Hacker News new | past | comments | ask | show | jobs | submit login
2.5 Gigapixel Image of the Orion Constellation Took 5 Years to Complete (petapixel.com)
371 points by giuliomagnifico on Oct 24, 2020 | hide | past | favorite | 62 comments



Image itself in high resolution can be found here:

https://orion2020v5b.spaceforeverybody.com/


I was looking for galaxies, and then I turned on annotations. There's a ton of PGC annoations (galaxy catalogue) and it helps you see many of them; however many of them don't seem to be visible at all in the picture (there's too many of those PGC annotations?)

Either way, very cool, that helps getting the eyes up for the galaxies scattered in the picture.


I've just zoomed in randomly on a couple of places and in each frame there's a handful of side on spiral galaxies... it's very cool - take another look?


My brain always thinks of Orion being orientated with Betelgeuse in the top left, it's really weird seeing it 'sideways'


You guys in the Northern hemisphere see it upside down, we call his belt "the pot"


It's funny – at such high fidelity, the glow from all of the distant stars and galaxies make it look like light pollution.


Good observation. I assume that's because it's "overexposed" to bring out the nebulas etc.?

Zooming in all the way, the pure "black" areas between stars also seem to have a fair bit of sensor noise that probably plays a part too.


There are no pure black areas between stars, and that's not (just) sensor noise. It's not very wrong to say that every single pixel in that image holds at least one galaxy. If you collect enough light, you will find something interesting to look at.



Wow!


I love that this is your reaction to that story, as I believe that is the only valid reaction! I also love being reminded that every day someone gets to experience something new that I've already taken as fact.


At some point the sensor noise swamps any signal, given your setup and processing.

So from a philosophical view every pixel holds a galaxy, but from a practical point of view sensor noise obliterates it.

I think it's reasonable to call the area in between visible stars "black" when received light is below the noise floor.


I’m guessing that it is difficult to distinguish the “noise” of sub pixel galaxies from the thermal noise of the sensor.


It’s not exactly correct though: at this resolution - maybe, but at resolution high enough there are gaps between galaxies. The reason for that is that region of space we can observe is finite. If it would be infinite, and between any two objects we would see one more object, all we would is see is uniform white.


Can I download the full scale image from somewhere?


I am wondering about this too. Maybe not the full resolution, but I'd like to stretch this over 3 4K monitors.


How big of a wall would that require to display at 1:1? I have a wall I'm willing to donate to the cause.


Assuming a 50,000^2 pixel image at 600 dpi it’s just under 7 foot by 7 foot. You can likely play around with the DPI to fit most any wall, but I don’t know the aspect ratio.


It's 65k×35k pixels: https://space4everybody.com/about-the-image/

There are some really nice printing processes for images of this magnitude, such as https://www.whitewall.com/us/metal-prints/aluminum-mounted-p... and https://www.whitewall.com/us/acrylic-prints/masterprint (the exact same processes are available from many printing companies; I have no idea how good this one is but it was easy to find).

I have a print on this scale of the center of the Milky Way, and it's amazing: http://sguisard.astrosurf.com/Pagim/GC.html


>http://sguisard.astrosurf.com/Pagim/GC.html

Welcome back 1999. Holy cow. Comments in good humor: 1) No preview image of what you are selling. 2) Several links are broken. One is a Flash based website!?!? Others are just no longer hosted. 3) Owww, my eyes!!! Also, Yay! Tables!

I get it. If it's not broken, don't fix it. However, you are making it difficult for an interested person to make an informed purchase. I am looking exactly for this kind of art to fill up my astronomy room, so this has piqued my interest. It's not like I don't know what the galactic center looks like, but not buying something sight unseen


Yes, it's pretty unfortunate. It was a flash-based pan-and-zoom interface for the image, since it's obviously too big for your browser to handle all at once. Now that flash is gone he does need to redo the work to host it, but that is non−trivial. You can see a very small thumbnail at http://sguisard.astrosurf.com/mini/mini-Guisard_MilkyWay_4_8....

I have the 70"×40" print on my wall right now, and it's worth it.


Wow! That is gorgeous, and not what I had envisioned, yet is exactly as advertised.

I am trying to create a gallery wall with the full image of the milky way galaxy[0], and then have close ups of different DSO printed in details. In my mind, I want to arrange it where each detail is attached via strings to their location within the [0] image CSI style.

