Summary:
Two videos of similar (perhaps the same) objects over coastal
Israel, appear to show characteristics that are commonly
associated with UFOs; erratic motion, an amazing range of speeds,
and right angle turns. The most unusual characteristic is what
appears to be erratic motion; although the object size is
unknown, conservative estimates of the object size, result in
object accelerations that are not feasible with known technology,
and would probably destroy any known aircraft. Like most videos,
the cameras are zoomed, leaving out terrestrial objects that would
be used as reference points to determine UFO motion. Fortunately,
the cameras were able to see background stars, making the
measurement of UFO motion possible.
Report:
On the evening of July 22, 1999, over the Israeli "Special"
(nuclear?) Air Force Base at Palmachim-Yavne (S. of Tel Aviv),
an object was reported to fly from the north to the coastal area
where it stopped and hovered above the base. No noise was
reported in the six hours the object hovered. Over 100
witnesses reportedly saw the object; several are seen on the
video submitted. Planes were reported circling the object at a
distance, although none appear on video. The video was about
12 minutes long; the tape has gaps, and with no onscreen time,
the actual duration covered is unknown.
Generally, the video shows the camera zooming into the object,
giving extreme magnification. The camera is capable of digital
zoom, which mostly provides magnification and artifacts but no
true detail. When normal hand-holding jiggle causes the object
to be lost, the camera is fully de-zoomed, the object relocated,
and zooming reattempted. This allows intermittent views of the
object position in the sky (compared with streetlight positions),
alternated with views of the object shape. The composite image
Fig. 1 shows the object position at the noted time in seconds
from the beginning of the video. Note the non-linear path of the
object. Also note that the object moves a considerable angular
distance in about a minute. (A witness face is obscured for
privacy.)
In close zooms, the object shows a wide variety of shapes. The
composite Fig. 2 shows some of those seen. None of these
images are "one-frame-wonders"; all shapes shown are visible in
numerous frames, and are therefore not merely due to random tape
noise or other intermittent artifacts. Equally important is to
recognize what is NOT "seen" (video artifacts as opposed to real
images). The horizontal and vertical lines are camera artifacts
of digital zoom and interlace, and should not be taken as actual
structure of the object. Relative brightnesses of light are valid
data, taking overexposure into account. The actual colors of
small objects are not properly recorded onto videotape, due to
video bandwidth limitations. A bright overexposed object appears
white, regardless of actual color. A large blob should not be
taken as a true size, this usually indicates imperfect focus of
the camera. A bright area around a white light, fading with
distance from center, is probably glare. As the glare is often
not overexposed and is over a large area, its color is indicative
of the actual object's color, even if overexposed.
Click here for negative video of UFO
Most camcorders will detect near infrared, in the range of 700-
1000nM (nanometers wavelength). For reference, the human eye sees
violet near 400nM, up to red near 700nM. Silicon photodetectors,
the usual heart of a camcorder video signal, typically have peak
response around 700nM, near infrared. A camcorder infrared response
can be tested by looking at a TV remote control through the camera
while pushing buttons on the remote; infrared response will be seen
as white flashes in the camera that can't be seen by eye. Infrared in
a video can be distinguished in three ways: It will be seen though
the camera, but not by the naked eye (if excited witnesses take care
to note!); it appears white; and being a longer wavelength, will
focus differently. "Near infrared" should be distinguished from
"thermal infrared" which is a far longer wavelength not detectable
by ordinary camcorders, often used in night-vision equipment. For a
typical achromatic lens, infrared underfocuses; it is bent less, so
the lens-image distance must be greater than for visible light.
The result can be a ring shape, distinct from properly focussed,
visible light images in the same scene.
Note: The object typically shows two or three lights, sometimes
with one or two dimmer lights. The bright/dim distinction changes;
in one segment, an equilateral triangle arrangement slowly turns
counter-clockwise and flattening, with the lower left light dimming
until the object resembles a pair of lights with a dimmer light
between. It should be noted that the characteristic flash of
airplane anti-collision strobe lights are never seen in the video.
Note: In the 5th image down in Fig. 2, the appearance of a
"headphones" shape. The "headstrap" bridging the "earphones" is
not a segment of a circle, indicating merely a mis-focused point
of light. The aspect ratio (height/width ratio) is about 1.5/1,
clearly different than a circle. During the headstrap's
visibility, occasional twinkles are seen within it; it appears
to be a real object, not an artifact.
