Fighting Bad Astronomy with Bad Astronomy

Bad. Guardian · #1 (**permalink**) 21-December-2007, 02:49 AM

This is mainly a question of ethics.

Dr. Plait, how would you handle the following hypothetical scenario? You have a press conference, at which you debunk some sort of Bad Astronomy. Hours later, you discover that the data that you presented at the press conference were bogus. Someone on your staff transcribed them incorrectly. The true data, though they don't exactly support the BA that you were trying to debunk, are much less harmful to it than the bogus data that you presented. What will you do? Will you call another press conference, and apologize for the mistake? If you do this, your retraction may bolster the same BA that you were trying to debunk. Will you choose to skip the apology, and just keep quiet, and hope that no one notices that your data were bogus? This would surely be a violation of scientific ethics, and in the long run this would probably do more harm than good. It's a painful decision to make, but I think we all can agree which choice would be the right choice.

While browsing recently, I came across a number of angry web pages which argued that a certain image of the "face" on Mars, in a certain JPL press release, was a fraud. To me, personally, the "face" has always seemed to be an accident of nature. But when I encountered this controversy a few days ago, I saw it as an opportunity for a fascinating mathematical challenge: an attempt to determine the exact algorithm that was used to "flatten" the landform in the image. I tried some Fourier tricks, and succeeded. The algorithm is a very simple 1x31 convolution, which resembles a crude and grossly exaggerated unsharp mask. It produces all of the expected effects: it eliminates vertical streaks, and destroys real 3D shading cues, and creates false shading cues, and gives the surrounding flat land a "grainy" texture, and creates "false shadows" above and below a small bright feature near the east edge of the image.

I experimented with similar convolution kernels in varying sizes, and it does indeed seem that the size 1x31 is particularly well suited to the purpose of deceiving the viewer about the true 3D shape of this particular hill.

Was the deception deliberate, or accidental? I don't know.

If the people who issued the press release were fooled by their own deceptive image processing, and really believed that the "face" was not actually a hill, as it appeared in the Viking images, but instead was a jumble of boulders encircled by a trough, as it appeared in their press release image, then their mistake was an honest mistake. But even if they were honestly ignorant of the problem, they could not have remained ignorant for long. Tim J. Parker's enhancement was released that same day, and anyone who looked at both images could see that the processing of the original image had been deceptive-- so deceptive, in fact, that it could reasonably be expected to inspire conspiracy theories.

The right thing to do, at that point, would have been to issue another press release, retracting the bogus image, and apologizing for the mistake. Someone decided not to do the right thing. This, then, would be two individual acts of Bad Astronomy: first, the original, possibly accidental, use of bogus data in a press release; second, the decision to refrain from issuing an apology and retraction, when they learned that an image in their press release was bogus.

If this story has a moral, then I suppose it must be something like this: when you attack Bad Astronomy, be conscientious, and fight fair. If you cheat, then you may get caught, which will only make the situation worse. If you innocently use bogus data by accident, then you'll have to issue an apology and retraction, which will only make the situation worse. If you accidentally use bogus data, but then you deliberately refuse to issue an apology and retraction, then... well, then we're back to cheating; see above.

--------
Update, 1/15/2008: I've uploaded the full-size animation, with no cropping or scaling, to an image host which doesn't have a file size limit:

(catbconv.gif, 512x480 pixels, 12 frames, 3,193,898 bytes, sha1sum 6b4d1f23915d3342e5888363d409fdaa97b82f17)

http://s2.supload.com/free/catbconv-...3559.gif/view/

--------
The text of this message is released into the public domain.
Attachment: an animation which shows the convolution in action.

Bad. Guardian · #2 (**permalink**) 21-December-2007, 03:26 AM

Quote:

Originally Posted by Bad. Guardian

Attachment: an animation which shows the convolution in action.

I had to shrink the animation from 512x480 to 112x105, so the forum server would accept it. I'm attaching a script which will allow you to see the full size animation. If you don't intend to run the script, and if you don't like to read source code, you may still want to view the script, and read some of the comments.

Wait... the forum server thinks that ASCII text is an unacceptable attachment type! OK, I'll just paste the script here.

Code:

#!/usr/bin/scsh -s

WARNING: this script can destroy files in the current working directory.
Run it in a directory which you've created specifically for this script.

