Preliminary Analysis of the Albany TV Object
FIGURE 1 A SERIES OF FIELDS PASTED TOGETHER
by
Bruce Maccabee
Nov. 26, 2002
THIS PRELMINARY REPORT PRESENTS IMAGERY WHICH IS THE KEY NOVELTY OF THE
ALBANY TV VIDEO. I HOPE THAT OTHER VIDEO ANALYSTS WILL CARRY OUT THEIR OWN
INVESTIGATION SO THAT WE CAN ARRIVE AT A CONSENSUS AS TO WHETHER THE PARTICULAR
IMAGE IN QUESTION (see below) IS "REAL" OR SOME SORT OF ARTIFACT.
THE VIDEO HISTORY
On October 21, 2002 I received a phone call from an employee of Fox 23 TV in Albany,
NY who told me that a cameraman, while filming background scenery for the local weather
report, had filmed something else as well: an "object" that had traveled so rapidly
through the field of view of the camera that he had not noticed it until he was back
at the TV station and running video slowly while editing it for the final production.
The object only appeared on 10 fields of the video so the total time of appearance was
about 1/6 of a second. (Each field is 1/60 sec and the fields are generated
consecutively to make 60 fields per second. These are combined in successive pairs to
make 30 complete "frames" per second.) As he examined this video field by field he
noticed that the object appeared to go behind a wisp of cloud at about 4,000 ft.
Immediately he thought that the object must have been big and traveling at a huge rate
of speed. He showed the video to several people at the station and they all had the
definite impression that the object went behind the cloud. They agreed that the
object must have been quite long to have made an image with the length shown in the video.
They called the local airport to find out if the object had been detected on radar.
The answer was no, so they decided they had better report to "higher authorities"
because of the possibility that some extremely fast object(s) could be flying
around undetected. They contacted the local sheriff and he, in turn, contacted the
FBI. The authorities, including some military official(s) questioned the cameraman
and the FBI took a copy (or the original?) of the videotape. The FBI also administered
a polygraph test to the cameraman.
The employee asked if I would be willing to analyze the video. I said yes if
they would send me a good copy. I finally received a VHS copy on Nov. 6. In the
meantime, there were press stories on this event:
Strange Sighting over Albany Airport, New York
Flying Object over Albany International
October 21, 2002 (date of story)
A strange sight in the skies over the Capital Region - no one
knows yet what is was - but it was enough to have the FBI
scrambling to take a look.
One of our videographers was shooting a plane taking off from
the albany international airport - at the same time the strange
object streaked through the sky - it was moving so fast, he
didn't even see it, until he played his tape in slow motion.
We're told the object did *not show up on radar. Authorities
will continue to analyze the video to figure out what the flying
object is. We'll keep you posted
Source: The Times Union - Albany, NY
http://www.timesunion.com/AspStories/story.asp?storyID=63973&category=REGION&newsdate=
10/22/2002
By CATHY WOODRUFF,
Staff writer
First published: Tuesday, October 22, 2002
Was it a terrorist missile, an off-course alien space craft or
just a zippy mosquito with a hearty autumn constitution? Local
authorities hope the FBI will know for sure.
The image caught on videotape by a Fox 23 cameraman shooting
background for weather stories on Sunday [Oct 20] resembles -- when it's
slowed, enlarged and paused -- a fine rod with a small set of
wings near each end.
The cameraman didn't notice it at the time, said Fox assignment
editor Jeb Rowledge, but the tiny dark spot streaking across the
frame, apparently above the clouds, caught his eye while editing
the video of a plane taking off at Albany International Airport.
The station showed the video to airport and Albany County
Sheriff's officials Sunday night, and on Monday, local
authorities asked the FBI to take a look, too.
"There's something on that videotape that is interesting, to say
the least, but we don't have any idea what it is," said airport
spokesman Doug Myers.
"Right now, we don't know what it is or if it was there," said
Undersheriff John Mahan. "It looks like a rocket," but it did
not show up on airport radar, and even the cameraman didn't see
it until he looked at the tape.
Albany FBI spokeswoman Lisa Massaroni said the tape will likely
be sent to FBI headquarters in Washington, D.C., for analysis.
"Nothing has been determined about what, if anything, is on that
video," she said.
The image resembles phenomena called "rods," which some buffs
suggest could be paranormal or extraterrestrial.