The 6th to 8th images in Fig. 2 show a ring of light. Particularly
in the 8th frame, the "earphones" are in fairly good focus but
the ring is quite evident. This would be consistent with a strong
infrared source, producing an under-focused ring image, while
visible wavelengths are in proper focus.
The images in Fig. 3, the result of digital zoom, are useful with
the repeated caveat that the vertical pattern is from digital
zoom, not a real attribute of the object. The uppermost frame
shows the right light extremely over-exposed, showing an orange
result in the non over-exposed glare. This orange result should
be interpreted with caution; even the best cameras produce poor
color at night. Note in your local nightly weather forecast, live
camera shots of city interstate intersections may be shown for
gauging traffic or weather. The picture will appear quite red,
although any driver knows the scene is yellowish-white. This is
due to the sodium-yellow lights' wavelength registering on the
red channel of the camera. The true color could be yellower than
what is seen on the video. The important aspect of this image is
that the object has an extremely bright yellow/orange light,
with two much weaker white lights. This combination is
inconsistent with typical aircraft lighting; extremely bright
lights are typically landing lights, which are white. The last
three images in Fig. 3 show closeups of a three-light arrangement,
with the lower two lights in fairly sharp focus (although much
glare) but the upper light in very bad focus. The difference
between glare and focus is obvious; the lower circles have an
clear source in the bright centers, but the upper ring has little
or no distinguishing center. This verifies the upper ring is the
result of poor focus of that wavelength, not glare. As the other
points are in fairly good focus, the mis-focus is apparently due
to a wavelength difference between the lower lights and the upper.
Infrared light appears to be the only candidate which would cause
this result; ultraviolet light penetrates glass poorly and
registers on silicon photoreceptors poorly.
At several points in closeups in the video, persistent points of
light (as opposed to random spots of video noise) appear in the
same, or at least in predictably different, positions on the
screen. This is consistent with a background star (example in
Fig. 4, circled) appearing in the video, appearing to bounce around
due to the unsteadiness of the videographer's hand. (I could not
find a weather report verifying clear skies that night; another
source of light, such as a plane running light, gives little
change in the conclusions so based.) Such a reference star is very
useful in that it can be used to determine the true angular motion
of the object, as long as the object is far enough away that
positional movement of the camera generates no significant parallax
error. Given the numerous witnesses referring to a distant object,
and the degree of zoom making any significant lateral movement of
the camera difficult while maintaining aim, the possibility of
parallax seems doubtful for giving a significant apparent shift
between the star and the object. The results shown below, therefore,
appear to be true angular motion of the object rather than the
result of artifacts.
Shown, graphically in Fig. 5, is the X-difference (solid line with
square markers, in units of screen pixels on a 640-by-480 display)
and Y-difference (thick line) between the object and the star over
a 145-frame time period (almost five seconds). As a size reference,
the object is about 40 pixels wide center-to-center of its outermost
lights. The X-difference between the star and the camera aim is
shown as a dotted line with triangle markers. The star is often not
clearly visible, being smeared out by camera motion to obscurity
or simply off screen. Such points are unmarked on the graph. Note:
The object made five reversals in direction in these five seconds -
an astonishing performance. The maximum lateral motion is about
six times the apparent width of the object. Assuming (as a probably
conservative guess - see below) that the object is 10 feet across,
the object went 60 feet back and forth in about 1 second. The
maximum acceleration is about 18G. This is consistent historically
with witness reports in which erratic UFO flight is reported, and
quite uncharacteristic with known aircraft performance. The
reference star does not appear to be a lens flare from any of the
numerous street lights in the area (or of the object itself).
Several lens flares ARE obvious during low-zoom conditions. If a
lens flare, it would have the shape (perhaps reflected and/or
inverted) of the source light; instead, the star is small and
point-like. But, assuming the object to be motionless, and that the
apparent motion is totally camera motion, the "star" motion should
correlate to the camera motion. This is since a lens flare is merely
a reflection off some stray surface inside the lens, and moving the
"mirror" (lens/camera) will also move the reflection accordingly.
But no such correlation is found. All indications are that the star
is a distant point-like object, and a valid reference point.