This script is released into the public domain.
First publication:
http://www.bautforum.com/bad-astronomy-stories/68345-fighting-bad-astronomy-bad-astronomy.html

This script requires scsh and netpbm and gifsicle.  If you're running
Debian 4.0, "aptitude install scsh netpbm gifsicle" will install what
you need.  (You might also want to install scsh-doc.)

Start with cydonia1.tiff in the current working directory.

  cydonia1.tiff sources:
  ftp://nssdcftp.gsfc.nasa.gov/miscellaneous/planetary/mgs/cydonia1.tiff
    (sha1sum: 611771f7cec388c7640f6110eba8aa0012379ab1  cydonia1.tiff)
  http://windsor.gsfc.nasa.gov/miscellaneous/planetary/mgs/cydonia1.tiff
    (sha1sum: 611771f7cec388c7640f6110eba8aa0012379ab1  cydonia1.tiff)
  http://mpfwww.jpl.nasa.gov/mgs/target/CYD1/cydonia1.tif
    (sha1sum: 9d9204ad231666ab0a0c3ecca65cd64e4c28a7da  cydonia1.tif)

  cydonia1.tiff information:
  http://nssdc.gsfc.nasa.gov/planetary/mgs_cydonia.html
  http://mpfwww.jpl.nasa.gov/mgs/target/CYD1/

This script generates an animation which demonstrates the effects of the
following algorithm:  Choose a convolution kernel height, and call it h.
Create a matrix with h rows and 1 column.  Set the middle element of the
matrix to 1, and set all the other elements to 1/(1-h).  Convolve each
column in the image using this matrix.  Normalize contrast.

In the original "catbox convolution," h was 31.  The image had been scaled
down, to a swath width of 512 pixels, before the convolution was applied.

In the limit, as h approaches infinity, the algorithm above becomes
equivalent to the following even simpler algorithm:  Subtract, from each
pixel, the average of all pixels in the same column.  Normalize contrast.

!#

;;; This function creates a convolution kernel for pnmconvol.  This is not the
;;; catbox kernel, because unfortunately pnmconvol isn't general enough to use
;;; it directly.  Instead this is the kernel you would get if you started with
;;; the catbox kernel, and multiplied the middle row by 0, and multiplied every
;;; other row by -1.  Later we'll undo the multiplication by -1 (using
;;; pnminvert) and undo the multiplication by 0 (using ppmmix).

;;; to do: a simple scsh function could perform the catbox convolution
;;; directly, without needing pnmconvol and pnminvert and ppmmix.  And it could
;;; be faster, because it could take advantage of the fact that nearly all of
;;; the elements in the kernel have the same value.

(define (make-kernel height)
  (format #t "P2 1 ~d ~d~%" height (* (- height 1) 2))
  (let ((f (lambda ()
             (do ((i (/ (- height 1) 2) (- i 1)))
                 ((<= i 0))
               (format #t "~d~%" height)))))
    (f)
    (format #t "~d~%" (- height 1))
    (f)))


;;; to do: instead of redirecting output to these silly throw-away files,
;;; use pipelines.  Scsh and netpbm both work beautifully with pipelines.

(define (prelim)
  (&& (epf (tifftopnm cydonia1.tiff)
           (> a.ppm))
      (epf (ppmtopgm a.ppm)
           (> b.pgm))
      (epf (pnmdepth 65535 b.pgm)
           (> c.pgm))
      (epf (pnmscale -reduce 2 c.pgm)
           (> d.pgm))
      (epf (pnmflip -lr d.pgm)
           (> e.pgm))
      (epf (pamcut -top 1769 -height 1440 e.pgm)
           (> f.pgm))))


(define (per-height height)
  (let ((subtrahend (format #f "~a.pgm" height)))
    (and (cond ((number? height) ; not infinity
                (&& (epf (begin (make-kernel height))
                         (> k.pgm))
                    (epf (pnmconvol k.pgm f.pgm)
                         (> ,subtrahend))))
               (else ; infinity
                (&& (epf (pnmscale -width 512 -height 1 e.pgm)
                         (> l.pgm))
                    (epf (pnmscale -width 512 -height 1440 l.pgm)
                         (> ,subtrahend)))))
         (&& (epf (pnminvert ,subtrahend)
                  (> m.pgm))
             (epf (ppmmix "0.5" f.pgm m.pgm)
                  (> n.ppm))
             (epf (ppmtopgm n.ppm)
                  (> o.pgm))
             (epf (pamcut -top 480 -height 480 o.pgm)
                  (> p.pgm))
             (epf (pnmnorm -bpercent "0.005" -wpercent "0.005" p.pgm)
                  (> q.pgm))
             (epf (pnmdepth 255 q.pgm)
                  (> r.pgm))
             (epf (ppmlabel -colour white -x 239 -y 475 -text
                            ,(if (number? height) height "inf.") r.pgm)
                  (> s.ppm))
             (epf (ppmtogif s.ppm)
                  (> ,(format #f "~a.gif" height)))))))


