Friday, June 4, 2010

The astronomical fraud of Alberto Geyer (@Solarview)

At the beginning of april this year, someone called Alberto Geyer, with username @solarview, added me on Twitter. I saw his account was about astronomy, but at first I didn't pay it much attention. However, a few days later I took a look at his profile out of curiosity, and I saw that he seemed to take pictures of Solar System objects. In some occasions he sent tweets to important users, like @NewHorizons2015, the probe that will arrive to Pluto that year, @CassiniSaturn, the one that orbits Saturn in these moments, @plutokiller, the astrophysicist Mike Brown who discovered Eris, etc., to show them, saying that with a new technique he achieved very good quality. So I clicked on one picture of Jupiter's moon Europa, that he had uploaded on Twitpic, and it looked good to me, but at the same time something arose my suspicions: I'd seen Hubble images that had less resolution of Jupiter's moons than this picture had.

It was at this point when I decided to visit his webpage, and I got disappointed. I saw that nearly all the photographs were either exaggerated amplifications of the noise in the image so that it looked like detail in the surface of the imaged bodies, or images taken by NASA probes the quality of which had been reduced afterwards. According to him, he took those images from Earth with his backyard telescope. My first impression: it was all a Photoshop hoax. I realized that among his tweets there were some replying to someone who didn't believe he got those results, and to whom he said to look at the raw images (the image with the unaltered data taken from the CCD, the camera used for astrophotography) posted on the web precisely for the unbelievers. Since it looked like it was a fraud, I decided to send him a message via Twitter that started an entire argument, of which some of my contacts were witnesses. In this conversation I tried to debunk his wrong assertions, while he systematically denied everything. Watching the conversation one could get an idea of the credibility of this Alberto Geyer, and it didn't speak well of him. I suppose that for this very reason, a few days later he deleted all his tweets regarding this issue and he continued his activities as if nothing had happened. Due to this, and after seeing his number of followers and admirers that think he really gets those images with honest means was still growing, I decided to write this post reproducing the entire conversation (which I saved at the time), adding explanations to clarify what happened while tweets followed one another. All the URLs of the original tweets are linked in the usernames, as well as the tweets being referred to when they are replies. Translations of the tweets in Spanish are in italics. Everything started like this:

DarkSapiens: @solarview Wow, took a look at your webpage and turns out it's all photoshopped… Nice trick, but misleading people is never good.

Solarview: @DarkSapiens What you saw is what is there!Take a moment to download the raw images on every each page.That was made available for doubters

DarkSapiens: @solarview I saw enough. Pluto is made amplifying the noise at such high levels that looks like detail. Any other blob would do the same.

Pluto. In which Alberto Geyer claims to have discovered a mountain 400 or 500 km in height. In the web itself there are published two sequences of screen captures (I y II) in which he shows his process. Starting from a little spot with a quite a lot of noise, he gets a pretty circle with what looks like a lot of surface detail. Detail that wasn't in the original image. And it is interesting to see that Geyer claims that the mountain can be seen from Earth in the unprocessed image, as something sticking out from the edge of the spot. A spot that is actually the combined light of Pluto and its moons Charon, Nix and Hydra, a point-like image and not the disk of the planet, that falls well below the resolution level.

We go on.

DarkSapiens: @solarview Mars has images of the mounts pasted on top of it to make it appear more detailed. And they're in wrong positions.

As can be seen here. The mounts being lit from the side while light arrives to the planet head on is also priceless.

DarkSapiens: @solarview Enceladus would be a bit more believable without the jets. It's even a shame someone has to say that.

Enceladus is already unbelievable. The jets, that were discovered by the Cassini precisely because this moon was between the Sun and the probe, are added lightly by Alberto Geyer in the image, and what's more, going in both directions (the real ones only come out from one side) and rotated with regard to their true position. Do you see those blue lines in the image? (Better in the original, in which by the way they look blue because it's in false color) That's the region from where they actually come out.

DarkSapiens: @solarview It's so obvious you took pictures from the space probes and reduced the quality. In some cases I even remember the original one.