[0]https://earthsky.org/upl/2011/08/milky_way_panorama.jpg


Oh, that could be really nice. Could also take a really big wall…


I'll just have be sure to include a plaque that says "All images are NOT to scale"!!


Not all to the same scale, at any rate :)

I just meant the aspect ratio of the full-galaxy picture is very high; you might need a rather long wall. And then a wide room so that visitors can step back and see the whole thing. It could be amazing.


There is a tool called "rasterbator" that can print large images out tile-by-tile "Wall Art Generator" https://rasterbator.net/


True. There are also large-format photograph printers that can print an amazing image four feet tall and as wide as you want: https://en.wikipedia.org/wiki/Durst#Lambda_and_Theta_photogr.... This is what was used to print the image I have.

Somewhat expensive though; best to rent time on one rather than buying your own. https://www.bayinkjet.com/products/durst-lambda-130-used



I'd like to do that, but it would make sense if he reserves it so that he can sell prints


Anyone have a direct link to the full image?


Finally, a picture where I can say “enhance” while zooming in, just like in the movies!


> Where Harbison had previously captured an image of the Nebula on a 10-11 pixel scale, using the QHY 16200 camera would change that to a 1.6-pixel scale of the constellation Orion.

Pixel scale here seems to mean how many arc-seconds a pixel covers. The full moon (or one's thumb at arm's length iirc) substend about 30 arc-minutes (half a degree)


The thing that surprises me is how stationary everything in space seems to be, in order for this to have been possible.

But everything is moving, right? So how could a series of pictures over this long of a time work out to be so clear?


It's an illusion.

The Earth is rotating, but it's easy to compensate for that so that the stars don't turn into streaks. You just mount the telescope and camera on a motorized mount that turns slowly to counteract the rotation of the Earth. That's cheap and easy to do, so everyone does it unless they specifically want to show off the rotation of the Earth.

Every star in that image has some proper motion, usually dozens or hundreds of km/s, and there's no way to counteract that. For any single exposure that's not large enough velocity to turn the stars into a streak, but it is enough for the stars to be in different positions relative to each other from year to year. Since different panels of the image were captured in different years, the overlaps between those panels won't match perfectly. When stitching the panels together, the photographer will probably have edited the photograph so that each star is taken from only one panel. That will prevent any stars from being cut in half or blurred together from both panels, but there could be stars that moved fast enough to have crossed the boundary between the panels and they could show up in the full image twice. Good luck spotting them though.

The Earth is also moving around the Sun, so there is an additional apparent motion, but it's smaller in magnitude to than the proper motion and won't cause any extra problems.

The gas clouds are also moving. Some of them are expanding, some of them are being pushed by stellar winds, etc. Stars can form in dense clouds, and then the stellar wind from the new star pushes all the remaining dust away to create a bubble. You can see these motions over the course of years, so presumably they caused some minor mismatches between panels that were photographed at different times. Since these features are fuzzy, it's not hard to blend neighboring panels together even when they're taken years apart.

https://www.youtube.com/watch?v=ofCooIkIwvQ is a great explanation with good visuals, and it has photographs taken by the New Horizons probe that directly show stellar parallax when compared with photographs taken at the same time from Earth. Even that is only visible on the nearest stars; none of the other stars in those photographs were close enough to move at all.


Camera systems like these use computerized tracking to account for the rotation of the earth. The earth rotates really fast!

The earth is also orbiting the sun pretty fast, but it's my understanding (I'm not an expert) that you don't need to account for error here, since the Orion nebula is so far away. Think about how the sun stays in the same place in the sky even as you're driving fast on the highway. The nebula is unimaginably far away compared to the sun, and relative movement between the earth, the solar system, and Orion are all undetectable at these scales.

I've read that even the Hubble telescope doesn't need to account for parallax error for such far-away objects during long exposures, which can last up to 20 minutes. Maybe someone here on HN can do (or point to) the math for us.


The seeming lack of movement even convinced the Greeks that the Earth is stationary:

https://www.lesswrong.com/posts/noBKApHwFm8aMuvnw/link-the-g...