(define (list-heights)
  (do ((n 0 (- n 1/2))
       (l '("inf")
          (cons (+ 1 (* 2 (inexact->exact (round (* #i480 (expt 2 n)))))) l)))
      ((< n -5) l)))


(define (gifsicle-args height)
  (list "--name" height (format #f "~a.gif" height)))


(let ((heights '(31 43 61 85 121 171 241 341 481 679 961 "inf")))
  (and (prelim)
       (every per-height heights)
       (run (gifsicle -l -V
                      -d165 ,@(gifsicle-args (car (reverse heights)))
                      -d33 ,@(apply append (map gifsicle-args
                                                (cdr (reverse (cdr heights)))))
                      -d165 ,@(gifsicle-args (car heights)))
            (> catbconv.gif))))

JonClarke · #3 (**permalink**) 21-December-2007, 05:53 AM

Quote:

Originally Posted by Bad. Guardian

While browsing recently, I came across a number of angry web pages which argued that a certain image of the "face" on Mars, in a certain JPL press release, was a fraud. To me, personally, the "face" has always seemed to be an accident of nature.

Are you referring to the so-called "catbox" image?

Quote:

Originally Posted by Bad. Guardian

But when I encountered this controversy a few days ago, I saw it as an opportunity for a fascinating mathematical challenge: an attempt to determine the exact algorithm that was used to "flatten" the landform in the image. I tried some Fourier tricks, and succeeded. The algorithm is a very simple 1x31 convolution, which resembles a crude and grossly exaggerated unsharp mask. It produces all of the expected effects: it eliminates vertical streaks, and and gives the surrounding flat land a "grainy" texture, and creates "false shadows" above and below a small bright feature near the east edge of the image

What is the basis for stating that the image was "flattened" (other than "angry web pages")? It does not look flattened to me.

Eliminating artefacts like vertical streaks is why images are processed. Is this bad?

Whaat is the evidence that "real 3D" shading cues have been destroyed and "false" shading cues and shadows created?

Whaat evidence you you have that the surrounding flat land has been given a "grainy" texture?

Are you aware that this image, taken under poor conditions (off axis, and with haze and cloud) still show much more detail than the best Viking image?

Quote:

Originally Posted by Bad. Guardian

I experimented with similar convolution kernels in varying sizes, and it does indeed seem that the size 1x31 is particularly well suited to the purpose of deceiving the viewer about the true 3D shape of this particular hill.

Was the deception deliberate, or accidental? I don't know.

Whaat is your evidence that there has been any deception at all? The processing of this image is a matter of public record.

Quote:

Originally Posted by Bad. Guardian

If the people who issued the press release were fooled by their own deceptive image processing, and really believed that the "face" was not actually a hill, as it appeared in the Viking images, but instead was a jumble of boulders encircled by a trough, as it appeared in their press release image, then their mistake was an honest mistake. But even if they were honestly ignorant of the problem, they could not have remained ignorant for long. Tim J. Parker's enhancement was released that same day, and anyone who looked at both images could see that the processing of the original image had been deceptive-- so deceptive, in fact, that it could reasonably be expected to inspire conspiracy theories.

What is your evidence that the people who issued the press released were fooled?

What is your evidence that they did not think it was a hill, but a "jumble of boulders surrounded by a trough"? To the contrary, MSS have repeatedly called it a hill.

Tim Parker's image looks a bit different because it was processed slightly differently. This is normal. The features in both images are the same.

Quote:

Originally Posted by Bad. Guardian

The right thing to do, at that point, would have been to issue another press release, retracting the bogus image, and apologizing for the mistake. Someone decided not to do the right thing. This, then, would be two individual acts of Bad Astronomy: first, the original, possibly accidental, use of bogus data in a press release; second, the decision to refrain from issuing an apology and retraction, when they learned that an image in their press release was bogus.