DarkSapiens: @solarview Oh, and Mercury in full phase, something only achievable when it's BEHIND the Sun. Yeah right… :)

DarkSapiens: @solarview I may continue if you want :)

Solarview: @DarkSapiens Go On Sapiens!Keep visiting our website!You obviously never saw a CCD camera let alone look in the eyepiece of a telescope!

DarkSapiens: @solarview Your last phrase is dead wrong. I'll visit your web later and tweet the URLs with links to the original images for ppl to compare

Since I belong to a good astronomical association where several members do astrophotography, and I even have a modest telescope, it can be said that this claim bothered me. Especially because of him denying in this way something that I was showing him in great detail…

Solarview: @DarkSapiens O.K. Why don't u start with Io's eruptions.The ejecta is right there in the raw image.Io is third from Jupiter.

DarkSapiens: @solarview Io: moons out of focus in the raw image. And you used this

The image from Wikipedia should be inverted from left to right and rotated a bit to fit exactly with the details of the volcanoes seen in the surface of Io in his picture, but nothing more. To explain the out of focus thing we have to go to the raw image, from which the following is a portion (click to enlarge):

In it it can be seen how Jupiter is overexposed to bring out the fainter moons, but those don't have a point-like shape, but its shape is a strange curve with several details. You can also see that this curve is in the same orientation for the three moons in the image (Io is in the lower right). This can not be due to the Sun lighting them in an angle, since Jupiter is much further out from the Sun than Earth, and as seen from here it will lit them always face on. This curved shape is precisely due to the image being out of focus, creating the same pattern in all of them.

At this point I decide to announce in public what I am doing, while Alberto Geyer continues replying.

DarkSapiens: Desmontando el fraude que es @solarview. Hay gente para todo.

Debunking the fraud that is @solarview. There's people for everything.

Solarview: @DarkSapiens Even a baby can see the eruptions!With regard to Mercury,a day prior to superior conjunction is like a day prior to full moon.

DarkSapiens: @solarview Mercury prior to conjunction is so close to the sun that the Earth's atmosphere is brighter. You can't take a pic.

DarkSapiens: @solarview All moons are out of focus in that Jupiter image, what you saw are definitely not eruptions.

I will talk about the Mercury issue further on, with more data. Meanwhile, Alberto Geyer changes tactics, and starts treating me like if I was around 12 years old, probably trying to discredit me:

Solarview: @DarkSapiens Ask your father for a telescope,an 8" will do.Start with the moon.Then try spoting Mars.Saturn is a good target too.Then try...

DarkSapiens: @solarview Are those telescopes good enough? I'm in those pictures and have used them.

The Mercury topic continues:

Solarview: @DarkSapiens The image info(available on display) tells all.If I can't take an image,why is the raw image there with Mercury's full disc?!.

DarkSapiens: @solarview The Mercury raw image shows a double blob. It's your processing/manipulation what made it round.

The raw image from which he supposedly creates his Mercury image is below, and a zoom on it after that.

In the zoomed image it can be seen, however, that only one every four pixels are dark in what would apparently be what divides the blob in two, what made me think he uses a color CCD (specifically, I think he uses this one). It is interesting, because this kind of cameras are not recommended in astrophotography if you want to get a high level of detail (precisely what Alberto Geyer wants), because there is a loss of resolution by using contiguous pixels for different colors. The preferable method is to use a monochrome camera and take several shots with color filters, making use in this way of the maximum resolution in all channels. Nevertheless, judging by the moons of Mars here, it seems that for his "processings" he uses all pixels at once, with no color distinction, increasing the irregularity of the final product, so this issue doesn't seem to be important to him.