The graph in Fig. 6 shows another segment of the video where a star
appeared useful as a reference point, with the same symbology as the
previous graph. Similarly, about 8 reversals are seen in about a
seven second period. The object width is about 60 pixels, so the
object moved laterally about six times its own width back and forth.
Near the end of the segment, extreme motion of the camera smeared
out the star image beyond recognizability, so this portion of the
graph is blank.
Extreme motion of the camera can be useful. The image in Fig. 7
shows a streak generated, presumably when the camera was bumped;
the previous frames are also disturbed as if bumped. The streak
shows a time exposure of the object during the 1/50 second exposure
of a single field. The brightness of the streak is representative
of the brightness times the duration the light spent at a particular
point on the image. Since the motion of the camera isn't known,
the time duration during the streak is not known. However, since
there are two lights, IF the lights are of constant brightness, one
would expect the brightnesses of the two treaks to correspond; a
bright region of one light's streak would correspond to a bright
region of the other. This effect is not seen; corresponding parts
of the two streaks are obviously of different brightness. The
differing brightnesses could be explained by several phenomenon,
all resulting in an anomalous conclusion:
1. The lights are rapidly changing brightness. Standard tungsten
aircraft lights are incapable of doing this.
2. Atmospheric turbulence (twinkling) is causing the brightness
changes. This implies a considerable amount of air between the
object and camera; the amount depends on distance and the air
stability. A hot night will have more thermal action than a cold
still night. (A large amount of heat roiling off the object, such
as looking directly into a jet exhaust, will also cause turbulence.)
Assuming no object heat, I would guess at least several kilometers
of air between the object and camera to generate this amount of
twinkling. Conservatively assuming one kilometer distance, and
100X digital zoom (which is a guess, but typical) and a pair of
lights 1/3 of the frame-width, the light spacing is 1000 meters /
100X zoom * 1/3, or about 3 meters apart. Often, three or more
lights at similar spacings are seen, so the object size is at least
twice this size.
In both graphs, the camera motion clearly lags the object motion by
about 1/3 second; this is what would be expected of a videographer
trying to "track" an erratically moving object. The 1/3 second
delay is what would be expected from studies of computer mouse and
trackball human response testing, and also is on the slow end of
TV cameramen's performance in tracking the football during
unexpected events (intercepts, fumbles, fake handoffs, etc.) during
NFL play. So the camera motion is quite typical of what would be
expected from normal human hand-eye coordination. This is also
additional confirmation that the star is not simply a lens flare;
a lens flare's position would react instantly to camera motion,
not with a delay. A summary of the object characteristics:
1. No standard aircraft anti-collision strobes are ever seen.
2. The object path is atypical for a plane. A helicopter could
easily make such a path, although hovering silently for six hours
is a gas-guzzling challenge.
3. The arrangement of lights changes continuously, often assuming
a non-horizontal arrangement. Airplanes, unless viewed from
substantially below, appear as inherently horizontal objects.
4. A "headstrap" structure is seen which appears to be real, and
does not match any lighted airplane or helicopter structure I am
familiar with.
5. At one point, the object's brightest light, by far, is approx.
orange (perhaps yellow or red). An extremely bright light of this
color is not standard on any airplane I am familiar with.
6. At another point, the object appears to have a light source
that emits infrared (approximately 900nM) with little or no
corresponding visible light.
7. The object position, compared to reference stars, is erratic.
Although the object size is unknown, even conservative size
estimates yield astonishing acceleration results. The first six
characteristics might be achieved with a determined faker with an
incredible helicopter and a complex array of visible and infrared
lights, but the last characteristic is implausible for any known
aircraft.
On the evening of September 2, 1998, a 14 minute videotape was
recorded showing a similar strange triangle of lights, described
as red/yellow/blue, "patrolling the skies of Rishon Leziyon"
(sometimes translated as Le Zion or Letzion) on coastal Israel.
The camcorder is obviously handheld, with the usual shakiness of
a handhold with zoom. The zoom varies, showing perhaps 16X zoom
alternated with occasional shots of a crowd of onlookers. Not
surprisingly due to the limited ability of videotape to capture
the color of small objects, no color other than a slight pink was
detected.
The triangular array slowly turns clockwise, with lights
occasionally extinguishing, and being replaced with lights on
the opposite side. Figure 8 shows a collection of images of the
object, with timing information.
