Wow amazing work! The scale of this project reminds me of another photographer's five year project of creating a 2.5 gigapixel photo of Orion: https://orion2020v5b.spaceforeverybody.com/
I immediately wondered how he managed to seamlessly mosaic together what is surely hundreds of images and thousands of subs. Here are the details: https://space4everybody.com/processing-details/
Nothing too fancy (in astrophotography relative terms). A decent refractor and mono camera (such parameters have come down in price by 2-3 times, ASI2600MM is comparable) and most importantly, a solid pier/mount.
267 Individual panels (including reshoots)
12,816 Individual light frames
65,000 x 35,000 in it’s entirety
5 States traversed to image
640.8 Imaging hours
500+ Editing hours
2.5 Gigapixel Image
1.6 Pixel scale
5 Years
1 Mosaic
Unfortunately they don't seem to have a link to the final "full-res" image. I'd love to put that with my collection of other massive-res images. My screensaver pans between random spots on massive images in a folder.
My libraries are always made of images I've found online. I've lost two libraries over time so I'm on my third one now.
I haven't added to my library in a while, but I might spend a few hours looking for some more images this weekend.
The current stats for my library:
At the moment it's 1.47gb for 75 images. I'm not sure where I have copied each image from, but I know some are either from ESO or NASA's image of the day, except that I find the original non-browser friendly images and download those. If I can get an image I like larger than my 5k screen, I'll add it to my library.
My average image size is 7156x5412.
My largest image is 324,000,000 pixels (18,000x18,000), pretty sure it's the Orion Nebula.
My smallest image is 2,958,400 pixels (so only 1720x1720), however any image that is surrounded by black (like a planet with no stars around may be cropped from time to time). Since my screensaver puts black around any images smaller than my 5k screen, it still works out. I'm pretty sure the smallest image is an image of Uranus from the Voyager that flew past it and I wanted to make sure I had an image of at least every planet in the Solar system and that was the best image of it I could find.
Frankly, since we only have ourselves to sample, we have no idea how large the survivorship bias is. It doesn't matter if there's googol of galaxies or just one galaxy, we have zero direct information on how easily life arises. Putting it another way, observing a googol of galaxies ought to pretty much not at all increase our confidence that other life exists, like how observing a 200% increase in a speed doesn't tell you anything when you don't know what the absolute speed was to start with.
However, one thing this argument often ignores is that the Earth is 4.5 billion years old. Life is 3.5 billion years old. The Earth was hot molten rock up until about 20 million years before life began.
So we went from Zero to Life in 20 million.
Once created it was so successful that it took over nearly every part of the Earth in a multitude of evolving forms.
It seems improbable life makes it's appearance so improbably early on Earth that perhaps the only explanation is that life isn't so improbable after all. Perhaps it's even highly likely give the physics of the universe.
We know that simple chemistry plus lots of energy can create more complicated chemistry. We have evidence that some protein molecules can be self-replicating under the right conditions.
Given billions of parallel experiments over billions of years, I think it is vanishingly unlikely that exactly one has progressed to the point we would recognize as self-organizing.
But that's faith, not proof. I'll defend it as the most reasonable and non-exceptional version of the story, but I won't claim it as absolute truth.
The path to modern astronomy went like this: cultures that lasted a long time (Multiple hundreds of years) typically developed observatories that defined the yearly calendar as well as dates of religious observances. Many times, their observations got very accurate and they could predict various things, although surprising stuff still happened. Eventually, civilizations lasted through multiple solar eclipse cycles and noticed they were periodic (keeping great records for hundreds of years and maintaing a class of people who can work with them is nontrivial). This eventually led to people being able to predict nearly all planet's motions with surprising accuracy (see the antikythera mechanism).
So I think it's had a huge influence ,and one that ultimately led to a scientific understanding of the cosmos. At this point, JPL can predict the location of nearly every major solar object for decades into the future and the main challenege is determining whether object movement is truly chaotic.