So far you have not shown any evidence that the MSSS release was bogus. So there is no need for apology and retraction.

Jon

Bad. Guardian · #4 (**permalink**) 21-December-2007, 06:38 AM

Sorry, I guess I should've provided a link to the original deceptive image.

Quote:

Originally Posted by JonClarke

Are you referring to the so-called "catbox" image?

Hi Jon, thank you for replying. Yes, I was referring to the so-called "catbox" image, but... No, I wasn't referring to the image to which you've linked the word "catbox" above. I believe that you've linked to an enhancement which was done later, and which is noticeably different from the so-called "catbox" image that was used in the press release.

I was referring to this image: http://mpfwww.jpl.nasa.gov/mgs/targe...ydonia1c-s.gif, which I found on this web page: http://mpfwww.jpl.nasa.gov/mgs/target/CYD1/. This is the so-called "catbox" image that was used in the press release.

I'll be glad to answer your other questions, but I'll wait until we're both looking at the same image, and then we'll see how many of those questions you'll still want to ask. I think that you'll agree, when you're looking at the same image, that it exhibits the characteristics that I mentioned.

Bad. Guardian · #5 (**permalink**) 21-December-2007, 06:48 AM

Quote:

Originally Posted by JonClarke

Are you referring to the so-called "catbox" image?

Hi Jon, thank you for replying. I posted an answer, but it's been held for moderation, probably because it included links. You'll probably see it eventually, but for now I'll just give you this capsule summary: you're looking at the wrong image. The image to which you've linked the word "catbox" above is not the so-called "catbox" image.

JonClarke · #6 (**permalink**) 21-December-2007, 10:24 PM

http://www.msss.com/mars_images/moc/..._face_release/

I'll be glad to answer your other questions, but I'll wait until we're both looking at the same image, and then we'll see how many of those questions you'll still want to ask. I think that you'll agree, when you're looking at the same image, that it exhibits the characteristics that I mentioned.[/quote]

Thanks for linking to the image you meant. So we have three differentially processed versions of the same raw image. One by parker, one by MIPL, one by MSSS. Each is slightly different, which is what happens when different people process the same data differently. But each is clearly the same original image and shows the same features.

You realise that all in the metadata for all the processing is in the links we have both provided? You don't have to reconstruct what you think was done, it is all there. And it is different to what you said was done.

And there is no evidence of deliberate or accidental deception in this.

Jon

Jerry · #7 (**permalink**) 21-December-2007, 10:31 PM

It is an unfortunate fact that when scientific data does not seem to mesh with scientific theory; it is often withheld from publications or qualified in the footnotes. As long as the data is being withheld because of uncertainty, this is reasonable, but only if there is an effort underway to verify whether or not the data can be better qualified.

If data is withheld because it may throw doubt upon prevailing theories and sparkle the eyes of unconventional theorists, then it is just bad science. We have an obligation to report our failures as well as our successes. Throwing out data that does not agree with prevailing theories may 1) include the baby with the bathwater or 2) inspire more confidence in a prevailing theory than is reasonable; given the iffy nature of the data.

JonClarke · #8 (**permalink**) 21-December-2007, 10:50 PM

Quote:

Originally Posted by Jerry

It is an unfortunate fact that when scientific data does not seem to mesh with scientific theory; it is often withheld from publications or qualified in the footnotes. As long as the data is being withheld because of uncertainty, this is reasonable, but only if there is an effort underway to verify whether or not the data can be better qualified.

If data is withheld because it may throw doubt upon prevailing theories and sparkle the eyes of unconventional theorists, then it is just bad science. We have an obligation to report our failures as well as our successes. Throwing out data that does not agree with prevailing theories may 1) include the baby with the bathwater or 2) inspire more confidence in a prevailing theory than is reasonable; given the iffy nature of the data.

And the relevance of this to the case in point is?????

JonClarke · #9 (**permalink**) 21-December-2007, 10:55 PM

To see how unface like the "face" really is in the Viking data when processing exposes what is in the shadowed area, look here.

http://www.msss.com/education/facepage/face.html

Jerry · #10 (**permalink**) 22-December-2007, 12:21 AM

Quote:

Originally Posted by JonClarke

And the relevance of this to the case in point is?????