I concentrate now in the conditions to take the picture. To tell you the truth, it may not be completely impossible to capture an image of Mercury in superior conjunction (when it is just on the other side of the Sun), since the tilt of its orbit is different from ours and not every year this planet passes behind our star (like there are no transits every year). Mercury would still be very close to the Sun, but there are very experienced astrophotographers that manage to take pictures of the crescent Moon only a few hours from the new phase. I don't really know if the brightness of the Sun would hid Mercury, but maybe trying it when the first one is below the horizon it could be done (with the inconvenient that the atmospheric distortion would be very big at such low height, so not much detail will be got). But well, since when I write this post I have time at my disposal, I have the possibility of doing what he tells me in his tweets: check the image information. Raw images in FITS format have the so called "header" with all data about it. Among them you can find the following:

In the observation date you can see 2008-09-06, or 9/6 2008, at 19:07:30 local time (22:07:30 GMT). This can correspond either to June 9th or September 6th of that year. So, armed with any sky simulation software, one can check the positions of the planets at both dates. I tried to test with June 9th, from some place in Brazil (consistent both with his webpage and the difference between local time and GMT), and I saw how the shot would be done at nightfall, but oh! Mercury would be lower in the horizon than the Sun itself despite being quite close. Then I dug a little further… and to my surprise the planet is not in superior conjunction, but in inferior. With what consequence? No less than causing the face of Mercury visible from Earth to be the nocturnal one. The planet is between the Sun and us, and we don't see its light because its lit part is on the other side. It's impossible to take a picture of Mercury in these conditions. We try with the other date… and we don't have inferior conjunction either, but the planet is almost in its greatest elongation, the further from the Sun that it can be seen from Earth. Maybe the next image clarifies a bit all these positions:

As you can see in it, when an interior planet is in that position, we can only see half of the lit face. So if the image was taken in this second date (and this tweet points to this being the case), and Alberto Geyer's method did really reveal details of the bodies he observes, Mercury should appear with a half-moon look in the image. As I tell him, the round shape of the objects in their images is not there, but he forces it as part of the processing.

Meanwhile, in the Twitter conversation, he continues using as an argument that I don't know anything about astrophotography.

Solarview: @DarkSapiens U need your own telescope,a dark spot for observing and lots,I mean lots of hours imaging,many more hours processing 2 get there

Solarview: @DarkSapiens U need get started with astro-imaging.U obviously can't tell what's in focus or not before you comment on somebody else's work

DarkSapiens: @solarview All moons have the same curved shape as a result of being out of focus in that image. And if you take the full circle increasing>

DarkSapiens: @solarview > the cutoff, you end with moons much bigger than Earth compared to Jupiter.

I stop for a moment to explain this last comment from me. "Increase the cutoff" is the expression that stuck to my head using the astronomical image analysis software during the project in which I worked this year (of what I will talk in a future post), and it would consist of a way to increase the visibility of the image taking for the maximum luminosity value a slightly lower one than the value it had originally. You enhance the less bright areas while the ones already bright get saturated. In the Jupiter image, doing this you can see the full circular shape corresponding to the out of focus image of all its moons:

And if we continue forcing it, the uneven illumination in the out of focus zone is leveled until it forms a saturated circle from which every detail you could previously get has disappeared:

If we remember than the size of Jupiter is not bigger than in the first of these three images (close to the beginning of the post, where Jupiter is already saturated), we see than the resulting circles in this last one have approximately a fourth of its diameter. That is, if Alberto Geyer claims that those circles are the actual shape of the jovian moons, these have between 2 and 3 times the diameter of the Earth itself.

The discussion continues like this:

Solarview: @DarkSapiens Look Sapiens,there's a way to put an end to this discussion.I can send u the development procedure for any one of the imagesO.K

DarkSapiens: @solarview Let's see the Io one. I saw the sequence for Pluto and what you did was magnify the noise in the image and make the blob round.

Solarview: @DarkSapiens Eor dead wrong.I just imaged Pluto again and the mountain is looking better than never.I'll send the whole procedure via e-mail

Solarview: @DarkSapiens It's their real dimension.Maybe you don't know ,but Jupiter got closer to Earth on 08/14/10 than it was in many years.Go study!

