PDA

View Full Version : "we can safely call the APOD image a fake" - inputs from astrophotograhers please!



Nereid
2007-Aug-27, 07:27 PM
The 24 August 2007 of Astronomy Picture of the Day (http://antwrp.gsfc.nasa.gov/apod/ap070824.html) (APOD) is "Astronomer's Moon", and shows a processed image of Ganymede, taken with a very modern but nonetheless amateur setup.

In the Discuss an Astronomy Picture of the Day (APOD) forum (http://bb.nightskylive.net/asterisk/viewforum.php?f=9), one contributor writes (http://bb.nightskylive.net/asterisk/viewtopic.php?p=87928#87928):
After posting the initial message, I wondered if Hubble had an image to compare. They do:

http://hubblesite.org/newscenter/archive/releases/1995/35/image/c/

The detail in the Hubble image is nearly equivalent to the APOD image. Given Hubble is 10 times the diameter of the telescope in question and it is not looking through the atmosphere, I think we can safely call the APOD image a fake.What do you think? Could that image have been taken with the setup described?

triclon
2007-Aug-27, 08:00 PM
I believe it is real. Says it is made from 409 out of 10000 stacked images. Never underestimate the power of image stacking.

stargazer_7000
2007-Aug-27, 08:42 PM
I would not say it is a "fake", that implies that Yuri has done that on purpose, I just believe they seem to have processed "noise"...unfortunately.

Kyle Edwards
2007-Aug-27, 09:51 PM
The maximum angular resolution of a 9 inch telescope is .50 arcseconds, and and since Ganymede's disc is 1.7 arcseconds, they should not be seeing detail smaller than a third or fourth the diameter of the disc.

andyschlei
2007-Aug-27, 10:54 PM
There is no reason to think this is a fake. Mike Salway has produced close to this detail (http://www.iceinspace.com.au/mygallery/displayimage.php?album=33&pos=14) with his 12" setup. Ideal seeing and lots of data (10,000 frames!) could make up for the difference in apeture. Also note that like most Matsukov Cassegrains, it is not a fast scope -- essentially f13 -- so it has a lot of focal length for its apeture.

paul f. campbell
2007-Aug-28, 01:41 AM
Hi Neried.
The first thing we need to decised is. Ok is it a fake or not. Then the next thing we need to decised is. Ok,well do we really care or not. If it is a fake then its a good one, but if its real, then its a good dame photo.

I for one am not quality to judge the authenticity of this image, but I am not saying that you are not. I have taken many photos of Jupiter with my etx 125 in the last 4 years, that I feel would make any 8 inch telescope plush. If you have not seen some of my photos please do, they are not the best out there but I think they are the best photos that this etx 125 has to offer.

If we are going to call this person on this, then lets get the facts from him. Write this person an e-mail and get his take on it. Then present us with the facts as he sees it. Is that not the American way of things. I should hope
it is because I live HERE. I am not trying to bust anyones gibees. Lets just get the facts and clear the story.
Clear skies to all. Paul

EDG
2007-Aug-28, 03:25 PM
Well it looks about right. At the time Galileo Regio should have been up in the northeastern quarter of the image (and there is a dark patch there), and Osiris should have been below and to the left of it (and there is a bright patch there).

Galactic2000
2007-Aug-28, 11:32 PM
I have to disagree with the Fake theory...

In fact Jerry, and Robert who run APOD are professional Astronomers who work for NASA and Universities and I'm sure they verified this image before posting it as an APOD.

If you guys would just check out Don Parkers Planet/moons images and even Mike Salways, or even Christopher Go's images you will see the same details in their shots of the Jovian Moons, if they were to magnify them a bit more they would look a lot like Yuri's pixelized shot, pushing it to the limits......these guys are using very modest equipment with good optics, and have one key but most important ingrediant, that is "Excellent Seeing"!!!!........

Besides, it not very nice to imply Yuri is not on the level with his image,
always give the benefit of the doubt, unless he admits otherwise, or NASA scientists disprove it.