Bad Guardian is asking if it is reasonable to withhold or not report erroneous data reduction when doing so may spark a new round of drum beating by crackpots. I say it is not; for the two reasons in my post above, but also because any time you make the judgement call: It is better to withhold or not correct erroneous data than muddy the water, you are flirting with conspiracy - or at the very least, giving crack pots more ammunition if the omitted information comes to light for other reasons.

NASA did not publish the fact Spirit was within a whisker of failure when it safely landed on Mars. I don't know what the margins were for Opportunity, but I would like to.

JonClarke · #11 (**permalink**) 22-December-2007, 01:49 AM

Quote:

Originally Posted by Jerry

Bad Guardian is asking if it is reasonable to withhold or not report erroneous data reduction when doing so may spark a new round of drum beating by crackpots. I say it is not; for the two reasons in my post above, but also because any time you make the judgement call: It is better to withhold or not correct erroneous data than muddy the water, you are flirting with conspiracy - or at the very least, giving crack pots more ammunition if the omitted information comes to light for other reasons.

Bad Guardian has posted no evidence to support his claim.

Quote:

Originally Posted by Jerry

NASA did not publish the fact Spirit was within a whisker of failure when it safely landed on Mars. I don't know what the margins were for Opportunity, but I would like to.

Yes they did, otherwise how else would it be known? And this has nothing to do with Bad Guardian's claims about the first MOC image of Cydonia. I hope you are not trying to hijack this thread.

Jon

Bad. Guardian · #12 (**permalink**) 22-December-2007, 06:34 AM

Quote:

Originally Posted by JonClarke

Bad Guardian has posted no evidence to support his claim.

Which claim?

This discussion has suffered some confusion, probably because we were discussing different images.

Which claim do you think requires additional evidence?

To reduce the confusion, if you don't mind, I'll summarize the argument in numbered paragraphs, and you can tell us which parts you think require additional evidence.

1. The landform in question is a hill.

2. The image in the press release was processed badly. Very, very badly. Normal 3D shading cues were aggressively suppressed-- so aggressively that the hill doesn't even look like a hill!

3. The use of bogus data in a press release, whether accidental or deliberate, is Bad Astronomy.

4. If the people who issued the press release had somehow fooled themselves, and honestly thought that their image was an accurate depiction of the shape of the landform, then they certainly must have learned very quickly-- within a few days, at most-- that this was not the case.

5. Deciding not to issue a retraction, when you learn that you've used bogus data in a press release, is Bad Astronomy.

6. Conclusion: Fight fair. Don't cut corners, don't cheat.

Quote:

Originally Posted by JonClarke

I hope you are not trying to hijack this thread.

I don't think Jerry was trying to hijack anything. He was only trying to explain the ethical principle that I was stating. I could try to state it more clearly, but instead I'll let Feynman say it, because he said it best:

Quote:

It's a kind of scientific integrity, a principle of scientific thought that corresponds to a kind of utter honesty-- a kind of leaning over backwards. For example, if you're doing an experiment, you should report everything that you think might make it invalid-- not only what you think is right about it: other causes that could possibly explain your results; and things you thought of that you've eliminated by some other experiment, and how they worked-- to make sure the other fellow can tell they have been eliminated.

Details that could throw doubt on your interpretation must be given, if you know them. You must do the best you can-- if you know anything at all wrong, or possibly wrong-- to explain it. If you make a theory, for example, and advertise it, or put it out, then you must also put down all the facts that disagree with it, as well as those that agree with it.

JonClarke · #13 (**permalink**) 22-December-2007, 08:58 AM

Quote:

Originally Posted by Bad. Guardian

To reduce the confusion, if you don't mind, I'll summarize the argument in numbered paragraphs, and you can tell us which parts you think require additional evidence.

1. The landform in question is a hill.

Nobody has disputed this, except Hoagland and his followers.

Quote:

Originally Posted by Bad. Guardian

2. The image in the press release was processed badly. Very, very badly. Normal 3D shading cues were aggressively suppressed-- so aggressively that the hill doesn't even look like a hill!

First of all what you see as evidence of "aggressively suppressed" 3D cues are in fact a reflection of the image being taken under suboptimal conditions - hazy aatmosphere, of axis imaging, and unfavourable light aangles.

As for the hill not looking like a hill, that is your opinion. I can see considerable relief in that image.

