Slashdot Mirror
Very Large Telescope Captures New 27-Megapixel Deep Field
xyz writes "European Southern Observatory's Very Large Telescope has captured the deepest ground based U-band image of the universe yet. The image contains more than 27 million pixels and is the result of 55 hours of observations with the VIMOS instrument. 'Galaxies were detected that are a billion times fainter than the unaided eye can see and over a range of colours not directly observable by the eye. This deep image has been essential to the discovery of a large number of new galaxies that are so far away that they are seen as they were when the Universe was only 2 billion years old.'"
Oh my god. It's full of pixels!
Hard to believe, looking at this, that there could ever be a shortage of anything.
My UID is prime!
Seriously, Slashdot, pointing to an article that contains a link to the 80 MB TIFF image at full resolution. Feeling a bit sadistic today, are we? Oh well, I'm rather early so I clicked it nonetheless. Feeling like a bit of a egocentric sadist myself today.
It works without a hitch in the AlternaTIFF TIFF Image Viewer. You can clearly see the galaxies, but otherwise it is a large sheet of colored dots (as expected I suppose).
The second link provides a 78MB TIFF (and a more modest but same-resolution 30MB JPEG) image.
However, it's a dinky low-resolution image one could have captured with a CCTV camera. Come on, you can do better than that.
I'm sorry, but what? The second link in the story has links to 6480 x 4236 JPGs and TIFs, which calculates to 27MP, the file sizes are 31MB and 79MB, respectively.
Normally, I would agree that web stories normally fall short with photos and multimedia, but it's just not true here.
The mind boggles. How anybody can believe we are here all alone, I don't know.
This new "picture" is taken in UV for which the Hubble ultra deep field is still the deepest image taken in visible wavelengths. Which provides, if you believe the current age estimate of the universe (13.73 ± 0.12 billion years old) means Hubble is still going back further. 0.73 Billion years vs 2 billion years since the beginning of the universe.
Just to give a sense of perspective in case you read the title and went so what?
Shortage of time.
:).
And here I am wasting it on Slashdot
The optics are the limiting factor here. Increasing pixel count wouldn't add any more detail.
Yeah, I know..... a lot....
What I mean is, if I look up in the sky, how big of a patch of the sky does this picture cover? The size of the full moon? Bigger? Smaller than a grain of sand at arms length?
Comment removed based on user account deletion
You may have a better vision. I can only see the cow.
Excellent name - simple and straightforward. They should have a contest for naming the next model. Put me down for "Amazing Freaking Ginormous Wonderscope"
Prisencolinensinainciusol. Ol Rait!
Phil Plait has quite a bit to say about this image:
http://blogs.discovermagazine.com/badastronomy/2008/11/07/voyaging-deep-into-the-universe/
"Scanning the full-res image is incredible. There's so much to see! Each dot, each smudge, is a full-blown galaxy, a collection of billions of stars. They're very, very far away; some of these galaxies are estimated to be 10 billion light years distant; you're seeing them as they were just a couple of billion years after the Universe itself began, and the faintest are one-billionth as bright as objects you can see with your own eye."
He also talks quite a bit about his favorite astronomical event - gamma-ray bursts.
I count galaxies in 1/8 x 1/14 of that image to be 150. In the whole image there are approx. 16800 galaxies. Since this is 14x21 arcminutes and 1 degree is 60minutes, hence this is 0.3 degree of 360 degree sky, I thinkg there are... 6.752*10^9 galaxies in the visible universe!
Did anyone notice the name of the press officer?
Dr. Henri Boffin.
Nominative determinism in action.
Every bloody emperor has his hand up history's skirt [Peter Hammill/VdGG]
And that's what most of the world is saying today.
No sig today...
If an image is sharp, then you weren't pushing the limits of the instrument in the first place.
Refraction flares caused by the crystalline pattern of molecules in the glass of the lenses. It only shows up in brighter objects because the flares are too dim for dimmer objects to make an impression. Bright stars simply overwhelm the local optics when you are gathering enough light to expose the dimmer objects.
Table-ized A.I.
Use Google Earth, and click the "sky" button. It's like Google Earth, but for the sky. Many different sources are mosaic-ed into it, and you can see how big some of these cosmic objects are from our vantage point, such as how much of an area of the sky this Deep Field image took in.
