look different. They cannot use JWST to fill in the gaps of a Hubble project.

I didn't say anything about JWST filling in the gaps. I said it wouldn't make sense to do another project with Hubble to finish the image when it would take another decade. They can get a scientifically useful image from the missing spots in less time with the new telescope.

Don't need another decade. A few months to get some of the holes at the bottom.

There's enough image there, just select the black and hit content aware fill.

Anybody else fantasize about what life could be like there? Do you think some civilization there has taken a similar photo of our own galaxy?

Yes, immediately. Think of the stories all those lives could tell. It’s awe inspiring and humbling.

417 megapixels image is really nice but it also something people on earth can at least approach. I did a 28 megapixel Andromeda galaxy shot myself without even resorting to mosaics:

https://www.astrobin.com/hqrhe0/

With a few changes I could have easily got somewhere around 100 megapixels if I did a 2x2 mosaic without my reducer on the scope.

There are better cameras and scopes (planewave scopes for example) that getting to 400 megapixel is totally achievable for a high end mature astrophotographer.

Astronomical seeing severely limits the efficacy of even multi-million dollar telescopes. The size of the pixels in this image is ~0.2 arcseconds, which is far below typical seeing limits even in excellent conditions.

Excellent seeing on earth is typically 0.4 arcseconds, so close. https://en.wikipedia.org/wiki/Astronomical_seeing

My setup gives me around 1.92 arc seconds for a point diameter.

And you can do tricks such as lucky imaging or active optics (depending on your budget) to further improve the resulting resolution. Lucky imaging is tricky on something as dim as Andromeda, but has been shown to be just about possible.

I haven't seen lucky imaging used on dim objects by anyone I know. I personally do not have a large enough aperture to collect enough light for that. But I've used it on bright planets before via AutoStakkert[1]: https://www.astrobin.com/full/06dzki/0/

[1] https://www.autostakkert.com

Lucky imaging was always a tool for use on planets and the moon. Anything bright.

It's hard to do dim objects because there's less for the software to inspect in each frame to determine the luckiness and distortion, but you can maybe use fortuitous bright stars in the frame to index off. You also need to collect a huge number of images to get any sort of signal to noise ratio. This video is an example of the technique actually used on a dim object, though the results were fairly modest because of murky British skies.

https://www.youtube.com/watch?v=5s9xbZ5G-wk

That 'excellent seeing' is for sites that are chosen for telescopes. Typical seeing from sea level is much worse:

- https://www.ing.iac.es/astronomy/development/hap/dimm.html

- https://www.researchgate.net/figure/Fig-C1-Seeing-distributi...

- https://www.mdpi.com/2072-4292/15/9/2225

Around 1-2 arcseconds is 'good' seeing in your backyard. There is a good reason for telescopes to be on some of the tallest mountains on Earth.

I think it would help, if they selected a region where to 100 to 1000AU the density was similar to ours, and showed the night sky from a position orbiting a star of comparable size, and then somewhere of significantly higher density.

I always assume that the levels of radiation closer to the galactic core are worse but so would insolation in the wider sense: the star field would be dense enough to illuminate more than the milky way does, for us surely?

I wonder if software can be put to it in order to plot every single star.

I wonder if there were a way to eventually get a stereo image — depth data for each point of light so that we can map Andromeda in three dimensions.

Distance in space is difficult because a dim star could just be a smaller dim star that's close, or a larger bright star that's far away. We have to use a lot of clever tricks (standard candles paired with parallax calculations [1] in particular) just to get distances estimates. We can kind of do this for stars that are very close (relatively speaking), but at the scale of the universe - even measuring the distance to a distant galaxy has a substantial uncertainty factor, let alone the stars within that galaxy!

[1] - https://en.wikipedia.org/wiki/Cosmic_distance_ladder

No way to use the change in relative distance between the stars of Andromeda itself?

I guess we would need two telescopes very far apart in order to get stereo images?

Wouldn't any two points converge at the horizon, considering the distances involved, otherwise?

Andromeda is a whole other beast, but I understand we've done something similar for bodies in our Milky Way using two ends of Earth's trip around the sun.

Or use the same telescope by taking images 6 months apart so the earth is in opposite position in its orbit

Yeah, Andromeda is way too far away. The Gaia mission was custom-built to do exactly this kind of "stereo" imaging of stars in our own galaxy, but it's still not sensitive enough to cover the entire Milky Way.