Quote:

Originally Posted by Bad. Guardian

3. The use of bogus data in a press release, whether accidental or deliberate, is Bad Astronomy.

You have not established that the data is bogus. The image is real, a range of standard processing techniques were used. What is bogus about this? You realise that accusing JPL and MSSS of releasing false data is very serious charge, in fact you are accusing them of something that is not only unethical, but contrary to their legal obligations?

What aspect of this image is invalidated by later images by MOC, THEMIS, HRSC and HiRISE?

Quote:

Originally Posted by Bad. Guardian

4. If the people who issued the press release had somehow fooled themselves, and honestly thought that their image was an accurate depiction of the shape of the landform, then they certainly must have learned very quickly-- within a few days, at most-- that this was not the case.

In what way does this image not accurately reflect the landform in question?

Again, you are making a servious accustation of dishonest, illegal behaviour, which you should substantiate, of withdraw.

Quote:

Originally Posted by Bad. Guardian

5. Deciding not to issue a retraction, when you learn that you've used bogus data in a press release, is Bad Astronomy.

Why should they retract the best image available, processed by experts? The MIPL are one of the most experienced image interpretation teams on the planet? MSSS built the system, Tim Parker is a very experienced planetary scientist. Despite using different processing their images are very similar.

Incidently the initial press release contained on error, with respect to the image size. This was corrected the following day. This is evidence of the intergrity of the people who made it.

Quote:

Originally Posted by Bad. Guardian

6. Conclusion: Fight fair. Don't cut corners, don't cheat.

So far you have not produced any evidence that there was any unfair fighting, corner cutting, and cheating. Unless you do so, you should with draw this false accusation.

Quote:

Originally Posted by Bad. Guardian

I don't think Jerry was trying to hijack anything. He was only trying to explain the ethical principle that I was stating. I could try to state it more clearly, but instead I'll let Feynman say it...

Jerry has a record of long convoluted discussions aabout the entry profiles of space probes. I would hate to see this discussion being hijacked into another one.

As for the Feynman quote, let's look at it.

It's a kind of scientific integrity, a principle of scientific thought that corresponds to a kind of utter honesty-- a kind of leaning over backwards. For example, if you're doing an experiment, you should report everything that you think might make it invalid-- not only what you think is right about it: other causes that could possibly explain your results; and things you thought of that you've eliminated by some other experiment, and how they worked-- to make sure the other fellow can tell they have been eliminated.

Let's see - the full details of the conditions under which the image was taken were reported, the camera specifications are available to all interested people, the processing done on the image were stated.

Details that could throw doubt on your interpretation must be given, if you know them. You must do the best you can-- if you know anything at all wrong, or possibly wrong-- to explain it. If you make a theory, for example, and advertise it, or put it out, then you must also put down all the facts that disagree with it, as well as those that agree with it.

None of the statements by MIPL, Parker or MSSS that were released on the 6th of April 1998 contained much in the way of interpretation. The describe the images and the processing. The interpretation is left to the viewer

Jon

Bad. Guardian · #14 (**permalink**) 22-December-2007, 11:23 AM

Quote:

Originally Posted by JonClarke

First of all what you see as evidence of "aggressively suppressed" 3D cues are in fact a reflection of the image being taken under suboptimal conditions - hazy aatmosphere, of axis imaging, and unfavourable light aangles.

Jon, either you're still looking at the wrong image, or you've been misunderstanding everything I've been writing.

I've been writing about the processing that was applied to the image _after_ it was sent back to Earth. Normal 3D shading cues, which were _present_ in the image that was returned by the spacecraft, were _absent_ in the image that was used in the press release.

Please look at the animation attached to this message. One frame is the image that was returned by the spacecraft (with left/right inversion corrected, and contrast normalized) and the other frame is the image that was used in the press release. I've applied a simple streak removal algorithm, and 50% scale reduction, to both frames.

When you watch this animation, I'm sure that you'll see how the normal 3D shading cues have been suppressed.

--------
Update, 1/15/2008: I've uploaded a full-size version of this animation, with labels identifying the frames, to an image host which doesn't have a file size limit:

(before-after.gif, 512x480 pixels, 2 frames, 532,569 bytes, sha1sum 6020a9d54b876cff135ab8fd09bff0a5668f1e2e)

http://s2.supload.com/free/before-af...3559.gif/view/

.