Skeptics, however, say they are simply the result of an insect
flying past a video camera at a certain speed and distance. They
contend the figure is formed because bug wings beat at a
different frequency from the frame capture rate of a video
camera, creating a strange image resembling a rod, the blurred
body of the insect moving forward, and small bulges along each
side.
The mention of "rods" in the above article is a result of the shape of the image,
namely, an elongated, faintly dark line against the blue sky. The image
does not have a lot of contrast with the sky, but nevertheless is easily seen, even when the
video is run at normal speed, when one knows when and where to look. Besides the
dark line there are "appendages" which appear twice in each frame. The appendages
are elongated images, slightly brighter than the sky, which occur in pairs: one of each
pair is above the dark line and the other is below. Oddly enough, these elongated
images are almost perfectly vertically oriented even though the dark line itself slants
upward and to the left at an angle of about 20 degrees or so.
Shown below are fields 4 - 7 showing the object which first appeared at the lower
right (in field 1, not shown). By the time of these fields the object had reached about
the center of the field of view. You can see the cloud at the upper left and the airplane
at the upper right. The airplane is leaving the field of view as the camera rotates to the
right to follow the airplane (which had just passed nearly overhead a second or so earlier).
The appendages that are above and below the dark line are barely visible in this
presentation, a result of digitizing a VHS copy of the original. The video original has
slightly more contrast.
FIGURE 2 FIELD 4
FIGURE 3 FIELD 5
FIGURE 4 FIELD 6
FIGURE 5 A PORTION OF FIELD 7; THE OBJECT APPROACHES A CLOUD
The fact that the line is dark against the sky indicates that the central body of the
object is at least partially if not completely, opaque, i.e., it has blocked some of the
skylight. On the other hand, the central body didn't reflect enough light to appear
brighter than the sky. However, the "wings" or side appendages, a pair of which appear
twice in each field, are slightly brighter than the sky.
THE HEIGHT OF THE OBJECT
If it weren't for the fact that the object appeared to go behind a cloud this video would
be immediately relegated to the class of videos which show nearby insects that happened
to zip by close enough to the camera to be out of focus and appear to be traveling at a
high rate of speed. However, one frame of the video imagery seems to rule out this "buggy"
explanation. To establish a reference point for what the cloud image looks like when
the object is notn passing the cloud position, look at the ninth field, which is
shown in Figure 6 below:
FIGURE 6 FIELD 9; FULL VIEW WITH ENHANCED INSET (OBJECT IN FRONT OF CLOUD)
Here we see that the image of the object clearly OVERLAYS one of the the wispy cloud
images (see enhancement). Keeping in mind what this looks like, now look at the previous
field, field 8, where IT APPEARS AS IF the view to the object was BLOCKED by a
wispy cloud and also Figure 7 before the obbject reached the cloud.
FIGURE 7 FIELD 8; FULL VIEW
FIGURE 8 FIELD 8; ORIGINAL AND ENHANCEMENT (OBJECT BEHIND(?) CLOUD)
The enhancement shows that there is apparently no darkening of the image of the wispy
cloud where the image of the object WOULD cause a darkening if the object passed
between the cloud and the camera, as happened in Field 9 (Figure 6). Just to be certain
that one "knows where to look" to see the faintly dark line made by the moving object,
note that the tiny hole in the cloud was darkened by the passage of the object. Connect
the center of that hole with the nearest end of the dark line image and one gets
a straight line that crosses the wisp of cloud. This is where the darkening WOULD BE
if the object passed between the cloud and the camera (see notations on the
enhanced image).
The fact that the image of the object overlays a wispy cloud image in the
9th field but apparently goes behind a different cloud image in the 8th field is not
an inconsistency. It would simply mean that the cloud in the 9th field was somewhat
higher than the cloud in the 8th field and that the object was essentially at "cloud height."
Herein lies the mystery of this video. If the object had clearly darkened the
image of the cloud one could argue that it was below the cloud, in which case it could be
at any distance between the camera and the cloud and, in particular, could be close to the
camera and hence small like an insect. But there is clearly no evidence of darkening of
the cloud where the object "should" have crossed it.
