Strange streak discussion: 2004 Dec 7 APOD

Comments and questions about the APOD on the main view screen.
Locked
victorengel
Science Officer
Posts: 158
Joined: Tue Dec 07, 2004 11:29 pm

Post by victorengel » Fri Dec 10, 2004 7:29 pm

Luis wrote:Why is the image of the bug white?
Please see my bat picture on page 69.

John

line in photo

Post by John » Fri Dec 10, 2004 7:37 pm

Maybe a plane shadow.

CurtC

Post by CurtC » Fri Dec 10, 2004 7:39 pm

Guest1 wrote:I find it hard to believe that a bug could not only fly such a straight line for such a distance but also that it would just happen to be in-line with location of the small ignition.
What you're calling a "small ignition" is just the body of the bug itself, very briefly lit up by the flash. About the straight line - have you ever seen the images of the "Roswell Rods"? If you piece together several of those, you can see long distances of straight bug flight.

pshisbey

Glowing Fly

Post by pshisbey » Fri Dec 10, 2004 7:58 pm

CurtC wrote:
Guest1 wrote:I find it hard to believe that a bug could not only fly such a straight line for such a distance but also that it would just happen to be in-line with location of the small ignition.
What you're calling a "small ignition" is just the body of the bug itself, very briefly lit up by the flash. About the straight line - have you ever seen the images of the "Roswell Rods"? If you piece together several of those, you can see long distances of straight bug flight.


But as we all know, the bodies of fireflys can light up without aid of a camera flash.

victorengel
Science Officer
Posts: 158
Joined: Tue Dec 07, 2004 11:29 pm

Re: Glowing Fly

Post by victorengel » Fri Dec 10, 2004 8:00 pm

pshisbey wrote: But as we all know, the bodies of fireflys can light up without aid of a camera flash.
I firefly would have a streaked image because the pulse of their light is very slow. It would streak with the motion of the insect.

Jon A

Tend to agree with Smith of Canada

Post by Jon A » Fri Dec 10, 2004 8:09 pm

I'd think it was the light burning out. Though the shooter mentions no shadow trail before/after this shot....so we're all still best-guessing, eh?

guessing

Post by guessing » Fri Dec 10, 2004 8:33 pm

This has probably been said previously in this post. Who cares right?

Ok if you will look carefully at the before picture on a high res monitor you can see the ionization trail shooting up into the storm cloud. The ionization cloud is the beggining of lighting. between shots the lightning formed the "light" portion then on the shadow pic your catching the result of the strike. Remember lighting strikes up not down. The after pic is the air vaporization trail after the lightning has disipated. There was no damage to the light fixture because it is grounded. The light bulb isn't working because the filliment was blown by the lightning voltage.

victorengel
Science Officer
Posts: 158
Joined: Tue Dec 07, 2004 11:29 pm

Post by victorengel » Fri Dec 10, 2004 8:38 pm

guessing wrote:This has probably been said previously in this post. Who cares right?

Ok if you will look carefully at the before picture on a high res monitor you can see the ionization trail shooting up into the storm cloud. The ionization cloud is the beggining of lighting. between shots the lightning formed the "light" portion then on the shadow pic your catching the result of the strike. Remember lighting strikes up not down. The after pic is the air vaporization trail after the lightning has disipated. There was no damage to the light fixture because it is grounded. The light bulb isn't working because the filliment was blown by the lightning voltage.
Can you please poing out this ionization trail more specifically?

Lightning grows from both ends, by the way. When you refer to the before and after pictures, are you making sure to revers them? They are mislabeled. Does this change your conclusions?

Diffit

Re: Animation with diff (animation #2)

Post by Diffit » Fri Dec 10, 2004 8:48 pm

zoltan wrote:Diffit: Could you do the same processing on the third image (actual after) that you did on the first two in the animation? I'm curious to see if the trail is there as well.....maybe this would add support to the camera artifact theory....the flash of the lamp simply brought it out more.

