Highest Resolution Photos Ever of the Night Sky

[Psnarf]

"These images are also at least twice as sharp as what the Hubble Space Telescope (HST) can make because the 6.5m Magellan telescope is much larger than the 2.4m HST."
https://visao.as.arizona.edu/

They converted a 36" broken mirror into a 35.5" secondary mirror with complex adaptive optics that makes 1000 atmospheric corrections per second. The images are toward the blue end of the visible-light spectrum.

Here's a link to one of the papers coming from this new eyeball:
http://arxiv.org/abs/1308.4155

[geckzilla]

Nice, this image shows a comparison. I wonder if those are from a single image or if each inset had to be done separately?


Very curious to see a full color image. Maybe they should do their own version of the Pillars of Creation, which is arguably Hubble's most famous picture.

[Chris Peterson]

Nice, this image shows a comparison. I wonder if those are from a single image or if each inset had to be done separately?


Very curious to see a full color image. Maybe they should do their own version of the Pillars of Creation, which is arguably Hubble's most famous picture.

The information about this system is very misleading, IMO.

There is nothing revolutionary about the new camera, it is a fairly typical AO system, with some evolutionary improvements (in particular, a large number of modes and a design that works into the long end of the visible light range). It does not image in color. Like all AO systems, it is extremely narrow field (8 arcseconds square for the visible light camera) and requires a bright reference star (limiting R magnitude of 14) very close (< 8 arcsec) to the target (the system doesn't have an artificial guide star laser).

The wide field reference image was not produced by this instrument, and could not be. The inset images were produced from separate exposures through a red filter (the images themselves, although displayed with a red palette, are single channel, monochromatic). The wavelength range of the camera is 600-1050nm, so all the available filters are red or infrared. The resolution is very impressive, but this instrument isn't going to outperform Hubble for any but a very specific, short list of important science goals.

[geckzilla]

It was obvious to me that the background image was the Hubble image included only for comparison and the insets were monochromatic and from the MagAO. With those sorts of limitations you listed, it seems a lot less exciting, though. Extremely narrow field of view. That's what I was curious about.

[stephen63]

Exoplanet research?

[Beyond]

In other words... It fills a particular niche very well

[Chris Peterson]

Exoplanet research?

Yes. Exoplanet atmospheres. Deep circumstellar imaging. Close binaries. In the Solar System, the surface of moons and asteroids. All very important and useful.

Comparing this with the HST is really a case of apples and oranges.

[geckzilla]

I suppose you could look for Yet Another Pluto Moon with it.

[neufer]

I suppose you could look for Yet Another Pluto Moon with it.

The problem with finding small moons of Pluto probably has less to do with high resolution
than it has with long exposures and a really dark background.

The examples shown here had only 60 second exposure times.

[Beyond]

60 seconds Heck, streakers get more exposure time than that, especially if they can run fast.

[Chris Peterson]

I suppose you could look for Yet Another Pluto Moon with it.

The problem with finding small moons of Pluto probably has less to do with high resolution
than it has with long exposures and a really dark background.

The examples shown here had only 60 second exposure times.

No, I think it's more of a resolution issue. This thing is attached to a very big telescope. Pluto will saturate in seconds; even a dim moon wouldn't require a long exposure. Pluto itself would probably need to be occulted, but the camera has that capability. The problem is that Pluto isn't generally bright enough for the AO system's wavefront detector, so any search would need to be carried out while the planet was passing within a few arcseconds of a star brighter than mag 14.

[geckzilla]

If Stellarium is accurate, 2014 May 9th at 00:00 is a time to look. A magnitude 13.3 star almost touches Nix 35 minutes prior to that.

[neufer]

I suppose you could look for Yet Another Pluto Moon with it.

The problem with finding small moons of Pluto probably has less to do with high resolution
than it has with long exposures and a really dark background.

The examples shown here had only 60 second exposure times.

No, I think it's more of a resolution issue. This thing is attached to a very big telescope. Pluto will saturate in seconds; even a dim moon wouldn't require a long exposure. Pluto itself would probably need to be occulted, but the camera has that capability. The problem is that Pluto isn't generally bright enough for the AO system's wavefront detector, so any search would need to be carried out while the planet was passing within a few arcseconds of a star brighter than mag 14.

Pluto isn't tiny enough for the AO system's wavefront detector.

Nevertheless I stand by my statement:

The problem with finding small moons of Pluto probably has less to do with high resolution
than it has with long exposures and a really dark background.

I know very well what the Hubble can produce...
Ah to have a couple of hours time on that beautiful machine to do some wideband RGB imaging...

You won't get much but noise in a couple of hours.

Edit to above statement: Actually, it would depend on what object you are looking at.
A bright object probably would only take a couple of hours.

[Chris Peterson]

Pluto isn't tiny enough for the AO system's wavefront detector.

That is almost certainly true, although in principle you can use AO techniques with non-stellar references as long as you have knowledge of their true profile. But for anything non-stellar, the wavefront analysis becomes much more complex, and the correction rate would necessarily drop.

