Strange streak discussion: 2004 Dec 7 APOD

[Guest]

I wish this forum had a spellchecker built in.

So dou AI!

[Guest]

The EXIF information (one shown below) ALL show FLASH USED as YES.

EXIF
Camera
Manufacturer = Canon
Model = Canon PowerShot G3
Dimensions
Width = 2272
Height = 1704
Advanced
Focal Length = 9.1mm
Focal Length (34mm equivalent) = 1169mm
…..

Black and white = no
Flash Used = Yes
Encoding = Baseline

What is the actual focal length used to take the picture? 9.1mm or 1169mm?

[RAJ]

When quickly swithcing between photos, I don't see any reflected highlights from a flash on objects in the vicinity of the pier, just an overall increase in illumination due to the growing size of the reflective cloud in the background. I think this negates the theory of a lumious event occuring in that area. I'll take the bug theory.

[victorengel]

The bright spot does not have any blurring due to motion of the hypothetical bug during duration of the flash. But the bug had to move almost across the whole picture for 0.05s. So my estimate of "possible blurring/total path" ratio is <~ 0.001 . So in the bug theory this imposes an upper limit on the duration of the flash as ~0.05s/1000 . Could somebody find out duration of the flash (technical info)?

I think it depends upon the intensity of the flash, which information we don't have, as far as I know.

[Luis]

Go here to see some work I did on sorting out a sense of scale for this debate. I think the 'bug' theory is up in smoke, if you'll pardon the jab.
http://homepage.mac.com/b.a.freeman/Dar ... bum14.html

Brett,

Can you provide an idea of the distance from the camera to the pole in question? 1km, 100m? I'm not remotly familiar with docks...

Thanks
Luis

[CurtC]

Well, I'm 99% convinced that it's simply a bug. So far, I haven't heard any objections to the hypothesis that could rule it out.

I don't have image processing tools, but I think an interesting thing to try would be to rotate the streak horizontally, then do an FFT of the area near the dark streak, across the image, to look for any periodicity that would have been caused by the wings. Someone earlier did this with a histogram (and found a 3 pixel periodicity), but an FFT would be much better at picking this out.

[PhD Man]

What is interesting about this exercise is that the ground rules are assumed to be:

1. APOD Editors are savvy enough to have ruled out that this digital image is not 'created' - which, in my esteemed opinion, is not possible when dealing with minimally semi-crafty enough 'artists'. It does not take much to create an image like this while maintaining appropriate associated data. I know anecdotally... I am an expert in this, among other, fields. The individual that captured this image is a "computer person" in case you didn't know.

2. The individual is assumed to be completely honest and without a trace of trickery (you are very trusting souls, aren't you?) or mischieve.

3. Image manipulation or camera artifact has to be ruled out; yet this is the easiest, most logical, most likely from a probabilistic point of view.

Given that, it has even been suggested that this is pure game and fun.
And so it may be for some.

But whole premise of this exercise is then skewed and faulty from the start. The purpose then seems to be: "Think up goofy stuff while purposefully ignoring reality or likely causes." OK, this is fun. This is a game. But I didn't see that stated anywhere. And of course, my question is: why are so many working so hard on trying to find a solution then?

I like the idea of voting on candidates. That would be 'fun'. But as far as so many people devoting such efforts, to a 'game'... well, wouldn't it be more beneficial, and fun, to tossing out the pre-assumptions that we rule out artifacts or manipulations and instead devote those same efforts to really trying to get down to the truth in this? What we call 'science'. Or is science 'out' and flights of fancy 'in'? I don't want to know, thank you.

If anyone takes this as abrasive or challenging... it IS a challenge. The abrasiveness is in your own issue. Cheers!

[victorengel]

Phd Man,

Do you have any evidence to support your ideas? The idea that this is a fraud has been proposed before, but so far I've seen no evidence to back it up. Just as with any other theory, that one requires something, I think.

[tonycc]

2. Why has nobody answered how the same trail appears somewhat wider and much fainter in the actual before picture? (not the picture that is linked "before") Look at this animation that Diffit posted and you can see a very faint trail quite clearly.

I believe what you are seeing is a residual after image frm the light trail in the diff picture. I can see it also. Try this, look at one side of the frame during the cycle. Then, when the after picture comes up look at the trail area. It is no longer there when I do this.

Tony

[tonycc]

2. Why has nobody answered how the same trail appears somewhat wider and much fainter in the actual before picture? (not the picture that is linked "before") Look at this animation that Diffit posted and you can see a very faint trail quite clearly.

