Slashdot Mirror
A Strange Streak Imaged in Australia
Koyaanisqatsi writes "Today's Astronomy Picture of the Day presents a challenge worthy of a large audience: as it says, "Meteor experts don't think it's a meteor. Atmospheric scientists don't think it's lightning". An intriguing dark streak and bright flash that defies explanation showed up on some cloud monitoring pictures. The forumsetup to discuss it is currently hosed, so perhaps fellow slashdotters can shed some light over the mystery?"
Except that camera was periodically taking pictures, and the picture right before and the picture right after don't show any problems on the lens.
That solution is the solution to a different 'challenge' APOD, not to the dark streak in the current picture.
Um, film? It was a digital picture to begin with. Check out this in the header of the image:
uExif
Canon
Canon PowerShot G3
ACD Systems Digital Imaging
2004:11:25 15:20:49
0220
0100
2004:11:22 18:52:52
2004:11:22 18:52:52
IMG:PowerShot G3 JPEG
Firmware Version 1.02
That's not the solution to this image. However, it is the solution to the APOD from September 13th which can be seen at: http://antwrp.gsfc.nasa.gov/apod/ap040913.html
"Me fail English, that's unpossible." --Ralphie
Your response makes you look *really* stupid, because the "solution" you just read has NOTHING WHATSOEVER TO DO WITH THE MYSTERY PICTURE IN QUESTION. Worse, every single other response to the parent - most of which were BEFORE you - recognizes that (though I'll give you the benefit of the doubt there, since it took you some time to write your response). The "solution" is talking about a completely different and unrelated picture!
*Wow*.
Thanks for a good laugh though.
I've got a photo at home of a statue of Sophia. The camera caught a lens flare off the top of her head going towards heaven. I thought it was a wonderful lens effect I couldn't have made if I tried. I think it's either an unusual lens flare, or as another poster mentioned, an insect flying near the lens out of the focal range. Shawn
I used to wonder what was so holy about a silent night, now I have a child.
Take a look at the way the clouds are moving - I've never seen clouds billowing inwards.
before
"the" picture
after
You're the 13th person to point that out. Maybe read the previous replies before adding one.
Yes, those pictures are out of order. Check the time stamp on the pictures.
Nope, CCD blooming happens primarily vertically, sometimes horizontally. Never on an angle. That flash isn't bright enough to cause that massive of an overexposure, and blooming causes the overexposure to bleed into the rest of the column/row, so it would be brighter, not darker. Also, blooming is typically symmetrical from the event, this image has the line going out only in one direction.
Ok, I did some work on this...
First, EXIF fields in the photos... something you should look at first.
Camera: Canon PowerShot G3
Date and Time (original): 2004:11:22 18:52:52
Exposure Time: 1/19 sec.
Aperture: f/5.6
And for the photo After:
Date and Time (original): 2004:11:22 18:52:37
So the photos were taken with 1/19 sec. exposure, every 15 seconds.
I took the two images into GIMP, substracted them, brightened the result a lot (using Levels) and ran it through despecle. First, the lamps do look perfectly identical between the photos (or there'd be a spot around the lamp where it changed shape). In fact, the only bright bits that remain (apart from the sea reflections) are the flash and the streak.
The streak looks conical, at 1-1.5 degrees (I measured roughly using GIMP). It ends before the edge of the picture. It's about 1200 pixels long, in fact. The street lamps are 60 pixels long... Assuming that a street lamp would be on the order of 5-10 metres high, you get about 100-200 metres streak.
The cloud is VERY visible on the difference image; it has yellow-orange central spot and 2 pure-white spots to the sides; this seems consistent with a central fire and a smoke circle.
Now I substracted the before and after image and brightened them the same way. I *think* there is a visible dark spot at the place where the white cloud was; however, the image is so noisy that it could just be my imagination.
I think that the flash and the cloud were from the blown lamp. They dissipated rapidly, but there could be traces left... I'd have to do much better image processing to be able to tell.
I have no idea whatsoever what the dark streak could be. It doesn't look like a CCD sensor problem - overloaded CCDs leak brightness straight up, as far as I know. I also don't know of any lens flare that can darken the photo. It could be smoke, in which case something would be hitting the street lamp. But that would've caused lots of visible damage.
I refuse to use
This is whats commonly known as a ''sun dog'', or lens flare
Careful, sun dogs and lens flares are two completely different things. Lens flares are caused by internal reflections inside the lens. Sun dogs are caused by the sun's light hitting ice crystals in the atmosphere and are visible to the eye.
Also, why is the white pattern around the lightpost symmetrical about the axis of the dark streak? Wouldn't a bug flying sideways in the picture be asymmetrical in relation to its flight path?
Wings.
A camera flash lasts very approximately 1/1000s, or 1/50th of the exposure of this shot. You won't see much movement during the flash duration, although some of the blur may be explained by that at suh close range.
Note also that the dark streak does not extend to the edge of the frame. The illusion of something flying in and hitting the lamp post is just that - an illusion.
Ydco co
Maybe this will shed some light. It's a Photoshop "difference" between the before image and the mystery image, with a bit of levels adjustment to make it more obvious. http://img119.exs.cx/my.php?loc=img119&image=x2fst rangeprydebigdi.jpg
I analysed the images and found that there is reflected light on the water that places the flash very close to the street light. Try this: - Load all three images into Photoshop. - Convert each image to Multichannel mode. - Select the yellow channel only in each image (channels window) - Bring up your curves window "Cmd+M":mac, "Ctrl+M":PC - Set the highlights: input 40% output 0% and the shadows: input 62% output 100% Do this for each image on the yellow channel only (save the curve and reuse it for each image for accuracy) Once each yellow channel is adjusted. (make sure you are not viewing channels in the the channel colour - view yellow as black) Tab through (cycle) each image to see the highlight in the water appear. (zoom out from each image with your keyboard - PS will place them all at the same position on your screen for a still animation). The light reflection you see will be a similar effect produced by lights on the waters edge from a NYC skyline at night - Tall and defused. It's not a bug. There is no smoke. That is a flash near the street light. No camera flash was used, and there is no sun beams present in the scene. I have no other explanation at this time. All I can say is the dark line is not smoke.
the streak "trajectory" angle
The streak angle, BTW, is exactly arctan(2/3) -- the streak goes two pixels up for every three across. (It goes 652 pixels up and 978 across, which is less than 1% different from 2/3, smaller than the error of me pointing at things in the Gimp.) To me this makes it very likely to be an artifact.
Peter
No, you're looking at the timestamps from a post processing session, when he took the screen captures from his time lapse series in whatever video editing software he was using. Since the "flash" frame is more interesting he grabbed that one first (hence its earlier timestamp), then grabbed the surrounding two a few minutes later which makes sense.
The second set of timestamps are obviously the ones from the camera that took the original time lapse images since they're separated from each other by 15 seconds in both cases. Although "before" and "after" are clearly mislabeled- before should be after and vice versa.
troon is absolutely right - this thing is a bug flying across the field of view, illuminated by a flash.
There is a certain class of crackpot who thinks that out of focus pictures of insects flying across a photoframe are evidence of some strange unknown creature.
Fortunately, we can visit their websites and laugh at them. Unfortunately, they can now point at the Astronomy Picture of the Day and say "See! NASA found more evidence for rods!"
Link to roswellrods.com - don't forget your tin foil hat, and your annoying-flash-website spelunking equipment.
Link to an actual sane person describing the phenomenon
More discussion
My amazing wife - Artist, Author, Philosopher - Laurie M