And in the modern world, we are blind to it because of light pollution. It's sad. One of the most amazing experiences ever, a truly dark, clear sky at night, is only accessible if you travel off-continent.
I couldn't believe what the sky looked like at night when I moved to Australia from the UK. And I was in a relatively suburban setting then. I regularly go camping now maybe every 1-2 months out into the outback. And the night sky, without exaggeration and using the words here sensibly, literally, never, gets old. In my 19 years in the UK I never saw anything like it.
Edit: Admittedly I think the constant cloud coverage in the UK was more of a hindrance than the light pollution here I lived. Or at least it was so fucking cloudy so fucking often, I had no idea how bad the light pollution was.
I remember when I first moved here and looked up to the sky from my back yard and I could see Orions Belt. Now I'm less than a novice, I only recognised it from Men in Black (don't judge me). But that feeling I had in that moment stuck with me.
And I get such a ridiculous glee when I can drive 1-2 hours from Brisbane city and see the most ridiculous night skys I've ever seen. Keep travelling west and it just gets better and better. I can't wait to visit on of the Dark Night Sky locations.
Ah man, damn it! I spent 10 months in Australia, but most of it was during lockdown, and I didn't have a car. I only got 1 week of travel and didn't see much of the stars.
My best experience was in Fiji and New Zealand, but I imagine that the outback could be even better.
I've gone out to Mojave on super clear nights. Vegas is easily visible, almost like a sunrise. The Great Smoky Mountains look good on paper (light pollution map), but in practice they are quite clowdy / hazy, so viewing isn't good. I imagine northern Canada could also be quite good.
Man that sucks! Were you mostly in Melbourne during this time? They were in lock down for a long time but most of the rest of the country was only 1-2 months of severe lock down (hard limits on leaving the suburb and then later, a hard 50KM limit from your place of residence).
I was in Canberra, which wasn't technically locked down for all that long, so I could probably have gone out more. It was generally very demotivating being there during the pandemic, so I mostly just farted around the lake on my bike. It's a big lake, though!
Cheers and happy trails. I'll return to Aus sometime!
For me the real shock was Lord Howe Island, which is 600km from the mainland and surrounded by nothing but open ocean. At night, the entire island has only a few dozens lights on for marine safety, powered by a diesel generator.
On a clear, moonless night the sky is just unbelievable. You can see several galaxies with the naked eye, which is not something I had experienced before!
Agreed. I very recently was thinking about how humans up to a few hundred years ago must have seen the most breath taking night skies. Imagine a native American standing in a clearing at night 2000 years ago, with zero light pollution, looking at the night sky and just wondering what it is all about.
Even w/ a clear desert sky hundreds of miles from the nearest city (e.g. middle of Nevada), the sky looks nothing like these pictures. Astrophotography uses exposure times many multiples--and usually several orders of magnitude--longer than the human eye.
Can you see the Milkyway? No, not really; not without prompting. If you have no idea what to look for, the disc is just a barely imperceptible, slight increase in otherwise meager (relative to astrophotographs) star density. I've never seen it, nor other constellations, because I don't really know what to look for, though I've spent many hours staring up at the sky from various places, including extremely remote desert and mountain locales--Western U.S., Northern Mexico, Mongolia, Ecuador, Chile, etc.
Of course, there's no way to understand the awe and wonderment somebody lacking our knowledge, experience, and expectations may have experienced. Especially if they had never seen an astrophotograph.
EDIT: Actually, now that I think about it, I wonder if astrology, navigation, and astronomy might have developed more slowly if the human eye had greater fidelity. With only several thousand stars visible (much fewer readily discernible, especially if you're not an adolescent w/ fresh corneas), the motion of stars probably stands out more. Imagine trying to recognize and track the planets on a background as dense as an astrophotograph. Where to start? Would you even recognize the relative motion? People had much fewer distractions during nighttime back then, but it still seems substantially more daunting.
This is not my experience. I have seen the Milky Way very clearly from a few locations in the US. None are particularly remote -- all are just a few hours drive from the large cities I've lived in. Winter months are clearer than summer.