The on-screen time is shown where available; if not, the frame
number is given. Apparent size changes are presumably due to
zoom changes. The last image is that of an airplane; the center
bright light is an anti-collision strobe light.
The graph in Fig. 9 below shows the turning of the lights. Note
the object appears to make about 1.5 complete revolutions. The
turning is accomplished in slow turning, then bursts, both of
which slow during the video. After 150 seconds, the time is real
time as shown by the camcorder on-screen time display. Prior to
150 seconds, the time is estimated assuming the camcorder ran
continuously. This is not true; there are breaks of unknown
duration in the video. Timing prior to 150 seconds should be
regarded as speculative. Note also that the time covered is longer
than the video length; this is due to breaks in the recording.
Several frames show smearing of the light images due to camera
motion. No interesting high speed brightness changes were
apparent in such frames.
At several portions of the video, background stars appear to be
visible (circled at right). These are very useful in that true
object motion is apparent using the star as a reference point.
(In the portion shown, two stars are visible, allowing a zoom
change to be compensated for. In this and another brief segment
where two stars are visible, they are motionless relative to each
other, corroborating the conclusion that these points are indeed
stars.) These portions are:
At 30 seconds, over 42 frames (8 points measured, although many
more visible), object moves leftward and downward at a 45 degree
angle, at a speed of 2.5 times its own length per second.
At 97 seconds, over 32 frames (but only 3 frames showed a star),
object moves right and upward at a 70 degree angle at a speed
of 7X its own length per second.
At 124 seconds, over 21 frames (but only 3 frames showed a star),
object moves rightward and downward at a 25 degree angle, at a
speed of 9X its own length per second.
At 164 seconds, at 18:38:35, over 86 frames (9 measured), object
left and down at a 40 degree angle at a speed of 2X its own
length per second.
At 18:40:04, over 225 frames (18 measured with 2 stars), object
right and down at a 17 degree angle at a speed of 6X its own
length per second.
At 18:45:45, over 31 frames (only 3 measured), object right and
upward at a 28 degree angle at a speed of 11X its own length per
second.
At 18:49:24, over 67 frames (7 measured), object right and upward
at a 14 degree angle at a speed of 13X its own length per second.
At 18:52:36, over 333 frames, (9 points measured using 2 stars)
rightward and upward at a 44 degree angle, at about 1.4 times
its own length per second. At 18:59:44, over 211 frames, (7
measured) leftward and upward at a 48 degree angle, at a speed
of about 0.8X its own length per second.
"Its own length" refers to the maximum distance between lights;
the size of a possible dark supporting structure is obviously
unknown. "Frames" are the 1/30 second NTSC frames studied,
although the original PAL video is recorded at 25FPS. In all
cases, the object motion was a straight line over all measured
intervals, within measurement error limits. Note that faster
speeds correspond to fewer data points. As would be expected
in the case of high object speed, only a small number of samples
could be obtained before the star left the field of view. Also,
fast motion smears out and obscures a dim object, making a star
harder to find.
The variety of directions and speeds may be better understood by
looking at the series of lines in Fig. 10; the direction of each
arrow shows the direction of the object, and the length of the
lines represents the object speed. Note, THE LENGTH OF THE LINE
DOES NOT REPRESENT THE ACTUAL PATH OF THE OBJECT,
but rather its speed at the stated times. In summary, the description
of "patrolling" is corroborated; between 30 and 97 seconds, the
object turned almost completely around, turned right, down slow,
right, up fast, then slowed and turned left. Note the number of
near-right-angle turns. OTHER TURNS (WITH NO STAR AS REFERENCE)
MAY HAVE OCCURRED. Remember this motion is apparent motion; the
object could be moving toward or away from the camera and this
motion would not be visible.
Conclusion:
The object is obviously not an airplane, or some device on an
airplane; ignoring the lack of standard lighting, the path, and
the 16:1 range of apparent speed, is impossible. A turnable set
of lights on a helicopter could make this video, but I doubt that
numerous witnesses would fail to recognize helicopter noise. A
helicopter would have particular difficulty making a 1 minute, 180
degree turn silently. I know of no known cause for this phenomenon.
Jeff Sainio
MUFON Staff Photoanalyst
7206 W. Wabash Ave.
Milwaukee, WI 53223-2609
Back to Index Page