-z
Here is another 3-frame animation with the 2nd and 3rd shots plus diff shot:

frame 1: shot #2 with streak
frame 2: shot #3 (the one mislabeled "before")
frame 3: enhanced diff of frame1 and frame2 above

Image

This second animation is very similar to the previous animation of shots #1 and #2. If the streak were a result of a flying particle from the lamp, it must have had an enourmous velocity to leave behind a perfectly straight line for such a great travel distance. The bug theory seems more likely in terms of the dimensions involved.

cdsmith
Asternaut
Posts: 2
Joined: Fri Dec 10, 2004 4:18 pm

Re: Additional images after Image processing

Post by cdsmith » Fri Dec 10, 2004 8:52 pm

hazeii3 wrote: In theory, if the trail is caused by an insect, the wings should have slightly darkened the area to either side of the obvious 'body' trail. To test this, I summed the rows of pixels in a direction parallel to the trail, then normalised the resultant value and replicated it along the pixel row. The expected effect would be to reveal the slight dimming caused by the wings - however, this is not readily apparent.

Image
It's probably not necessary now that other people have done fancier analysis like FFTs, but I would think a better indication would be to sum/normalize/replicate the pixel rows parallel to the body trail. This might average out the noise but leave dimmer bands parallel to the trail with each wing beat.

Guest

Post by Guest » Fri Dec 10, 2004 8:53 pm

I agree with the two previuos posters that this is the shadow of a contrail. I live on one of the major air traffic routes on the east coast of the US and see these same kinds of streaks on a regular basis.

Bill

pshisbey

The jury is out.

Post by pshisbey » Fri Dec 10, 2004 9:08 pm

Is it possible to take a vote on this? I'd like to see a general consensus of where we stand.

Guest

Post by Guest » Fri Dec 10, 2004 9:24 pm

Which part of IT'S NOT A CONTRAIL SHADOW do people not understand?

victorengel
Science Officer
Posts: 158
Joined: Tue Dec 07, 2004 11:29 pm

Re: Additional images after Image processing

Post by victorengel » Fri Dec 10, 2004 9:31 pm

cdsmith wrote:
hazeii3 wrote: In theory, if the trail is caused by an insect, the wings should have slightly darkened the area to either side of the obvious 'body' trail. To test this, I summed the rows of pixels in a direction parallel to the trail, then normalised the resultant value and replicated it along the pixel row. The expected effect would be to reveal the slight dimming caused by the wings - however, this is not readily apparent.

Image
It's probably not necessary now that other people have done fancier analysis like FFTs, but I would think a better indication would be to sum/normalize/replicate the pixel rows parallel to the body trail. This might average out the noise but leave dimmer bands parallel to the trail with each wing beat.
You inspired me. I took the original image (not one of the difference pictures), rotatee it counterclockwise 33.6 degrees and selected only that portion of the flight path that is passing over evenly colored background. This consistet of two rather significant pieces. I butted these pieces against each other and changed from 8 bit to 16 bit mode in order for subsequent steps to get the precision desired. Next I resized to 1 pixel wide (this essentially sums up the rows). Then I resized to 256 pixels so that my filter that builds a line profile would work. Before I applied the filter, though, I pasted an upside down copy of this image onto itself with normal blend mode to cancel out the gradual change in brightness seen in the background.

My filter creates a line where the distance of the line from the left edge of the picture is the level of the horizontal raster. That's why it's 256 pixels wide. Here is the result, pasted next to "the bug".

Image

One thing that really intrigues me about this is the little perturbation near the bottom of the bottom "wing". It seems to line up very well with the bright spot on the bug image. There's a corresponding spot on the other side too. I see my selection should have been bigger, though, going beyond the width of the bug. I'll repeat this with a wider selection.

victorengel
Science Officer
Posts: 158
Joined: Tue Dec 07, 2004 11:29 pm

Re: Additional images after Image processing

Post by victorengel » Fri Dec 10, 2004 9:43 pm

victorengel wrote:I'll repeat this with a wider selection.
Here it is.