Go study. Told to me that someone that seemingly just claimed that the moons of Jupiter, but not the planet, look bigger if both get closer to Earth. And by the way, if I have interpreted the date correctly as October 14th, 2008 (the only one that made sense, without being in the future nor needing a 14-month year), Jupiter was not as close from Earth as it could be (in opposition), but it was rather close to quadrature. The matter of Pluto has already been commented above.

I start to take more forceful measures.

DarkSapiens: @BadAstronomer @elakdawalla @plutokiller What do you think of this? Hard to make @solarview admit what he does is a fraud…

If only one of the three scientists to whom that tweet was directed had taken a look, Alberto Geyer would have seen himself under all their influence, orders of magnitud greater than mine. But with thousands of followers each, it was difficult that they could pay attention to this simple message.

DarkSapiens: @solarview I said big relative to Jupiter. The planet being closer has no effect on this.

Solarview: @DarkSapiens You obviously never took a shot a Jupiter!

No, I don't do astrophotography because I lack the means to. But I have seen many shots from some CIDAM member. Compare the size of the moons with that of Jupiter in this animation. I could also stress than to get good images of celestial planetary bodies a video is done for each frame, to then average the results and get rid as much as possible from the noise and atmospheric distortion. Just the opposite of what Alberto Geyer does.

Already seeing that his only answer is to deny the obvious, I decide to post links from his web together with the original images taken by the probes, as I warned I would do, so everybody can compare. Adding "@solarview" in the tweet, he would get the notice.

DarkSapiens: Busquen las diferencias! @solarview

Look for the differences!

Titan, the biggest moon of Saturn, taken from a false color image that combines certain wavelengths in the infrared with others in the ultraviolet. And he leaves the same colors even when they are very far from reality. At least he could had got an image in the visible range.

The asteroid Vesta from Hubble data. And I'm sure that instead of the image from the space telescope, he used the computer generated three-dimensional model as basis, losing even more credibility.

When he sees this, it seems that Geyer gets a bit nervous:

Solarview: @DarkSapiensDon't put words in my tweets Sapiens.If you have a problem with image processing advancements,keep it to yourself!
Solarview: @DarkSapiens No hay diferencias porque es el mismo objecto! 
There are no differences because it is the same object!
DarkSapiens: @solarview Which words?
Solarview: @DarkSapiensDe nuevo,busque la imagen original en SPACENOW y veras qui el objecto es el mismoPero si non crees sus ojos el problema es suyo 
Again, search for the original image in SPACENOW and you'll see that the object is the same. But if you don't believe your eyes the problem is yours.
Solarview: @DarkSapiens You know your own words.Why don't you send us yous best astrophotos?Let's sees the expert's work!

Now his tactics are to ask me to show astroimages made by me,  and I assume he's trying to discredit me by making me pass, again, for someone who has no idea of astrophotography. Due to the delay between the posting and reading of tweets, the conversation becomes a bit harder to follow. I continued linking to comparatives:

DarkSapiens: @solarview Do you mean the comparisons I'm posting? If I'm wrong it shouldn't be a problem I think.

DarkSapiens: @solarview …and it doesn't look the same from a probe than from Earth.

(That last tweet is in response to "there are no differences because it is the same object". They being the same object doesn't mean it has to look exactly the same from Earth than from a probe in its proximity)

Solarview: @DarkSapiens All you do is import pictures from other sources and pass it on to yet othersGo make your own images!Get going with astrophotos

Here either he doesn't understand why am I posting images from the probes or he tries to deviate attention. Of course I post images from other sources and not mine. If what I am trying to show is that the detail in his pictures comes from those images, what sense would it make to post pictures from another astrophotographer?

I, meanwhile, was still showing him reasons that supported my claims:

DarkSapiens: .@solarview BTW, for Titan you should have used a different image, not the IR + UV false color one…
DarkSapiens: .@solarview Oh, and I think you used Halley instead of Kleopatra :S vs
Solarview: @DarkSapiens If you take the trouble to download the Kleopatra raw image u'll see that it stands out even in the midle of the starfield

Being honest, I didn't download that image. I don't know if he refers to the asteroid stands out (a thing that, if it's only a point of light, wouldn't have much importance), or that its shape is visible. But a Google images search reveals it is hard to image. Its shape can be inferred by radar, but taking a picture using traditional means is complicated.