Considering the implications of this "video fact" (strange, high speed objects zipping
around) one must ask whether or not the failure of the object to make a slightly dark
line across the wisp of cloud could be the result of anything OTHER THAN the altitude
of the object (i.e., its altitude being higher than the cloud). For example, one could
ask the following question: could the failure to show the object crossing (darkening) the
image of the cloud be a momentary electronic artifact or "glitch" (a momentary failure
of the camera)? This seems impossible to imagine because the nature of the assumed glitch
would be strange indeed. This question assumes that the object DID pass beneath the cloud and
DID darken the wispy cloud, but that the camera and electronics somehow failed to record the
momentary darkening or "covered it up." Yet, the darkening of the (different) cloud in frame 9
shows that thepassage of the object "across" a cloud, with the resultant darkening of the cloud
image, was not too fast for the camera to record. Furthermore, the cloud image in frame 8
where the object SHOULD have darkened the cloud looks exactly the same as it does in all the
other frames (compare the cloud around the small hole in frames 4 - 9; there is no change.)
It is inconceivable to me that the camera could fail momentarily and effectively remove the
slight darkening of the cloud image along the line of motion of the object and yet make no
other change to the image. I, therefore, reject this possibility.
What other possible explanation might there be?
(Added Nov 28: NOISE!!! See below!)
....................................
This is a preliminary report which I am publishing with the hope that others will comment
and analyze the video. The implications, if this imagery is "true" (the object was distant),
are of considerable importance (high speed, large objects traveling through the sky at
"invisible speeds"). Clearly, then, it is important to arrive at a concensus as to whether
or not the object did pass above the cloud. If that is the consensus, then I will publish
a more complete report (already largely finished) that includes calculations of the estimated
size and speed of the object.
...................................................
***************************************************
FURTHER ANALYSIS NOTES (added Nov 28):
In an effort to elucidate the elusive evidence that the object passed in front of (i.e., below)
the cloud rather than behind the cloud I present four more processed images. In all cases
I have removed the color in order to concentrate on relative brightness levels. In the first,
the contrast was increased during digitization of the VHS video such that the white part of
the cloud saturated the response of the video display electronics. The response for
typical pictures corresponds to 256 levels ( zero to 255). The brightness of an image
is proportional to the brightness of the light source as long as the electronic equivalent
of the brightness (in the camera it is a voltage level that has been quantized or digitized
into 256 levels) does not exceed 255. If the brightness exceeds 255 then it is "assigned" the
value 255 regardless of how much brighter it might be. Hence, by increasing the brightness
of the portion of the image of interest here, i.e., the portion that shows the wisp of cloud
where the object would have crossed, I have "automatically" saturated the portions of cloud
image that were already bright.
In Figure 9, below, I have pointed out a "gap" in brightness of the (now-much-brighter than
in the previous figures) wispy cloud image. The presence of this gap seems consistent with
the idea that the object did pass in front of the cloud, but the appearance or "look" of this
gap is not thoroughly convincing because it does not seem to be as straight as the track of
the object and because it does contrast does not appear to be as great as the contrast
between the line image and the background sky.
(NOTE: The letters A and the associated lines indicate the locations of the appendages, all
four of which can be seen in good reproductions of this field. Unfortunately the additions
of indicator lines and letter notations has introduced more pixel noise around the lines and
letters that did not exist in the original. Do not use this image as a source for further
picture processing.)
FIGURE 9 Field 8 The letter A designates the locations of the appendages
As pointed out above, the contrast between the brightness of the "gap" and the brightness
of the nearby cloud does not appear to be as great as the contrast between the line image
and the background sky. However, just because it does not appear to have as much
contrast doesn't mean that it doesn't have as much contrast...the eye is not a good
photometer (device to measure relative brightness). The actual contrast can be checked
by measuring the relative brightness levels at points within the image.
Figure 10 shows the same scene but without contrast or brightness enhancement.
FIGURE 10 Field 8
The numbers presented in the figure are the measured relative (digitized) brightness levels
at several points in the image, said points being at the ends of the indicated lines. The
number 98 is the relative brightness at a location within the line image made by the object
as seen against the sky and the number 108 is the relative brightness of the nearby sky image.
The ratio is 98/108 = 0.91 which means that using these two points for comparison the line
image is about 9% less bright than the nearby sky. However, one finds that measuring at
different points (actually difference pixels) in the image one can find other relative
brightness levels for the line image such as, for example, 96, 97 and even as low as 88 (at
the hole in the clouds), and other relative brightness levels for the sky image, e.g., numbers
in the range 106 to 110 within the scene area shown in the figure. These variations are a
result of electronic noise in the camera and electronics. The image of the sky should be
perfectly uniform in brightness over the area of this image. However, as one can easily see,
the image of the sky appears granular or "puckered," indicating that there are small
brightness variations that were caused by electronic noise.