Image

Now I don't believe there's any useful data over the background noise. You may observe that my wider curve is bumpier. That's because I had to use a smaller data set to accomodate the extra width without getting into parts of the image with more detail.

victorengel
Science Officer
Posts: 158
Joined: Tue Dec 07, 2004 11:29 pm

Survey

Post by victorengel » Fri Dec 10, 2004 9:55 pm

I've created a survey. The software allows only 10 entries. I put them in alphabetical order. Let me know if your choice is missing.

http://www.opinionpower.com/Surveys/377021291.html

P.S. Post has been changed to link to the new survey. The previous one had problems.
Last edited by victorengel on Sat Dec 11, 2004 8:31 pm, edited 1 time in total.

nobodyimportant

Re: Animation with diff (animation #2)

Post by nobodyimportant » Fri Dec 10, 2004 9:56 pm

[quote="Diffit
Here is another 3-frame animation with the 2nd and 3rd shots plus diff shot:

frame 1: shot #2 with streak
frame 2: shot #3 (the one mislabeled "before")
frame 3: enhanced diff of frame1 and frame2 above

Image

This second animation is very similar to the previous animation of shots #1 and #2. If the streak were a result of a flying particle from the lamp, it must have had an enourmous velocity to leave behind a perfectly straight line for such a great travel distance. The bug theory seems more likely in terms of the dimensions involved.[/quote]

Is it just me or does it appear that there is a slight arc to the streak?

victorengel
Science Officer
Posts: 158
Joined: Tue Dec 07, 2004 11:29 pm

Re: Animation with diff (animation #2)

Post by victorengel » Fri Dec 10, 2004 10:05 pm

nobodyimportant wrote: Is it just me or does it appear that there is a slight arc to the streak?
Maybe I see something. I used the measuring tool and see something pulsing every 121 pixels. Can FFT guy identify 121 pixels on his chart?

Diffit

Re: Animation with diff (animation #2)

Post by Diffit » Fri Dec 10, 2004 10:18 pm

nobodyimportant wrote:Is it just me or does it appear that there is a slight arc to the streak?
Some have suggested that, but if you load shot #2 into drawing software and draw a straight line along the path, the streak is very straight. Look at the blinking (certainly straight) line on this rotated crop of the diff image:

Image

chrisanthony

Re: This is probably silly...

Post by chrisanthony » Fri Dec 10, 2004 10:25 pm

i think your hypothesis has something going for it. it's quite possible!

nebulamendoza@yahoo.com

Post by nebulamendoza@yahoo.com » Fri Dec 10, 2004 11:03 pm

I am almost sure of it. If you compare this picture to a picture of the Milky Way from Earth you can see that this is some sort of arrangement that has to do with the Milky Way. Maybe the outcome of a Gamma Ray Burst impacting our Galaxy. I thought about this because I saw it and I really think that many astronomers think the same thing about it as me. I wish to know more because I like Astronomy. If you find out what it really is please send me a e-mail with the answer so that I can get the real story.

Ernst Lippe

Post by Ernst Lippe » Fri Dec 10, 2004 11:22 pm

Anonymous wrote:I agree with the two previuos posters that this is the
shadow of a contrail. I live on one of the major air traffic routes on the
east coast of the US and see these same kinds of streaks on a regular basis.

Bill
This is one of the most common one line explanations that have
given many times in this discussion. It has several drawbacks:
First of all, it does not explain the white blob on the picture.

It also does not explain why the streak is not visible on the other
photo's. The next picture that was taken only 15 seconds later does not show
any sign of it. This is only reasonable when there is a lot of turbulence in
the air, but in that case it is very unlikely that the contrail would be so
uniform, essentially it appears to be undisrupted across its entire length.

If it is the shadow of a contrail it must be projected on something that
reflects. It is indeed possible to see reflection of a shadow on clouds or
mist, but there are only two cases where you will observe that shadow as a
straight line:
  • The first case is when the observer is in the same plane as the
    original line-shaped object and the light source. In that case the shape of
    the shadow is not dependent on the shape of the reflecting surface. But this
    is simply not possible in this picture because then the sun should be located
    in the line of the shadow when the sun is in front of the observer as it is in
    the picture.
  • The second case when we could get a line shaped shadow is when the reflecting
    surface is flat and uniformly reflecting. Now that is highly unlikely, because
    the clouds are not all uniform along the streak.
(If you don't understand this part of my explanation, some experimenting
with shadows on an uneven surface will probably show you what I tried
to write here).

If it would be only the shadow of a contrail, why don't we see
any sign of the contrail itself. After all it should be pretty
impressive and located somewhere between its shadow and the sun.