They look the same…
DarkSapiens: @solarview O más bien… no hay diferencias porque es la misma foto ;)
Or rather… there are no differences because it is the same picture ;)

The Iapetus image has even some of the smallest details taken from that image, that is only rotated. The second tweet is a better response to the "there are no differences because it is the same object" thing.

He continues with his argument of me not knowing about astroimaging:

Solarview: @DarkSapiens Porque non empiecas con un curso en astrophotografia.Tienes mucho que aprender.Donde estan su photos?

Why don't you start with a course in astrophotography. You have a lot to learn. Where are your pictures?
DarkSapiens: @solarview Me not taking astrophotos doesn't mean I don't know how they're done, poor excuse.
DarkSapiens: .@solarview I'm not "importing pictures", I'm just linking to the images you probably used, taken by Voyager, Galileo and Cassini probes.
Solarview: @DarkSapiens Then tell me what's a flat field,a bias, and a dark subtraction. Right off the bat. Don't consult your older friends.
Solarview: @DarkSapiens Don't acuse others without knowing .

Here, now it seems that he wants to put me to the test. Even knowing that it wouldn't make a big difference (one could perfectly look it up in the Internet in the time it takes to send a reply), I decide to be honest in my response and I reply by heart:

DarkSapiens: @solarview They're images taken for post-processing, in order to calibrate the picture and get rid of CCD or ambient issues.

DarkSapiens: @solarview You substract the Dark because if not, the background level could be not black thus affecting photometry and other things.

DarkSapiens: @solarview You take the bias image to get rid of a bias introduced by the CCD, again affecting photometry in each pixel.

Solarview: @DarkSapiens That was close.A serious astro-imager doesn't subtract these shots on a post processing.They're done at picture taking time.

Let's see. The darks and bias are images made with the camera covered so you only get the alterations and noise created by the instrumental, to precisely be able to eliminate it later from the photograph. They are separated images. I don't know how are you going to subtract them from the picture if this one is not taken yet. And once all images are taken, what influence would it have to make the processing (subtracting them) the same night or months after taken? They are computer files. When are they used is irrelevant. To know more about these shots I found some days ago this easy article (in Spanish), quite explanatory.

DarkSapiens: @solarview The darks are taken when you take the picture, but you can substract them later.

Solarview: @DarkSapiens Oh yeh?And what about the flat field?what is it for ?

DarkSapiens: @solarview Oh, didn't see the flat thing. You take it to get rid of uneven illumination caused by the telescope, lenses, filters, etc.

Solarview: @DarkSapiens Now that you just read the theory,go out there and try it out.After you show me your own work,not NASA's,come back and we talk

As it was probable to happen, he already knew I could easily read instead of explain by heart these things. And it seemed he was not going to admit in public that I did know about astrophotography. He continues using this as his main argument and he insists that I have to show him my work and not NASA's. I don't see any sense to this other than deviate the subject, since the only images we were talking about here were his, made from the ones by the space agency. What I could picture or not doesn't matter at all. And that's what I tell him:

DarkSapiens: .@solarview LOL, I wrote, not read. And if I'm saying you use NASA images, why should I show my photos if they have nothing to do here?

Solarview: @DarkSapiens Because you never took a shot at the heavens.That's why Get out and get a life.Quit the computer screen.Get experience.Goodbye

DarkSapiens: .@solarview Haha, I was with my astronomy association yesterday, btw. So you can't prove your images are not from NASA, then?

It is funny that in his last comment he treated me now as if I had never looked at the sky, when precisely the day before I had spent the night doing astronomical observation with my association. He never replied again to any of my questions. It is possible that to cut all contact he blocked me, and then as I describe in the beginning he deleted everything that referred to this conversation.