*some greeks. You can't fool Aristarchus though.


a combination of things, the distances you're looking at are mind-numbingly huge. you might think it's a long way to the chemist but that's just peanuts to space. This means that even though things are moving, and fast, they still look like they're in the same place from this far away. Along with that, the images aren't a single exposure over years, but days at a time at worst. that means that any minor movement can be adjusted for when combining the images for the final result.


The sensor gathering the light moves to compensate for Earth's rotation.


Because the distances involved are truly massive. 5 years is nothing at these scales


What are the small white dots filling the space between stars? Is this a camera artifact, gas/dust or something else?


Do you mean the fainter stars that fill the background?


Everything you see in the picture is either a star, or the glow of interstellar gas/dust. Every resolvable circle is a star.


Looks like sensor noise that cameras have in blackness


The Orion Nebula is such a difficult object to get the inner detail. At its heart are four stars known as the Trapezium, but you rarely ever see them as they over-expose so easily. With as many exposures they took to make this panel, I was hoping to have seen the Trapezium. It's not hard to understand why, as you loose so much other detail and brilliance of the nebula if you do expose for the Trapezium. There's no wonder this patch of the sky is one of the most photographed of the entire sky.


Orion is home to a whole bunch of stuff that the human eye cannot see that you see in this photo, especially Barnard's Loop and the Lambda Orionis Ring, which together occupy pretty much the entire size of the constellation. They are huge objects in the night sky, yet at H-alpha wavelength they appear too dim to the naked eye to see.

I've been doing a lot of landscape astrophotography lately in which I show these objects rising and setting over familiar landscapes to give people an idea of what our human vision is "missing out" on. Here's one of Orion that I took with a 28mm lens -- i.e. pretty wide angle:

https://www.instagram.com/p/CD9ztD0nnki/


Awesome. Is that picture taken with a regular DSLR type camera?


Yep, although I ripped the IR filter out, because the stock IR filter on consumer cameras cuts off the Hydrogen alpha wavelength (656.3 nm) which is what most of those red nebulae are.


How do I not accept cookies on that website?


Zoomable version linked in article: https://orion2020v5b.spaceforeverybody.com/


It is fascinating that there is probably life on some of those stars and its sad that humans will never able to explore places far into space.


If by ‘explore’ you mean ‘visit’ then sure. Otherwise, we pretty much know what to expect - astronomy and astrophysics are no joke. On a human scale, visiting another galaxy means just seeing more of the same, so why even bother. Diversity of life forms? I think we have a better chance to invent and explore it in labs here on earth.


We are always expecting more of the same, until it isn't. We are constantly observing unexpected things. Both with regards to life (on Earth, obviously, since we didn't find life anywhere else) and with regards to celestial bodies and physics.

And this is to be expected. We are building models to match our observations. "More of the same" is the basic assumption, but it doesn't mean it is true. That's why we experiment: to verify our models and if found wrong, to update them. The more things our model takes into account, the more we can expect it to explain everything, meaning the more we can expect to find "more of the same", but it is never a given.

I mean, we thought we had the universe almost completely figured out in the early 1900s, with just a few details remaining. The "details" turned out to be quantum physics and general relativity...


We haven't even finished discovering the diversity in life forms on this planet. I'm an astronomy nutter, with all of the possibilities it brings. However, I'm continually amazed when I remember that our little planet is covered 70% by water that we truly have not explored. We keep finding some new species when they do go looking. The diversity of life in water compared to on land is impressive.


There are many groups of insects that have a majority of species undescribed, chalcid wasps are all around you, everywhere outdoors, but so small they are never seen. Current efforts proposed to sequence the genome of every species are grand scope, and have good intention, but also, literally, impossible due to the vastness of biological diversity.


Does anyone know where I might be able to buy a full scale wall-sized poster of this? Or would that be more of a custom print job?


Wonderful image, it's a pleasure zooming in and exploring it!


I've never been a fan of the obligation to crank the saturation through the roof on astronomy images. I think it leads to unrealistic expectations. Any links to a more natural view?


Isn't this a never-ending discussion? There is no natural when you go past the naked eye. Any camera has a "filter", a development technique is a "filter"


Please, letting in a more light is nothing like cranking the saturation so high that grey turns to bright red.


Agreed. Without long exposures, nebulas are all but invisible. Even if you were inside of one, it would still be essentially invisible. They're really diffuse objects; the inside of a nebula is still a hard vacuum.


Has it been colorized?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: