UFO Evidence header

<< Return to the Main Page for this Case


The Mexico City, Aug. 6, 1997 Video of an Unidentified Object

Source: Bruce Maccabee, Ph.D, UFO Updates, March 1998


Summary: Report by Dr. Bruce Maccabee, optical physicist and UFO researcher, on his analysis of the Mexico City 1997 UFO video.

Below is a preliminary report posted on UFO Updates mailing list, before the article was published in MUFON Journal in 1998.


$#@& !!! (Expletive deleted.) Its just a model hanging on a string...a stupid hoax!!

That was my immediate reaction watched the beginning of the Mexico City video for the first time back in early November, 1997. Its wobbling and rotating saucer seemed just as big as life...but not as real. What immediately popped into my mind was a vision of a 1950s movie portrayal of flying saucers.... the movie with rotating saucers that crash into the U. S. Capitol building (Invaders from Space?)! I was disgusted!!!

Five seconds later I was no longer disgusted. Instead, I was surprised.

During the first seconds of the video I had envisioned a small model saucer rotating and wobbling as it was hanging from a string.

Then it abruptly moved to the right.

Thats no hanging model, I thought as it accelerated without any swinging back and forth characteristic of the pendulum motion of a small model at the end of a string. As the video continued I watched carefully for any evidence of swinging.

There was none.

I then reversed the video and watched the image of the Unidentified Object (UO) as it moved to the left toward its starting position.

Suddenly it stopped moving...as if running into a brick wall (crash dummies inside?).

I ran it back and forth several times.

Each time I became more convinced that the acceleration was very rapid, perhaps even instantaneous. Suffice it to say that I watched the rest of the video with more respect.

I then got out my ruler and started making measurements.

I wanted to know just how fast it had accelerated.

Thus began my investigation into one of the most interesting UO videos I have ever seen.

In late November I posted initial results of the investigation on the internet. This report provides some corrections and updates to those initial results.


Jaime Maussan, a TV journalist in Mexico City (who has collected numerous videos related to the Mexico City solar eclipse sightings of 1991) received the video and a letter from an anonymous source on Sept. 26. He watched the video and wasn't impressed because it was too clear, too obvious. It had to be a hoax, he told Michael Lindemann during an interview at the World UFO Congress in Brazil in December (published on the internet: cninews1@aol.com).

The video has a date stamp on it: August 6, 1997.

There are the voices of two men who were talking during the sighting. The conversation, which provides no details of the sighting, is as follows (translation by Carlos Guzman, MUFON director for Mexico):

1st voice ...........(expletive deleted)

2nd voice I cant believe it.... (expletive deleted)

1st voice It looks just perfect............

2nd voice Are you recording Carlos?

1st voice Yes, the light is on.

2nd voice .....(expletive deleted)

1st voice The disc looks great..

2nd voice Its not a plane, that...(expletive deleted)

1st voice I cant see it now. It moves beind the building, wait....

2nd voice By the other side?

1st voice Its isnt any more. (i.e., Its gone.)

Jaime had not heard of anything particular happening on Aug. 6, so his initial impression was that he should ignore this video. But after he watched it several times more he decided to air it on his TV program, Tercer Milenio, on September 27 and 28.

Then on Sept. 29 he located the buildings in the video and determined where the videographer had been on a balcony on the fourth floor of an office building. This is in the western part of the federal district of Mexico City (an area called Prolongacion Bosques de La Reforma, where the road Prolongacion Paseo de los Laureles meets the road Sindicato de Trabajadores de la Industria Militar).

Although the videographer has remained anonymous, Jaime told Michael Lindemann that since the Sept. 29 showing he has found more than a dozen witnesses by walking around the streets in the area asking people if they saw something unusual in August.

Several children claim to be able to draw the UO. One lady who was close to it claimed to have been burned, like a sunburn, as it passed overhead. Another claimed to have been under it and felt a downward force. This lady said that it turned away much faster than it appears on the video. Several said they saw lights on the bottom.

Michael Hesemann also interviewed witnesses on Dec. 3, 1997, while in the company of Alejandra Dehesa, a Mexican journalist who works for a different news agency (a rival of Maussan). They located the buildings which appear in the video and then walked around asking people if they had seen, or knew someone who had seen, the UO.

Ultimately by this manner they located and interviewed 7 witnesses and learned of 5 others who were not available to be interviewed at that time.

They learned that the office building where the video was shot has restricted entrance (a security system with guards; Hesemann was not allowed to enter) and hence the videographer must have worked for one of the companies in the building. There is no TV production company, computer company or video laboratory.

According to Hesemann, the witnesses came from all levels of society, with no connection at all.

Yet another witness investigation has been carried out by Lee and Brit Elders, who have located a number of other witnesses.

If it can be proven that these witnesses are not part of a rather large scale conspiracy, i.e., if the testimony is not a hoax, then this video will be important evidence of another reality right here on earth.