Amatuers are producing images today rivaling and exceeding professional images, many of us long time astrophotographers are assisting professionals with their research, by them allowing us to participate in projects and us taking very deep images for comparison with their results, it is the same for solar system researchers as well.

My hats off to Yuri for pushing equipment and his personal will to the limits!

EDG
2007-Aug-29, 02:14 AM
Yeah, I'm not sure why people are claiming this is a fake. They seem to compare it to a hubble image, then declare that since Yuri's image is apparently more detailed (which it really isn't) then it must be fake? That's just silly.

Kyle Edwards
2007-Aug-30, 03:32 AM
They seem to compare it to a hubble image, then declare that since Yuri's image is apparently more detailed (which it really isn't) then it must be fake? That's just silly.

I take it you missed the part about diffraction not allowing a 9 inch telescope to get that much detail, and let alone through the atmosphere?

Nereid
2007-Aug-30, 10:57 AM
Thanks for the comments so far, especially from BAUT members who are astrophotographers.

I'm particularly interested in getting inputs from those who have similar setups to Yuri Goryachko (described in the APOD page): while you yourself may not have produced an image quite like this one, from the (planetary) images you have, and your experience, how close do you think you could come?

hhEb09'1
2007-Aug-30, 11:19 AM
I'm particularly interested in getting inputs from those who have similar setups to Yuri Goryachko (described in the APOD page): while you yourself may not have produced an image quite like this one, from the (planetary) images you have, and your experience, how close do you think you could come?Here's an excellent resource (http://www.stellarproducts.com/about/observatory/Resolution%20tests/resolution%20tests.htm), it's apparently the basis for an article that appeared in Sky and Telescope in 2005 entitled Resolution Realities, and in particular it discusses the imaging of Ganymede.

EDG
2007-Aug-30, 02:33 PM
I take it you missed the part about diffraction not allowing a 9 inch telescope to get that much detail, and let alone through the atmosphere?

There's one guy on these very boards who has posted some pretty detailed colour images of Ganymede and Callisto - how big a scope was he using?

Kyle Edwards
2007-Aug-30, 03:13 PM
There's one guy on these very boards who has posted some pretty detailed colour images of Ganymede and Callisto - how big a scope was he using?

Is this what you are referring to?

http://www.iceinspace.com.au/mygallery/displayimage.php?album=lastup&cat=0&pos=5

I am not saying that I think the APOD is faked, but that because it would be physically impossible to see features that small (like the bright spot near the bottom of the disc in the APOD image), I think some noise artifacts got processed.

Kyle

andyschlei
2007-Aug-30, 05:05 PM
Is this what you are referring to?

http://www.iceinspace.com.au/mygallery/displayimage.php?album=lastup&cat=0&pos=5

I am not saying that I think the APOD is faked, but that because it would be physically impossible to see features that small (like the bright spot near the bottom of the disc in the APOD image), I think some noise artifacts got processed.

Kyle

Mike Salway's Ganymede has a similar bright spot to Yuri's APOD. Yuri's APOD has been enlarged, where Mike's hasn't. Mike uses a 12" newtonian, where Yuri is using a 9" matsukov. Is there sufficient difference in the resolving power of the scopes to think that the bright spot in Yuri's image is an artifact?

At some level we are back to the "do astrophotos look like the real thing (http://www.bautforum.com/astronomy/57046-do-astrophotos-look-like-real-thing.html)" problem, only this time with potential enhancement of an artifact into what appears to be detail on an object. How does one know when they have pushed things too far on the sharpening / image enhancement side?

One other comment. Yuri used 409 frames (the best of 10,000) to get his image. IMHO, this would make noise an unlikely candidate for the artifact, unless the 409 frames were chosen for that particular piece of noise. This raises the question where did the original artifact come from to then be enhanced into part of the final image?