And in related news the ESA's Gaia spatial mapping L2 Satellite is just about to run out of Gas [1] after some amazing work much of which can be seen in VR for the real 3D treatment [2]

https://www.manchestereveningnews.co.uk/news/uk-news/esa-spa...

https://zah.uni-heidelberg.de/gaia/outreach/gaiasky

I love seeing photos like this, and assumed that you would need the Hubble to capture pictures of space this good. But I have been following Andrew McCarthy [1] on twitter and am amazed at how good of pictures you can get with a backyard telescope.

[1] https://x.com/ajamesmccarthy/status/1876658931717832938

I invite you to look at it from from the radical mystic standpoint.

Imagine you're in the ancient times looking at your cell phone and seeing this distant galaxy in so much detail. Don't think about the technical details and the long time period facts and all that. Just do it as though you're looking through a special magical lens that allows you to see it, because you are!

Obviously very cool, but I'm also curious what it can be used for, if any resident astrophysicists are reading this and can chime in...

When I see these pictures, as impressive as they are, I wonder what kind of scientific facts can the astronomers extract from them.

With my layman's eyes, it is very clear that there is a dense galactic center and dust clouds between the galaxy and us. However, What else can an expert eye tell from the picture?

Plenty. What looks like generic dots to us laymen are actually different types of stars. With more resolution, astronomers can more accurately categorize the different stars in the galaxy. The ratios of different stars tell us things about the universe.

With better high resolution images recently we've also been able to see confirmation of Gravitational Lensing [0] which reveal superstructures in space-time that affect the images we see. (i.e. with lower resolution we might've assumed we're seeing multiple distinct stars, with better resolution we understand that it's the same)

For example, we just discovered the first "Einstein Zig Zag". [1]

Ultimately, understanding the gravitational structure of space-time is the key to understanding dark matter which is arguably the biggest mystery of the universe today (besides dark energy).

[0] https://en.wikipedia.org/wiki/Gravitational_lens

[1] https://www.space.com/first-einstein-zig-zag-jwst

Would love a version with the largest rectangle possible without the black/missing bits

It'd be interesting to see an overlay showing images by their age.

Usually that galaxy is moving over 4,000 miles per hour. With this photo evidence, we can now issue them a speeding ticket, we've got 'em dead to rights

The speed limit is about 186,000 miles/second.

Not for astronomical objects. Their speed may be boosted by the expansion of space.

Yeah but good luck seeing them.

Well, let's imagine an object that has some Platonic motion of 0.2c directly away from us, but it's so far away that spatial expansion boosts its velocity to 1.1c.

At time 0 it emits a photon toward us. One year later, it's 1.1 light years farther away. The amount of newly-created distance between us is 0.9 light years. Some of that was created behind the photon, but I'll ignore that and just assume that all of the new space got in the photon's way.

Over that one year, the photon has traveled 1 light year towards us, and only 0.9 light years of that distance was new. So it's gotten closer to us by (a little more than) 0.1 light years. This rate will accelerate over time.

The object is still visible at the point where it emitted that photon, even though its velocity exceeds the speed of light.

I think it will ultimately become invisible, though.

depends on how you define speed, I suppose

No, it's you moving 4 000 miles per hour relative to the Andromeda galaxy.

Serious question: Is this what Hubble originally captured? Or unlike bodies in our solar system, maybe with a galaxy compositing isn't necessary?

What do you mean by “originally”? The cameras on Hubble are black and white, and this is a color image.

It's ~never just the raw output of a sensor. Even your personal camera isn't much like that tbh. It's just a question of how much processing is done, and with an image like this, the answer is: quite a bit.

too bad it's not optimized to view without loading the entire thing to memory...

Yeah we need a Google maps like view.

Seems like there should be a library that does this fairly efficiently for ultra large images.

Not sure how many 417-megapixel images are out there where this would be something someone works on "over a weekend". We just need the right person to come along at the right time to think it would be a cool thing to do just because.

I did some quick searching, this library seems like the best bet: https://github.com/openseadragon/openseadragon

I hacked together a quick demo using that library on this image: https://andromeda.bsprague.com/

Beautiful, thank you!

Thanks heaps!

It's cool zooming in and seeing the moire patterns shimmer into view

where is FLIF when it's needed the most

Reply on news.ycombinator.com