The MUFON representative in Mexico City, Carlos Guzman, has also investigated this sighting and has provided numerous photographs which show a typical urban setting with a few high rise apartment and office buildings interspersed between in the typical one or two story homes and stores in the area.


The analysis of the video has been a joint project by me and Jeffrey Sainio, MUFON photo/video analyst. (Jeff has previously presented several papers at MUFON symposia in which he has discussed the analysis of UFO photos and videos.) We have attempted to determine characteristics of the image of the UO and to use these characteristics to assess the liklihood of one of the following hoax hypotheses:

(A) a full sized model, assumed circular, and estimated at 6 m or 20 ft in diameter at a considerable distance (estimate 500 m or 1,500 ft or more),

(B) a small mechanized model several inches in size combined with partially fake and partially real scenery that is near the camera (i.e., within a room in the office building),

(C) some arrangement using optical/photographic methods (e.g. a combination of front and rear screen projected images with a mask shaped like a building for the rear screen projection), and

(D) a complete electronic fabrication using real video of the background scene combined with sophisticated computer image generation methods to create the UO image.

(These methods are discussed in more detail at the end of this article.)

The video analysis reported here was made possible by an unknown source in Mexico City who supplied a good copy of the video to a friend who subsequently sent it to me. This version (which is devoid of color for some reason) has a black and white direct copy of the video, a 1.6x magnified and slowed-down (half speed) negative version of the video (bright is changed to dark and vice versa) and a 7x magnified slowed-down video of the UO alone.

(I subsequently learned that the magnified versions were created by Jim Diletosso of Village Labs in Phoenix, Arizona, who also sent me a color copy. The color copy shows that the UO image has no color.)

Within a week of receiving of the video I placed on the internet (in early November) a preliminary analysis.

This paper presents some corrections and an update to that analysis.

The video begins with a "wide angle" shot and then, over the next 3 seconds, zooms in on the UO. Once zoomed in it stays that way.

The initial wide angle shot shows numerous nearby and distant buildings. The nearby buildings are noticeably darker than the bright sky background. The distant structures have low contrast with the sky, a fact which is consistent with the known very hazy conditions for August 6 in Mexico City. The distance to the building that the UO went behind is eastimated at about 500 m or 1,500 ft away. That would place the UO at a comparable distance or farther, assuming it was a real object out there.

The UO image is generally well centered in the video. The camera jiggle is obvious and looks as one would expect for a hand held camera. The jiggle is much more noticeable after zoom.

The UO started to move about 9 seconds after the start of the video. Its acceleration was abrupt (see below).

The cameraman apparently panned with the moving UO until it was out of sight about 24 seconds after the beginning of the video. The rightward motion was at a roughly constant rate (see below) and so was the pan motion (with jiggling).

After the UO disappeared behind the second, higher, building the cameraman kept the camera running for another 11 seconds (total video time is 35 seconds) and continued to pan to the right as if he expected to see the UO appear from behind the second building. This would be logical thing to do, considering the previous steady rightward motion.

The motion of the UO was always steady and rather slow. It did not simply hover for many seconds and then zip away at high speed, as reported in the MUFON Journal article of November, 1997.

The center of the UO was initially stationary while it rocked back and forth in such a way as to make it appear to be rotating about a vertical axis, a motion similar to that of a precessing gyroscope (i.e., a spinning top with its spin axis tilted from vertical and the spin axis rotating about a vertical axis through the support point).

The 7x video blowup shows periodic left-to-right motion of numerous diffuse (not sharp) dark areas or dark spots which seem to be on the rim of the UO. If these are fixed to the surface, then they suggest rotational motion, counter-clockwise as seen from above. This could be rotation of the whole UO or rotation of only an outer rim or ring. If the UO or part of it was rotating then perhaps the rotation was related to the periodic tilting or precession, as in a gyroscope. (In a gyroscope the precession rate is related to the spin rate and the torque acting on the spinning mass of the gyroscope.)

On the other hand, the UO might not be rotating at all and the dark spots might actually be some structural features which moved along the rim or edge of the UO, thereby given the appearance of overall rotation.

Whatever the case, this rotation and wobbling continued throughout the video (see below).

On a 14 diagonal monitor, using the direct video, the image size after the zoom is completed is about 27 mm wide and 8 mm high. The actual distance and size of the UO have not yet been determined (although this should be possible from witness testimony) nor has the focal length of the camera. Hence the size can only be estimated by comparison with objects of known size in the video. Unfortunately as of this writing there is no known size of an object in the video. However, one may assume that the spacing between floors in the building is at least 8 feet. The width of the UO image is about 2.5 times the spacing between floors on the building. This means width of the UO, if a real object out there farther away than the building, was at least 20 ft.

Without knowing the exact ratio of distances to the UO and the building it is impossible to determine how much larger it might have been.

Therefore, we assume as a conservative estimate that it was 6 m or about 20 ft wide. In this case its height was about 1.8 m or about 6 ft.