:think:

Kyle Edwards
2007-Aug-30, 05:27 PM
Mike Salway's Ganymede has a similar bright spot to Yuri's APOD. Yuri's APOD has been enlarged, where Mike's hasn't. Mike uses a 12" newtonian, where Yuri is using a 9" matsukov. Is there sufficient difference in the resolving power of the scopes to think that the bright spot in Yuri's image is an artifact?

At some level we are back to the "do astrophotos look like the real thing (http://www.bautforum.com/astronomy/57046-do-astrophotos-look-like-real-thing.html)" problem, only this time with potential enhancement of an artifact into what appears to be detail on an object. How does one know when they have pushed things too far on the sharpening / image enhancement side?

One other comment. Yuri used 409 frames (the best of 10,000) to get his image. IMHO, this would make noise an unlikely candidate for the artifact, unless the 409 frames were chosen for that particular piece of noise. This raises the question where did the original artifact come from to then be enhanced into part of the final image?

:think:

You have very good points. It is hard to say for sure what is going on with the image, wether or not it is overprocessed.

I may be wrong, but after processing planets like Mercury myself and seeing how easy it would be to expand the image 500x and use heavy sharpening, it is just hard to believe that it is real detail.

Especially since there appears to be more detail in his 1.7 arcsecond disc than the Keck telescope can get on the 3-4 arcsecond disc of Uranus without adaptive optics.

http://www.centauri-dreams.org/wp-content/images/thumb-power_of_ao_ucb_small.jpg

andyschlei
2007-Aug-30, 05:47 PM
I may be wrong, but after processing planets like Mercury myself and seeing how easy it would be to expand the image 500x and use heavy sharpening, it is just hard to believe that it is real detail.

True enough, it is easy to go overboard on processing. Although on the outer planets you will have better seeing than you would ever have with Mercury, hence less noise to expand and enhance. By reputation, I know Mike Salway is very careful on this point.


Especially since there appears to be more detail in his 1.7 arcsecond disc than the Keck telescope can get on the 3-4 arcsecond disc of Uranus without adaptive optics.

I see your point. But by taking a large number of short-exposure images, selecting the best, and combining them, you get adaptive-optics like results. Instead of adapting to "clear-up" atmospheric distortion, you pick the shots that have little distortion. Same net result. Also, isn't Ganymede much brighter than Uranus? Would that better S/N ratio provide a better ability to resolve details?

Kyle Edwards
2007-Aug-30, 06:11 PM
True enough, it is easy to go overboard on processing. Although on the outer planets you will have better seeing than you would ever have with Mercury, hence less noise to expand and enhance. By reputation, I know Mike Salway is very careful on this point.


You are right, and I wasn't suggesting that Mike Salway was overprocessing his images.




I see your point. But by taking a large number of short-exposure images, selecting the best, and combining them, you get adaptive-optics like results. Instead of adapting to "clear-up" atmospheric distortion, you pick the shots that have little distortion. Same net result. Also, isn't Ganymede much brighter than Uranus? Would that better S/N ratio provide a better ability to resolve details?


Good point. I do not pretend to know everything about imaging, but I guess is possible to get AO-like results just from using such large numbers of frames.

Maybe I should image Ganymede some time and take thousands of frames, since my telescope (8.35 inch F/12.4 cassegrain) is somewhat similar to Yuri's

andyschlei
2007-Aug-30, 07:14 PM
You are right, and I wasn't suggesting that Mike Salway was overprocessing his images.

Kyle, I certainly did not mean to imply that you were saying that about Mike. Sorry for my sloppy writing.


Maybe I should imaging Ganymede some time and taking thousands of frames, since my telescope (8.35 inch F/12.4 cassegrain) is somewhat similar to Yuri's

That would be a very good experiment. The wild card is the seeing.

I thought about trying myself with my C-11 or my C-8, but Jupiter is far too low this year where I am in Southern California. And I need to get a DMK or other better quality camera -- perhaps this is just the excuse - I mean reason - to spend the money. :D

cloudbait
2007-Aug-31, 04:43 PM
I am not saying that I think the APOD is faked, but that because it would be physically impossible to see features that small (like the bright spot near the bottom of the disc in the APOD image), I think some noise artifacts got processed.