> I have seen the Milky Way very clearly from a few locations in the US.
I agree. heck, I remember stargazing in rural New Mexico in the 70's, the moonless night sky was luminous to the dark-adapted eye. The Milky Way was a river of stars, bright enough to read by (well, distinguish the words on the page, I don't remember spending much time reading...)
Completely agree, I've seen the Milky Way just by traveling to a city called Eskilstuna, some 200km away from Stockholm, in the middle of the forest and away from settlements you could see the Milky Way much brighter than just a faint streak of light.
Yeah I agree. Not only have I seen the milky way, but I've seen andromeda with my naked eye from the right place on the right night, within a day's drive from very populated cities.
+1. I’ve not seen it for 10 years (since I’ve lived in the United States) but growing up in New Zealand, looking up at the Milky Way with your naked eye, it’s definitely clear the increased density of stars. Doesn’t look anything like photos, but very beautiful.
Maybe its clouds, or air quality, or general deteriorating climate, about 30 years ago, my village in plains of Punjab, we used to see lots of stars, constellations, & planets along with satellites at night time. Maybe absence of tv, mobile, culture of sleeping on roof in summers, made us more aware of stars. Full moon nights were like brightly lit. No moon nights were amazingly dark. We could see shadows in moon light. Things were just like day, but in dark blue grey color. We always saw the north pole star, few other groups, & watched milky way moving from side to side in night. Did not know its name, we call it SkyRiver in our language (Galaxy). For reference, now I am in northern california small town.
That is exactly the feeling I get the couple of times when I am out hiking hiking and see the most spectacular winter night sky. The feeling of being in the middle of nowhere and seeing the dark sky with an uncountable number of stars is indescribable. I believe it is a great loss of modern society and I really hope we are able to do something about it, we could do just fine without most of the street lighting we have today.
> My observatory is located in the very center of the city Oulu in Finland. Due to massive Light Pollution I mainly do ...
Well I guess most of us live in filthy cesspools of light. Seriously I wonder what humanity misses out on. I think we would have been a much better species if only we could see our place in the universe every night.
Would have been a much better species? Bright lights are a very recent part of our species’ history. It’s not like humanity had a reputation for being particularly civilized before the invention of electric light.
Depends on what you mean by "civilized", but I take your point. I do wonder, however, if there isn't a correlation between the increase of light pollution and the trend away from reverence for gods. In other words, does the ability to see the grandeur of the heavens on a daily basis influence us to believe in greater beings?
On a side note: there is a great documentary on light pollution called “Saving the Dark” (it's on Youtube). This really has a big impact on us and the environment.
"Depends on what you mean by "civilized", but I take your point. I do wonder, however, if there isn't a correlation between the increase of light pollution and the trend away from reverence for gods. In other words, does the ability to see the grandeur of the heavens on a daily basis influence us to believe in greater beings?"
I wouldn't say it's exactly subtle to say "depends on what you mean by 'civilized'" and then immediately talk about the "trend away from reverence for gods". If you want to clarify, then clarify, but as it stands your comment is quite clear.
Oh, I see where you're coming from, but I was actually just being pedantic about what it means to be "civilized", since human civilization goes back thousands of years before electric light. "Civilized" in historical contexts means to be in an advanced stage of social and cultural development, that there is a lot of structure to people's lives, even if that structure employed violence in ways we find abhorrent.
The other meaning of "civilized", the one you seem to have intended, is more along the lines of being well-mannered and less violent, which is why I take your point.
I do some astrophotography from time to time and it has been a great way to get away from it all to get some perspective while stargazing.
On occasion I venture out into dark sky areas in the US with friends, and they are always surprised that it is possible to see the milky way with the naked eye(or just so many stars at all).
There was this one time where in some city in the US, cannot remember which, there was a city wide power outage, and some people were freaking out because they've never seen a non-polluted sky before.
Do I understand correctly, that all colours on these pictures always are purely artificial? That is, all visible light we can see through any telescope is always white, and only some minor differences in the spectrum, invisible to us, are used to detect what elements there are, and then semi-artistically colorize the picture based on that data?