If you have continued reading until here this long post you may ask the question of why do I bother to do all this rebuttal work. First, I could refer to the name of this blog, and claim that as other bloggers uncover the lies of homeopathy, of lunar conspiracy theories, or every other fraud, here I was faced with one directly related with my main area of knowledge: astronomy. Even when my goal is to dedicate myself to it as a professional, I don't dedicate to it all the space I would like in this blog, and this could be also an opportunity to explain some concepts. Second, the reason to have saved all this conversation was to make it public and accessible, especially in case it was eliminated (as it can be checked clicking in the links for @solarview comments, these don't lead to much, although Google's cache saves two small fragments [UPDATE september 17th 2010: Well, at least it did when I wrote this post…]).

Another important trigger, and in fact what made me decide completely to write this post, was when I saw that apparently he hoped to be present as a speaker in the TEDx event organized in the Rio de Janeiro planetarium sending this video. Mysteriously, it stopped loading a few days ago (the ones from all the other candidates still work), but in it some parts of the procedure done to get the image of Io were shown, and you could see how it consisted just in saturating the image of the moon as I show above and increasing the contrast to round its shape more. Then the process jumps directly to an image already with the color in the NASA image included and in which you saw how messing a bit with Photoshop he starts getting more details from it, but it doesn't show what happens between one step and the other. [13/8/2010: UPDATE! Geyer has uploaded full on his website the video for the Io processing, that the one I mention before has as background. You can see what I mention in this paragraph, although he gets the color from nowhere actually, modifying at a guess the levels histogram many times. That is, where before there was only black and white, he adds color gradually until it looks to him sufficiently similar to the images… that he has seen from the NASA probes] In the video, Geyer, a man seemingly between thirty and forty years old, insists that his is a "cheap and effective" method for the research of the Solar System, that almost everyone could implement. Imagine if all this was true. It would be an authentic revolution.

But in the light of the evidence, I wonder: what is really his aim? His now disappeared video, despite having a huge lot of negative votes, has some 18 comments supporting and admiring him, even offering him funding to develop his project. The number of followers of his twitter has nearly doubled since he added me. Is it fame and admirers what he seeks? Being recognized by the people who doesn't have enough knowledge to realize that everything is a fraud? In his claims he seem to be convinced that his method is a big advance, and even wanting to share his "technique" with professionals. Could this be true? He must have manipulated the pictures to get the final result, data from Hubble telescope have allowed to construct approximate maps of Pluto and there's no sign of the gigantic mountain, that will be absent when New Horizons flies by the dwarf planet in 2015. Could he be convinced that what he does really works? Is it possible for levels of self-delusion of one person to reach those extremes?

Whatever the motivation, there are reasons for which something like this shouldn't be left without a categorical reply. The amateur astronomers community has been some time fighting to be recognized among professionals as a very good complement for research. And a lot is being achieved: from following and discovering asteroids to alerting of new supernovae, monitoring possible changes in Mars, Saturn and Jupiter (like the disappearing of one of his belts or even, today, of a new impact with some wanderer object), amateur astronomers are contributing with valuable observations. But amateurs like Alberto Geyer still make the level of distrust not to be low. I know other cases in which some amateur astronomer thinks he has a level of knowledge quite superior to the real one and he tries to explain to laypersons in the matter concepts that are erroneous, or who think to have made an important discovery that when it is dismissed by professionals as a false alarm, they adopt the role of misunderstood people.

But the most important, to my judgement, is the following: if people think it is so easy to study the Solar System in such a cheap way, why would they support the public investment in multimillion-dollar missions that wouldn't make such a big difference with what can be done for an extremely low fraction of its cost? If the spending in space exploration is perceived as something even less important than many people think, the fight for a needed increase in these budgets would have one more burden. And I have already talked about how important this investment is.

Adding any extra difficulty to the funding of these kind of projects is an action, at least to my judgement, quite reprehensible.

17/09/2010 UPDATE: Taking a look at the blog statistics I found that someone in the forums found Geyer's work and they were discussing it, and they linked to the Spanish version of this post on the issue. They are still debating about it there, and this motivated me to translate this post to English, something I had still left to do.