Refraction flares caused by the crystalline pattern of molecules in the glass of the lenses.
Um, no. The spikes are caused by the diffraction of light around the struts supporting the secondary mirror in the telescope. The wave nature of light ensures that no matter how large you build your telescope, you cannot focues stars to a perfect point.
Some corrections, because the GP confused linear and solid angles:
14 linear arcminutes * 21 linear arcminutes = 294 sqare arcminutes
1 square degree = (60 linear arcminutes)^2 = 3600 sqare arcminutes
294 square arcminutes / 3600 sqare arcminutes ~= .08167 square degrees
there are ~41253 square degrees in a sphere, only this fraction of a sphere is subtended by the picture:
(294 square arcminutes) / (41253 square degrees) ~= 1.980*10^-6
As someone stated elsewhere, this is about 1/500,000 of the sky (i.e. the celestial sphere).
So we count the number of galaxies encountered in this secion, then divide by the fraction subtended; using GP's estimate:
16,800 / (1.980*10^-6) gives ~8.49*10^9 galaxies
However, about 2 orders of magnitude more galaxies are in the field, though only ~16,800 galaxies are detected in this particular image of the field. The number of galaxies in the *observable* universe is at least on the order of 100 billion (10^11), per other, more sensitive surveys with more rigorous counting methods than a quick subsampling as performed by a human examining an image visually.
Next:
...with an average of 40 billion stars in a galaxy...
This is lower than I've encountered. The average galactic mass is about 100 billion solar masses, and the average stellar mass is about .5 solar masses*, so the the average number of stars in a galaxy is is on the order of 100~200 billion.
...it is conjectured that there are some very small galaxies, making the average much smaller than our own Milky Way...
Actually, it is fairly well established that there are indeed many such "small" galaxies. But though the number of "extremely large" (trillions to tens of trillions, versus hundred billions for the Milky Way) galaxies is small, the contribution to the mean ("average") number of stars per galaxy is disproportionately large because they themselves are disproportionately large. This is the nature of the arithmetic mean: a few highly weighted outliers skew the mean more than the median, and the median more than the mode. That's precisely why the "average" number of stars per galaxy is actually on the order of the Milky Way.
(* Note that the "average" stellar mass is skewed upward by the few but extremely massive stars just as galactic mass is. A "typical" star is smaller than the .5-solar mass "average" star; the vast majority of stars are smallish red dwarfs, with the sun being more massive than at least ~90% of stars, if only by a little in the range of stellar masses from ~.04 to ~150.)
So:
~(10^11 galaxies) * ~(10^11 stars/galaxy) = ~10^22 stars
The highest *reasonable* estimates I've seen yield a little over 5*10^22 stars, so on the order of 10^23 stars is still conceivable.
You're making a bad assumption here, here's why: As another respondent indicated, optics are one of the limiting factors here. The other major factor is the imaging device itself and the size of the individual pixels on the sensor. The amount of photons a individual pixel can collect is governed by the size of that pixel. The larger the pixel, the more photons it can gather. The more photons, the more light-sensitive it is. Now the sensor used on everything from your pocket P&S camera to a imaging device such as this are of a certain, finite size. You can fit only so many pixels on in its surface area.... the more pixels, the smaller they have to be in order to fit. Fewer pixels allows each pixel to occupy more space and gather more light. See the trade-off? With more pixels comes higher resolution at a cost of lower sensitivity, generally speaking. The capability of the optics is a moderator here as it has its own resolution limit. Given this optical limit, a point is reached where the addition of more pixels on the sensor becomes, essentially, a useless exercise with no return in terms of resolution and a definite loss in overall per-pixel sensitivity.
Well, it's time to hire Mythbusters to settle this. Maybe its both. If the struts were that big of a problem, then couldn't they use a flat lens-plate(s) to hold the secondary mirror instead?
Astronomers hate putting lenses into their optical systems--there is always some light lost to reflection off the glass surface. The VLT is an 8 meter diameter telescope, so supporting a giant lens above the telescope would be a major engineering issue. This isn't really a problem you can solve by adding a new lens or tweaking the secondary support structure--it's a fundamental feature caused by the wave nature of light. Anytime light passes through an aperture, it creates a diffraction pattern.