Within the gap in the wispy cloud the relative brightness is indicated as
122, whereas the relative brightness of the adjacent cloud is indicated as 135.
The ratio is 122/135 = 0.90. Again, these numbers change and hence the ratio changes
somewhat from point to point. However, the near agreement between 0.91 (ratio of the line
brightness to brightness of the blue sky background) with this ratio, 0.9 (gap to white
cloud background) does suggest that the object did pass below the cloud.
BUT.....
To confirm this one should check the following and preceding images (fields) to see if
the "gap" portion of the wispy cloud was brighter before and after the object passed
by. Figure 11 shows field 9 with the brightness and contrast increased to show a gap,
if any.
FIGURE 11 Field 9
Oddly enough, there appears to be a gap in field 9, "long" after the object has passed by.
Clearly that would indicate that the gap was not caused by the object!!
(The cloud itself did not change appreciably in the 1/60 second between field 8 and
field 9.)
BUT.....
Now we must check field 7. This is shown in Figure 12, which also has increased contrast
and brightness. Oddly enough, there does seem to be a narrow brightness gap, perhaps
just at the "left side" of (or below) the location where the object would make its darkened
line. There is also a blob of extra brightness just above the gap which appears to be
directly in line with the path of the (oncoming, in frame 7) object.
FIGURE 12 Field 7
If that moderately-bright blob of cloud image in field 7 were an actual measure of the
cloud brightness before the object arrived, and if the gap in brightness of the cloud
in field 8 were an actual measure of the brightness of that blob as the object passed by,
then the decrease in brightness from field 7 to field 8 would confirm that the object
passed below the cloud.
BUT.....
Which is correct: field 7 or field 9? They seem to contradict one another, although field
7 does seem to show a slight gap, in which case the liklihood of field 7 confirming
field 9 (gap existed before the object arrived) is greater than the liklihood of field 9
confirming field 7 (gap did not exist until the object passed by).
Got that?
Hmmmmm?
If you feel confused, don't feel bad. You're not al;one.
I, too, am confused.
It appears that the comparison of fields 7 and 9 with field 8 raises more questions
than it answers. The comparison does NOT confirm that the intriguing brightness gap in
field 8 was caused by the passage of the object. What this comparison does confirm is
that electronic noise is overwhelming the "signal" here.
Electronic noise has "awesome powers." (No, not Austin Powers!!)
It can create a gap where there should none and it can "close" a gap where there should
be one!
From whence cometh all this awesomely powerful noise? Let's review the steps from the
orginal scene to what I have analyzed.
The original video was made by recording images created by an analogue electronic image
sensor device, a CCD silicon device that "converts" light into electric signals. This
conversion is accompanied by a very small amount of photon (shot) noise, some thermal noise
and perhaps 1/f noise added to the brightness levels in the image. The signal from each
element of the CCD (each pixel of the image) was digitized before being recorded on a medium
(tape, disc, whatever). The digitization process introduces quantization noise and there is
some small noise inherent in the recording medium (tape/disc). When this was played to create
my VHS tape the digitized signal levels on the original recording were converted to analog
levels ("anti-digitization"), thereby introducing further noise into the signal. Finally,
when I played the VHS tape the electronics added a small amount of noise and then the
digitizer ("Snappy") added some noise when digitizing the image.
The addition of noise at each step makes images such as the sky image, which should be quite
uniform in brightness, seem grainy. After all these noise additions one may marvel that the
whole system works at all. However, the added noise in each case is only a fraction of a
percent of the brightness level or some small, fixed (independent of brightness) level
called the "noise floor," so the total noise isn't great. But it is enough to make
small signal levels, such as the brightness variations under study here, indeterminate.
The total addition of noise to the signal in this VHS copy of the video corresponds to
brightness variations in the range 5 - 10%, as one can determine by measuring relative
brightness levels over small areas of the blue sky image, after converting the sky image to
grey levels. When the electronically created image noise variation is comparable to the
expected signal level (the expected signal is about an 8 to 10% decrease in cloud brightness),
one cannot arrive at a conclusion.
Bottom line: if this question is to be resolved it will be necessary to analyze the original
digitized video or at least a digital copy of the original.
.
.
.
Are you listening, Fox 23?