A similar theory that is equally unlikely is that the dark streak is the
contrail itself. That is also extremely unlikely. The streak is visible all
the way to the horizon (and if you look at the difference between the pictures
you can see that it even appears to continue below the horizon). But a
contrail that appears close to the horizon must be located at a very great
distance. So this particular contrail must have been very long and extremely
uniform. Normally you would expect that the older parts of such a contrail
would already have started to disintegrate (after all 15 seconds later it has
completely disappeared). But there are no signs of disintegration visible on
either end of the streak. Also the formation of a contrail depends strongly on
the humidity of the air, it seems likely that there is a fair amount of
turbulence in the air, the trees/bushes on the ground are definitely moving
and if you look at the clouds in the pictures there clearly is a lot of
turbulence in the higher levels as well. But with a high amount of
turbulence you would not expect such a highly uniform contrail, because
there would be large variations in humidity and because parts of the
contrail would be rapidly disturbed.

Cloudbait

Post by Cloudbait » Fri Dec 10, 2004 11:44 pm

Ernst Lippe wrote:This is one of the most common one line explanations that have
given many times in this discussion. It has several drawbacks:

It also does not explain why the streak is not visible on the other
photo's. The next picture that was taken only 15 seconds later does not show
any sign of it. This is only reasonable when there is a lot of turbulence in
the air, but in that case it is very unlikely that the contrail would be so
uniform, essentially it appears to be undisrupted across its entire length.
I agree in general with your comments regarding contrail shadows, but not this one. Long contrails will not dissipate in 15 seconds, but their shadows can do so very easily. All it takes is for the Sun to go behind a cloud (which can happen even if it has already set), or for the contrail to move enough that the viewer is swept out of its shadow plane. And even crooked, raspy contrails generally produce very clean, straight shadows, since the shadow is usually a forshortened projection of the contrail.

While I do not believe that this can be a contrail, that is because of the geometry of the elements, not the transience of the effect.

Victor Stover

Strange flash

Post by Victor Stover » Sat Dec 11, 2004 12:50 am

I think the light was off and when started to light the bulb exploded or there was a short circuit. I have worked in a warehouse/factory that had High Pressure Sodium (HPS) or Metal Halide (MH) blow off or there was a short in the fixture. Some of you have seen a street light that was off, start to flicker and come on only to have it go back off a short time later. In a short time the same process starts all over again. Most of the time it is a bad element in the bulb that is the cause. On HPS or MH lights there is more than just the bulb to blow or short out. Most have transformer or ballast that could short and are wired at 220 or higher voltage. Look at all the other lights that are on in the photos was this at dusk when light start to come on? If the light was full cut off it could be hard to see if the light was off or on. Plus it looks as if photo was taken at a point higher than the fixture. A short could last seconds if the current has not tripped off the power source. There could have been two events. Bulbs blows than a short and short time later. That maybe why the smoke like area is so far off the fixture. A small electrical short can make a large flash and show no damage to lighting fixture as in the after photo. UFO :wink:

Guest

Re: Additional images after Image processing

Post by Guest » Sat Dec 11, 2004 12:53 am

cdsmith wrote:
It's probably not necessary now that other people have done fancier analysis like FFTs, but I would think a better indication would be to sum/normalize/replicate the pixel rows parallel to the body trail. This might average out the noise but leave dimmer bands parallel to the trail with each wing beat.
I'm not clear what you're suggesting....because you mention wing beats,are you suggesting summing the pixels perpendicular to the trail? It's certainly a nice idea - I tried it, and took an FFT of the result (btw, the other FFT stuff was me too) However, there was still no sign of a periodic signal. Possibly a better approach would be to take the FFT's of the pixel strips first and then sum the FFT's; that way any phase difference between the strips wouldn't affect the result. Essentially, it would be equivalent to lengthwise summing of the data in the following image (which is a 3D representation of the FFT's of a 200-pixel strip centred on the trail).

Image

(as usual, click the image for a larger version, and see here for further info)

The actual expected signal would be similar to the various 'rod' photographs others have posted (e.g. the top pic on this page. Just out of curiousity, I'll try an FFT on those pics tomorrow and see what appears; I'd certainly expect a strong signal from them.

Locked