I believe what you are seeing is a residual after image frm the light trail in the diff picture. I can see it also. Try this, look at one side of the frame during the cycle. Then, when the after picture comes up look at the trail area. It is no longer there when I do this.

Tony

[BOb]

Aliens did it

[caseyoconnell.]

Nice, PhD Man.

I rather like the way the discussion went from totally unscientific and unconsidered musings to careful analysis.

I saw some great visual effects especially the bats picture and the snow picture. This stuff will be useful for further consideration.

I had been considering that the bug was flying down right until the snow picture and associated information.

I then considered that it was flying up left.

But now I realize that if it were flying up left, then there would be a light trail instead of a dark trail. It seems that the light trail is down right, and gives the blur at the wingtips.

[Guest]

Phd Man,

Do you have any evidence to support your ideas? The idea that this is a fraud has been proposed before, but so far I've seen no evidence to back it up. Just as with any other theory, that one requires something, I think.

Extraordinary claims require extraordinary proofs...
What you are really saying is that you "feel" it is more likely that the image is not a camera artifact or digitally altered (which happen to be the easiest solutions). Rather, it is, in your opinion, a more exotic explanation that has given us this image. In this case, the burden of proof is on the more exotic explanation. Occam's Razor.

If you ask for proof that it is is fake, then I can just as easily ask for proof that it isn't fake. Your burden would be greater than mine, unfortunately.

Thanks.

[Bear]

Maybe the bug was hit by lightening.
Could explain why his butt was on fire!
More than likely it's - we'll never know for sure.
There is a lot in nature we can't explain and more of that is being discovered everyday.
Let's keep the discussions light, keep your eyes, your ears, and mind open, more mysteries will send this one to the back burner soon enough. 8)

[nebulamendoza@yahoo.com]

The picture that we saw on the Streak of Australia is some sort of photogenic reaction of a nebula produced by the Galactic Center

[Ruidh]

Nice, PhD Man.
I had been considering that the bug was flying down right until the snow picture and associated information.

I then considered that it was flying up left.

But now I realize that if it were flying up left, then there would be a light trail instead of a dark trail. It seems that the light trail is down right, and gives the blur at the wingtips.

We don't know which way the bug was travelling until we know whether the flash was at the beginning or the end of the exposure. Reportedly, the camera in question can flash at either extreme. If the flash was at the end of the exposure, the bug was travelling down and to the right.

I don't understand your claim about a light trail.

[Ernst Lippe]

There is only explanation for the very bright spot on the left
side that I have not seen yet (surprising when you consider the
number of contributions in this discussion).

Many animals (especially nocturnal ones) have highly reflective
eyes. When I went night spotting with a flash light in Australia
it was very easy to find wolf spiders (Lycos) and lots of other
animals.

The Cannon Power G3 has a special setting for the synchronization
of the flash, it can either go off at the start or at the end of
the exposure period. I sent an e-mail to the photographer to ask
his setting, but so far he has not responded (not surprising,
since I am probably not the only one that is asking dumb
questions).

When the flash occurred at the start, the bug is flying right to
left, and the bright spot at the left would be caused by a
reflection of its head. Otherwise, it would be caused by a
reflection of its abdomen. Because many insects have a
reflecting abdomen that would also be a satisfactory explanation
for the bright spot.

[pshisbey]

I noticed that when I first started reading this thread and studying the images I desperately wanted to believe it was a meteor. I hated the bug and contrail theories cause they were so boring. I liked the exploding bulb theory cause it was clever. But the contrail had difficulties (thank God), almost no one supported the meteor (Damn!) and the evidence for an exploding bulb didn't look right. So with no one taking serious blows to it and many supporting it with intelligent arguments and analysis I gradually warmed up to the bug. Not that the thing is over......

Phil

[Mike the bike]

loved all the beautiful responses, please don't demean your own impressions, every answer has merit, we only got where we are by thinking through things...except the character who mentioned particle beams from a plane..come on...you can do better than that..unless it's a joke and then it's real funny...

[victorengel]

Phd Man,

Do you have any evidence to support your ideas? The idea that this is a fraud has been proposed before, but so far I've seen no evidence to back it up. Just as with any other theory, that one requires something, I think.

Extraordinary claims require extraordinary proofs...
What you are really saying is that you "feel" it is more likely that the image is not a camera artifact or digitally altered (which happen to be the easiest solutions). Rather, it is, in your opinion, a more exotic explanation that has given us this image. In this case, the burden of proof is on the more exotic explanation. Occam's Razor.