Astrophotograpers often add artificial colours on top of (or instead of) the actual colours.
Plus, much of what you see from astronomers (although not from this case) is taken in non-optical parts of the electromagnetic spectrum (e.g. radio waves or x-rays), so those pictures are necessarily artificially coloured.
Real colors is always a particular definition, because everyone of us see the colors differently, for the astrophotography is another different thing because the colors are made by frequencies of the light spectrum. Try to read this for more info: http://dslr-astrophotography.com/right-colors-astrophotos/
> Conclusion on color calibration:
"There are multiple definitions of what it means to have ‘correct colors’ in your astrophotos. Therefor this always will be a matter of personal preference and everyone will have to figure out what their own vision on the matter is and pick a method of calibrating the colors accordingly"
They are synthesized, from the article "Image in mapped colors from the light emitted by an ionized elements, hydrogen = green, sulfur = red and oxygen = blue."
Depends on the filter used. I am assuming the author is using one of these hydrogens (probably Hα) and O-III and S-II. Each of these black-and-white images with this filter are used as the data for a single RGB channel in an image. Their wavelengths are, and would like like to the naked eye [1]:
Hα (656 nm) bright red
Hβ (486 nm) deep blue
O-III (496 nm and 501 nm) greenish blue to green
S-II (672 nm) deeper red
The color mapping the author uses is called the Hubble Space Telescope pallet, where hydrogen = green, sulfur = red, and oxygen = blue.
Hard to tell without being closer :) Nebulae might not even emit that strongly in the spectrum that we humans can see. A lot of these pictures are taken with special narrowband filters that isolate wavelengths for specific components, such as Oxygen or Hydrogen. Then these narrowband images are combined and assigned to the R, G, B channels. A popular palette is the "Hubble Palette" because, well, that is what the Hubble uses.
Depends on what you mean by real colors. The light is so faint, that human eyes cannot make it out. However, long exposure photography with a CMOS sensor will bring out colors without any filters. Are those the real colors? There are tons of examples of these types of shots. Search for The World At Night (TWAN) for some great examples
To the naked eye from earth, it looks black and white, and you wouldn't be able to see that much detail. What they "really" are is more complicated, as indicated by other responses.
What do I have to do to get the full-size image? The 7000x1300 pixel image is pretty nice, but I was thinking of making a 1440dpi print on 4 sheets of A3+.
I had the urge to print as well. I doubt he'll be interested in giving away this masterpiece, atleast for an affordable price. I'll mail and ask; he might offer a print service maybe.
If you can get this printed on metal, it would look amazing. My wife is IT for an astronomy dept and they have metal prints all over the place that are just stunning!
Related question: If someone wanted to print something like this for themselves, how would they go about it? Say, I have an image or photo and I wanted to make a high quality print suitable for framing and hanging on, say, my office wall. Is this an 'Ask HN' question?
I've done high quality large prints from https://www.bayphoto.com/. They did a great job and have different archival paper to choose from. Got the print from there and used a local framing shop to get it professionally framed.
Do you want the cheapest way to do it or the highest quality, or something in between?
If you have an image that fits in less than 20x30", CostCo or another mass printing service can do it for $10. They won't do any adjustment, so the colors may be off a little.
They'll print on canvas to 30x40" for $120. That's only 150dpi; it doesn't look great up close, but is pretty good from 3+ feet away. Again, colors are what they are.
High quality printing is a thing that you talk about with a professional printer (the person, not the machine) for each job. Prices generally start at $100 or so for smallish prints, and can go up to $1500 or more for large, permanent-art quality work.
More like the highest quality, but probably price sensitive which means 'somewhere in between.' I've thought about taking a test image to Costco to see how it comes out on metal, but figured that there might be higher quality options. Thanks for your examples, looking to see what others post.
Remember that when you work with a good image printer, you're paying for their expertise and time -- and you won't have to re-print failures, because they'll consult with you to get it right.