Nevertheless I stand by my statement:

The problem with finding small moons of Pluto probably has less to do with high resolution
than it has with long exposures and a really dark background.

Hubble images of Pluto saturate in about a second. The VisAO camera will probably saturate about an order of magnitude faster, so the exposure time for targeting shots will be no more than a few hundred milliseconds. Hubble collects high S/N images of Nix and Hydra with 8-minute exposures, so that would be about 1-minute exposures with the VisAO. The most recently discovered moons are about three magnitudes dimmer, so would require exposures some 16 times longer.

We're talking exposures well under an hour, which means that there shouldn't be a problem keeping a reference star in the field, and any moons should produce very little movement. So I hardly think sensitivity is an issue (although as I noted, an occulting disc would need to be used to prevent blooming around Pluto). The magnitudes of bodies in orbit around Pluto are nowhere near the limiting magnitude of the camera and telescope, so sky background isn't an issue. Finally, the camera employed is an EMCCD, which means it has no readout noise and can be operated at high (even video) frame rates. So if conditions or AO requirements dictate, an image can be summed from shorter exposures with little or no S/N penalty.

[neufer]

I stand by my statement:

The problem with finding small moons of Pluto probably has less to do with high resolution
than it has with long exposures and a really dark background.

Hubble images of Pluto saturate in about a second. The VisAO camera will probably saturate about an order of magnitude faster, so the exposure time for targeting shots will be no more than a few hundred milliseconds. Hubble collects high S/N images of Nix and Hydra with 8-minute exposures, so that would be about 1-minute exposures with the VisAO. The most recently discovered moons are about three magnitudes dimmer, so would require exposures some 16 times longer.

We're talking exposures well under an hour, which means that there shouldn't be a problem keeping a reference star in the field, and any moons should produce very little movement. So I hardly think sensitivity is an issue (although as I noted, an occulting disc would need to be used to prevent blooming around Pluto). The magnitudes of bodies in orbit around Pluto are nowhere near the limiting magnitude of the camera and telescope, so sky background isn't an issue. Finally, the camera employed is an EMCCD, which means it has no readout noise and can be operated at high (even video) frame rates. So if conditions or AO requirements dictate, an image can be summed from shorter exposures with little or no S/N penalty.

The next moons of Pluto will probably be at least another three magnitudes dimmer,
and so will require exposures at least 16 times longer still. (Advantage Hubble.)

[Chris Peterson]

The next moons of Pluto will probably be at least another three magnitudes dimmer,
and so will require exposures at least 16 times longer still. (Advantage Hubble.)

Why will they be three magnitudes dimmer? Why would that give Hubble an advantage? Three magnitudes dimmer is still within the limiting magnitude of ground based instruments.

[neufer]

The next moons of Pluto will probably be at least another three magnitudes dimmer,
and so will require exposures at least 16 times longer still. (Advantage Hubble.)

Why will they be three magnitudes dimmer?

[b][color=#0000FF][size=135]Hubble image of the Plutonian system, showing the three larger moons[/size][/color][/b]

---

Because there is no trace of them currently.

(Unless Charon has companion Trojan moons the size of Nix or Hydra.)

Why would that give Hubble an advantage? Three magnitudes dimmer is still within the limiting magnitude of ground based instruments.

We're talking exposures well over an hour for ground based instruments, which means that there should be a problem keeping a reference star in the field, and moons probably would produce noticeable movement at hi-res.

The 6.5m Magellan telescope was clearly designed to resolve brighter stars using short exposures. If Charon has companion Trojan moons the size of Nix or Hydra then Magellan has the advantage.

[Chris Peterson]

Why will they be three magnitudes dimmer?

Because there is no trace of them currently.

One magnitude would be enough for that. As would the presence of objects closer to Pluto.

We're talking exposures well over an hour for ground based instruments, which means that there should be a problem keeping a reference star in the field, and moons probably would produce noticeable movement at hi-res.

I don't think so. Pluto has a drift rate of about 1 arcsecond per hour, so it should be possible to keep a properly selected reference star in the field for as long as 16 hours. The known moons have orbital periods measured in days, so an exposure of several hours should not be a problem.

The 6.5m Magellan telescope was clearly designed to resolve brighter stars using short exposures.

The telescope was clearly not designed for such a narrow purpose! Neither is this accurate for the AO system or the VisAO camera. They are perfectly well suited to long exposures, as long as there is a bright reference star close to the target. It is only in the case where the target and the reference are the same that we see short exposures, as in the examples.

[neufer]

Why will they be three magnitudes dimmer?

Because there is no trace of them currently.

One magnitude would be enough for that. As would the presence of objects closer to Pluto.

I grant that Hubble is at a distinct disadvantage for moons closer to Pluto (and for resolving features on Pluto).

We're talking exposures well over an hour for ground based instruments, which means that there should be a problem keeping a reference star in the field, and moons probably would produce noticeable movement at hi-res.