It isn't physically impossible at all to "see" features like the small bright spot. The telescope aperture limits the size of features that can be resolved, which is an altogether different thing. Small, high contrast features that are smaller than the resolution limit of optics are detected all the time. Stars are the most obvious example, but there are also small lunar craters, Saturnian ring divisions, and many others.

There seems to be an expectation that if the resolution limit of the optics is 1/3 the size of Ganymede's disk, that the image should somehow look like an ~3x3 pixelated grid. When you sum up many images this way, you smooth out transitions and can produce a reasonable brightness map of the surface, even though it is poorly resolved.

That said, I'm sure that many of the features in this image are processing artifacts, even if the average brightness variations (on a scale of about 1/3 the disk diameter) represent the actual surface of the Moon. A good quality test would be to throw out the images that were stacked, and pick the next best few hundred. The final image might be somewhat worse, but should definitely correlate with the first.

iceman
2007-Aug-31, 08:48 PM
It's not a fake, I think it's a bit rude to imply it's "fake".

However I do believe the excessive enlarging has introduced artifacts that are separate from the real albedo features which are quite easy to capture, even at the altitude of 14deg.

I posted a more comprehensive reply in the thread on CN. I'll copy/paste the text here (below).

Showing detail on Ganymede is nothing new, myself and many other planetary imagers can do it pretty much every time we capture it when the conditions are good.

But at only 1.6", I guess it's how much real detail you can show. Major albedo markings are no problem. Even Io at a smaller angular size can have albedo markings revealed. Some examples of my captures of Ganymede this year ( #1 (http://www.iceinspace.com.au/mygallery/displayimage.php?album=33&pos=4), #2 (http://www.iceinspace.com.au/mygallery/displayimage.php?album=33&pos=7), #3 (http://www.iceinspace.com.au/mygallery/displayimage.php?album=33&pos=14), #4 (http://www.iceinspace.com.au/mygallery/displayimage.php?album=33&pos=20), #5 (http://www.iceinspace.com.au/downloads/20070323-jupiter_anim.gif) )

I think it's the massive resampling of 7.6x that most people would find introduces artifacts or noise that could be mistaken for real detail. I never resample my images more than 2.5x, and even then it's usually only for printing purposes. Most times I leave my images at their natural size and that's the way I prefer it.

Yuri and his friends are very advanced in the planetary imaging field - just look at his images of Mars, Mercury and Venus. Even ganymede captured at 14deg alt in average seeing, if you capture enough frames, could show albedo markings but when it's processed and resized to such a degree, the real detail being shown is just not there (in my opinion).

tdvance
2007-Sep-08, 04:04 PM
It isn't physically impossible at all to "see" features like the small bright spot. The telescope aperture limits the size of features that can be resolved, which is an altogether different thing. Small, high contrast features that are smaller than the resolution limit of optics are detected all the time. Stars are the most obvious example, but there are also small lunar craters, Saturnian ring divisions, and many others.


yes--Betelgeuse has an angular width of 0.054 arcseconds, much smaller than what amateur scopes can resolve and much, much smaller than the human eye can resolve (a good eye resolves down to about an arcminute or two). Yet, we can see Betelgeuse very clearly in Orion.

Resolution tells how well a scope can separate two close features--for example, a close double star or two features on a planet or moon that are close together.

In my opinion, the photo is not faked. There are other really good amateur photos out there, some which are really in the running with Hubble photos, including one to be in the 2008 Guinness Book of World Records for largest spiral galaxy photograph (a mosaic of many photos through an amateur scope, beating the former record by the Hubble telescope): http://www.robgendlerastropics.com/M31NMmosaic.html

If I didn't know better and saw this image, I would certainly have assumed it was a Hubble image. I believe it has been an APOD once, and I've seen it in SBIG advertisements in astronomy magazines too.

Todd