One friend of mine spent several hundred dollars at an online printer in $40 increments trying to get it right.
I mean, you can always segment it into several smaller prints and then stitch them together (carefully!), but I'd rather the author get some cut of it than me doing it myself at Walgreen's or Staples.
There are roughly 10^23 stars in the observable universe. We really have no idea how likely intelligent life is to evolve. It could be 1/10^10 per star, in which case intelligent life would be a fairly typical thing (dozens in the milky way), or it could be just 1/10^100, in which case we could be the only ones. Unlikelihoods can compound very quickly.
Thats not hard to believe once you understand how unprobable life seems to be and especially in the very long term. The universe is a very violent place and cosmic events act as great filters all the time.
We know that life only emerges in very specific conditions, that most planets do not have. We also know that even in very stable Earth, we have had billions and billions of species that developed and disappeared and we now only have a single specie with technological capabilities - and for a very short time at that, all things considered.
If the universe is 'rich in life', most of it is probably micro-organisms that get wiped out every now and then by drastic changes on the planetary scale.
What are the chances that, from an outside perspective, that we are just one variety of life form on the planet?
Sure, they can make distinctions about which variety of us (e.g. humans, not dogs) have the most intelligence or something--but perhaps from an outside perspective, we all started from the same spontaneous event or whatever, and we're just many, many generations of that same lifeform.
In other words, maybe we're just being "racist," sort-of.
As an aside, is this a named theory or something? I'd like to read about it.
You might well be undercounting there, according to a 2016 study[0]:
> Earth could contain nearly 1 trillion species, with only one-thousandth of 1 percent now identified, according to a study from biologists. The estimate is based on the intersection of large datasets and universal scaling laws.
Agreed. It’s a minuscule probability per planet multiplied by a vast number of planets and it’s hard to say which factor carries greater weight. The fact that we haven’t detected any signals from space also sets a limit on the likelihood/duration of intelligent life in our area of the universe.
It occurred to me the other day that there's barely any science fiction that even occurs outside our own galaxy (Star Wars being the exception that proves the rule). And if it does it's usually something from Andromeda right 'next door'.
Even Star Trek only leaves the galaxy maybe one time, and that requires the help of Q.
There's that saying... too young to explore the world, too old to explore the universe.
When our children's children learn to extend their life, or figure out some sort of suspended animation, or pload their brain... what wonders they will see.
If we could travel appreciably close to the speed of light, the journey could be close to instantaneous for the traveler. (Depending obviously on how close to light speed we’re talking about.)
Indeed - something often misunderstood as always recorded from the perspective of a third party observer.
Quick follow up question. If you did depart and travelled at near-ligth speed to another galaxy, arrived there in a few seconds and then looked back at the earth - would it appear as only a few seconds have passed for the earthlings back home as well?
By the same token, if you are actively watching another galaxy as you travel at near-light speed towards it - will you be watching that galaxy "age" at an extremely fast pace throughout the journey?
I think the downside is that for earth, time is passing by at 'normal' speed. So as soon as you get up to speed you're leaving behind, possibly the entire human race.
Weird to think that too if aliens show up. Their entire civilisation is probably already gone (unless it's VERY stable over hundreds of thousands of years).
If you traveled one light year at the speed of light, when you looked back you would see light from right after you left. But since you were traveling at the speed of light, your relativistic time for the journey would be zero.
So the light you observed would be proportional to your subjective time.
It might make more sense to take off in your near-light-speed ship, do a few laps of the galaxy, then return home for a thousand-years-in-the-future ship upgrade, then head on to another galaxy.
Wait, what? In what universe is there "barely any science fiction" outside our galaxy?
Lensman. The entire Revelation Space series. Beyond the Aquila Rift. Player of Games. Algebraist. Sirens of Titan. Perry Rhodan.
And that's off the top of my head. There's quite a bit of SF (both good and trash, and I think the above list has both :) happening outside our galaxy.
Fantastic series and probably my favorite writer, but because it mostly sticks within known physics nearly all of it takes palce within about 20 light years of earth with the exception of Galactic North and maybe a few other side stories. Earth doesn't get much of a mention, but the main character is from Mars.