I don't think so. Pluto has a drift rate of about 1 arcsecond per hour, so it should be possible to keep a properly selected reference star in the field for as long as 16 hours. The known moons have orbital periods measured in days, so an exposure of several hours should not be a problem.

Pluto currently has a "drift rate" of about 1 arcsecond per hour due to its own speed.

The retrograde drift rate (taking into account the Earth's speed) can be as much as 3 to 4 arcseconds per hour.

The 6.5m Magellan telescope was clearly designed to resolve brighter stars using short exposures.

The telescope was clearly not designed for such a narrow purpose! Neither is this accurate for the AO system or the VisAO camera. They are perfectly well suited to long exposures, as long as there is a bright reference star close to the target. It is only in the case where the target and the reference are the same that we see short exposures, as in the examples.

My guess is that the Magellan telescope takes a number of 1 minute exposures and only uses the shot where the atmospheric turbulence is at a minimum. I could be wrong.

[Chris Peterson]

My guess is that the Magellan telescope takes a number of 1 minute exposures and only uses the shot where the atmospheric turbulence is at a minimum. I could be wrong.

That isn't how it works. It is a perfectly conventional astronomical telescope (a pair, actually, although I think only one is finished). As such, it is used in a variety of ways, the most common being long exposure tracked images of deep sky objects. What you are describing is called "lucky imaging", and isn't often used professionally. When it is, it is on bright objects and with very short exposures - less than a second.

The VisAO instrument (which is only one of many instruments used on this telescope) works by modifying the optical wavefront, thereby allowing the camera to operate near the theoretical resolution of the 6.5 meter objective (which is quite a bit better than the atmosphere limited resolution without active optics). The VisAO instrument isn't designed for any particular exposure time; it will be short for bright objects, and long for dim ones.

[neufer]

My guess is that the Magellan telescope takes a number of 1 minute exposures and only uses the shot where the atmospheric turbulence is at a minimum. I could be wrong.

That isn't how it works. It is a perfectly conventional astronomical telescope (a pair, actually, although I think only one is finished). As such, it is used in a variety of ways, the most common being long exposure tracked images of deep sky objects. What you are describing is called "lucky imaging", and isn't often used professionally. When it is, it is on bright objects and with very short exposures - less than a second.

The VisAO instrument (which is only one of many instruments used on this telescope) works by modifying the optical wavefront, thereby allowing the camera to operate near the theoretical resolution of the 6.5 meter objective (which is quite a bit better than the atmosphere limited resolution without active optics). The VisAO instrument isn't designed for any particular exposure time; it will be short for bright objects, and long for dim ones.

I stand by my assumption that the Magellan telescope depends upon:
BOTH modifying the optical wavefront AND "lucky imaging."

(With 1 minute disposable "frames" I'll bet it gets a lot of "lucky imaging.")

[Chris Peterson]

I stand by my assumption that the Magellan telescope depends upon:
BOTH modifying the optical wavefront AND "lucky imaging."

(With 1 minute disposable "frames" I'll bet it gets a lot of "lucky imaging.")

I'd love to see some evidence of your second assumption.

I doubt very much that the other cameras used on this telescope use short exposures (because there is a terrible readout noise penalty for doing so, very few of their targets would value from one minute exposures, and to take advantage of lucky imaging requires subsecond exposure times, which makes the readout noise problem even greater).

The VisAO camera doesn't use lucky imaging because it has something better: active optics.

[neufer]

I stand by my assumption that the Magellan telescope depends upon:
BOTH modifying the optical wavefront AND "lucky imaging."

(With 1 minute disposable "frames" I'll bet it gets a lot of "lucky imaging.")

I'd love to see some evidence of your second assumption.

I'd love to see a 10 minute hi-res exposure from the Magellan telescope.

I doubt very much that the other cameras used on this telescope use short exposures (because there is a terrible readout noise penalty for doing so, very few of their targets would value from one minute exposures, and to take advantage of lucky imaging requires subsecond exposure times, which makes the readout noise problem even greater).

There doesn't have to be zero turbulence just minimal turbulence.

The VisAO camera doesn't use lucky imaging because it has something better: active optics.

The larger the effective aperture the more work the active optics have to perform.

There's plenty of work for the Magellan telescope to do without forcing it to also observe the faintest stars.

[Chris Peterson]

I'd love to see a 10 minute hi-res exposure from the Magellan telescope.

Here are a bunch of them, most much longer than 10 minutes.

There's plenty of work for the Magellan telescope to do without forcing it to also observe the faintest stars.

It is a 6.5 meter telescope. Its primary function is observing the faintest objects!

[neufer]

I'd love to see a 10 minute hi-res exposure from the Magellan telescope.

Here are a bunch of them, most much longer than 10 minutes.

None of them super hi-res

There's plenty of work for the Magellan telescope to do without forcing it to also observe the faintest stars.

It is a 6.5 meter telescope. Its primary function is observing the faintest objects!

Not at super hi-res

Large apertures with adaptive optics permit super hi-res but not for the faintest objects!

(They also permit hi-res for the faintest objects!)