If you ask for proof that it is is fake, then I can just as easily ask for proof that it isn't fake. Your burden would be greater than mine, unfortunately.

Thanks.

No. I did not claim anything. PhD Man is the one claiming something. I'm just saying, let's see the support for the claim. We have a number of running theories here. Each one has evidence to back it up, except for the fraud theory. I will readily admit that there could be a hoax perpetrated here, but I see no evidence for it. If there is any, let's see it. An example might be evidence of doubly encoded jpeg artifacts or discontinuities in lightness around certain features or any of a myriad other things. Frequently (usually?) a hoaxter will be careless about something he didn't consider. So if there's evidence of this anywhere, let's see it. That's all I'm saying.

[fpj]

Phd Man,

Do you have any evidence to support your ideas? The idea that this is a fraud has been proposed before, but so far I've seen no evidence to back it up. Just as with any other theory, that one requires something, I think.

Extraordinary claims require extraordinary proofs...
What you are really saying is that you "feel" it is more likely that the image is not a camera artifact or digitally altered (which happen to be the easiest solutions). Rather, it is, in your opinion, a more exotic explanation that has given us this image. In this case, the burden of proof is on the more exotic explanation. Occam's Razor.

If you ask for proof that it is is fake, then I can just as easily ask for proof that it isn't fake. Your burden would be greater than mine, unfortunately.

Thanks.

I believe that it is clear for everyone in this forum that the picture can be the result of digital manipulation. But unless there is an irrefutable proof that it is in fact a fraud, all other possible explanations must be taken into account.

A few experts in digital image manipulation have stated that it is very unlikely that this particular image was digitally manipulated.

If you can prove they are wrong please give us that proof.
If you can prove that any of the other theories are wrong, please give us that proof.

Some of us here are trying to follow some sort of scientific approach to analyse this. Any scientific approach cannot be done by things like “it is obviously a fraud” or “the odds of…”.

We have to consider all the possibilities and try to prove or disprove the ones that we can. This process is not a waste of time and is more then just a game, because we can all learn something.

[Guest]

The bright spot does not have any blurring due to motion of the hypothetical bug during duration of the flash. But the bug had to move almost across the whole picture for 0.05s. So my estimate of "possible blurring/total path" ratio is <~ 0.001 . So in the bug theory this imposes an upper limit on the duration of the flash as ~0.05s/1000 . Could somebody find out duration of the flash (technical info)?

I think it depends upon the intensity of the flash, which information we don't have, as far as I know.

I discovered just general info on the web. A typical flash duration is in order of 0.001s. This would result in significant blurring of the "bright spot" along the trail (~1/50 of the total trail lengh) due to the bug motion during the flash. So the question on the flash duration is an important one for the bug theory.

[Guest]

<snip>
I don't have image processing tools, but I think an interesting thing to try would be to rotate the streak horizontally, then do an FFT of the area near the dark streak, across the image, to look for any periodicity that would have been caused by the wings.
<snip>

Ok, here's an FFT created as follows:-

A 1024-pixel strip of horizontal pixels parallel to and just above the streak was extracted, windowed with the hamming function and fed into a 1024-point FFT. A sine wave was also mixed into the data at -16.0dB in order to act as a marker and to verify correct operation.

The plot of the resultant amplitudes is shown below (click the image for the full-size version).



The upshow is, there is no obvious periodicity, other than the introduced sine wave (indicated by the vertical red line).

[hazeii3]

Ok, here's an FFT created as follows:-

A 1024-pixel strip of horizontal pixels parallel to and just above the streak was extracted, windowed with the hamming function and fed into a 1024-point FFT. A sine wave was also mixed into the data at -16.0dB in order to act as a marker and to verify correct operation.

As an addendum to the above, I've repeated this test, and compared the outputs of the FFT's for 200 separate 1024-pixel strips parallel to the track; there's no obvious correlation between the FFT's outputs, which you'd expect if there was a periodic signal in there.

It does appear to me that if there is a 'wing signal' in there, it's too far below the noise to be detectable.

Further information available here.

[Guest]

If you zoom the pictures by about 400% and look at the base of the pole with the light flash, there is clearly a smoky appearance which does not appear in the before or after pictures. This "smoke" can only be illuminated by the flash of light whereever it comes from.

In addition, the ring of light around the bright flash looks clearly like some halo affect around the flash. Withing the halo ring, the color is more consistent with the water and not an extension of a bug.

Therefore, the bug cause should be ruled out unless someone can explain the smoky affect and why the halo appears more clear than part of a body.