I guess I was thinking more about ones that have intergalactic travel. Do any of these actually have travel between galaxies or are they just set in an alt galaxy? Sorry, I din't really explain that thinking very clearly.
Beautiful, and respect! Seems many photographers collaborating could achieve similarly amazing results a lot faster, is that a thing and if so can I join?
I can kind of see what you mean, but I wonder if this is actually a result of the subject, not processing. The night sky is absolutely peppered with pinpoints of light far smaller than individual image pixels (resolvable stars) which are quite unlike typical terrestrial scenes.
The effect seems to be more noticeable in the areas filled with gas.
I'm not saying that stars aren't pinpoints or sometimes are just single pixles, but just it overall looks like some post-sharpening has definitely been applied aggressively to my eye.
This is also a much smaller area of the sky. The higher acuity of the image might actually be more representative of what the telescope actually sees. The smooth darkness you are used to seeing might be similar to how some people prefer music where the treble rolls off (like tube amps or sennheizer 650s headphones) vs people preferring a more linear freq response.
well, the orion image you linked doesn't look like it's been over-sharpened, in contrast to the OP image. I'm not saying there's a way to tell just by eyeballing it, but that's what it smells like.
This is true, but it's also true that making faint features visible (like the dark, dusty regions) often requires the use of local contrast, sharpening, and masks.
Also, in the case of this specific image, I think it's important to note just how deep the images are - they're processed for high resolution in the individual components, so the mosaic has an absolutely huge resolution. When downscaled to look at the whole thing zoomed out, that might give some of the impression of over-sharpness.
This is amazing! When taking long exposures like this over a significant amount of time, does anyone know how astronomers account for stellar parallax and/or the sun's orbit in the milky way? Are these factors too small to affect long exposure photos? And if they are too small to affect a long exposure photo, at what point does it become an issue?
Stellar parallax isn't a big effect. It wasn't even observable until the 1800s. Even the Hubble cannot measure parallax past about 10,000 lightyears, which is about 10% of the size of the Milky Way. For earth bound telescopes at this scale, it's totally negligible.
On a camera in manual mode if you increase the exposure time, you must compensate the increased amount of light coming to the sensor by decreasing the aperture or the ISO. Can anybody please explain, how is it possible to have such a long exposure without overexposing the image?
Note that these images are made from many exposures that are then processed and combined, and the total exposure time is given. The individual exposure times were probably long, but I'm guessing on the order of minutes, not hours. (Actually, if you photograph stars with an exposure time that is too long, then you get star trails in the image.)
Trails appear very quickly, at focal lengths of around 20mm (typical dslr kit lens) they are obvious in 20-30 seconds, with a 100mm lens (start of telephoto range) they are obvious in 3-5 seconds. This of course depends on the res of your camera (I shoot with an 8MP and a 24MP).
To get rid of the trails you need a tracking mount. The better the mount the smoother and longer it tracks. Astrophotography is expensive...
If you want longer exposures than your lowest ISO can handle you can also use neutral density filters, which go over the lens and reduce the amount of light hitting the sensor.
By using very low ISOs. You want to use the lowest ISO possible because the lower the ISO the less noise in the picture, and noise in the night sky can be very noticeable since you are photographing a ton of tiny dots against a black background.
I'm so glad we're used to clouds on Earth and that our human brain evolved to intuitively understand them because as it happens the cosmos are just like clouds. It'd suck if we looked at them and were lost in their perplexity.
This got me thinking about treating a surface with Black 3.0 paint and printing astrophotography images on top of that, but I wouldn't know how to get a large format printer to safely print on top of the treated material.
Love the use of the apparent size of the moon for scale. Really gives you an awesome sense of what you’re not seeing in the sky in cities with your own eyes.
12 years might be too short of a time frame for objects that are moving about 1/1000 of the speed of light.
It would be interesting to see a comparison of 50 or 100 years. The center should be rotating faster or slower?