(Note: if further investigation proves the UO was larger than this then the results of the following calculations can be scaled appropriately. For example if the UO were 40 ft in diameter the calculated acceleration would be twice the value given below.)


The acceleration to the right is what initially caught my interest.

To make a quantitative estimate of the acceleration I measured, frame by frame, the spacing between the right end of the UO image and the left edge of the building image, thereby removing the camera jiggle from the measurement of UO position as a function of time.

The several sets of data I obtained by hand measurement are presented in Figure 1. Because of what I had seen while running the video I was not surprised to see that the UO accelerated rapidly. However, I was surprised to see just how quickly the UO reached a steady speed to the right.

The data in Figure 1 suggested that the acceleration occurred over a period of time no longer than two video frames or 2/30 sec. However, because of difficulty in making accurate position measurements in the presence of electronic noise and the oscillation (tilting) of the UO image I could not rule out the possibility that the acceleration took as little as 1 frame time (1/30 sec) or even 1 field time (1/60 sec).

(In standard video format a frame or complete picture is completed every 1/30 sec. Each frame is composed of two fields each taking 1/60 sec. Hence there are 30 frames or 60 fields per second. The first field is made up of odd horizontal scan lines 1,3,5, etc., and the second field consisting the the even lines, 2,4,6, etc., so that after the second field there are a total of 460 lines making up the complte frame. Hence each field is itself a picture of the scene, but at lesser vertical resolution than the complete picture or frame.)

Jeffrey Sainio used computer aided methods to produce more precise data. He obtained the horizontal position data for 20.45 seconds of the video, starting after the camera zoomed in on the UO, by measuring its position relative to the building, field by field (60 fields per second and 2 fields per frame) for 1227 fields, an effort that required many, many hours of effort.

Figure 2 shows its horizontal position relative to its initial position, i.e., the distance moved rather than the separation between the UO and the building (shown in Figure 1). The horizontal position as a function of time is in relative units (cursor units, vertical scale) vs time (in units of 1/60 sec). (The data points are close together and therefore make a thick line.) The hole in the graph corresponds to the time when the UO was (apparently) behind the building. The slope of the graph is proportional to the velocity. The graph shows that the UO accelerated quickly, in a fraction of a second, to a constant velocity and held that velocity until it disappeared behind the first building.

By the time it reappeared the velocity was slightly lower (the slope was not as steep; compare to the slope the thin line projected along the direction of the initial slope). The rightward motion continued at a constant rate until the UO disappeared for the second and final time, 27 fields (nearly 1/2 second) after the end of the graph.

A close up of the horizontal acceleration to the right is presented in Figure 3 which shows about 100 data points.

Jeff obtained these data by using a special process that measures image shift from field to field in order to reduce the effects of electronic noise. I have drawn a thin sloping line through the data.

Note that the data suggest the acceleration took place in 1/60 sec.

Further analysis is being done on this acceleration.

Measurements made using the direct video monitor show that from the time of acceleration to the time when the leading edge of the UO image contacts the building image is about 45 1/2 frames or 1.52 sec (+/- 0.1 sec).

During this time the UO image moves about 2.15 times its own length. It then requires 21 frames or 0.7 sec to travel its own length and disappear behind the building.

The total distance moved from the start of motion to complete disappearance is therefore 3.15 times its length and it does this in about 66.5 frames or 2.22 sec.

Assuming a size of 20 ft, the speed was about 28 ft/sec or 19 mph.

If the speed change from zero to 28 ft/sec occurred in 1/15 sec the average acceleration during this 1/15 sec would have been about 28/(1/15) = 420 ft/sec^2 which corresponds to 13 gs or 13 times the acceleration of gravity (32.2 ft/sec^2 or 9.8 m/sec^2).

If the speed change occurred in 1/30 sec the acceleration was about 26 times that of gravity.

Finally, if the speed changed occurred in 1 field time (1/60 sec), as suggested by Figure 3 above, the acceleration was about 52 times that of gravity.

If you think of your own weight as caused by 1 g of gravitational force, you can try to imagine what it would feel like to suddenly weigh 13 times or 26 times or 52 times your own normal weight. (Time for a crash diet!)

The less steep slope of the graph after it reappeared from behind the building means that while it was (apparently) behind the building it slowed its rightward velocity to about 0.66 of the initial velocity.

Assuming it was 20 ft wide this final horizontal velocity was about 18 ft/sec or about 12 mph.

At the same time that the UO accelerated to the right it also accelerated and then moved upward at a constant speed.

A graph of the height vs time shows a continual climb until a few seconds after it reappeared from behind the building, at which time it abruptly stopped its continual climb and began a very slow climb upward.

Again, assuming a 20 ft wide UO, the upward speed was about 9 ft/sec or about 6 mph.

Except for the growth in image size during the initial zoom, the size of the UO image changes very slightly (a few percent) during the first 17.65 seconds of the video.

The slight changes are consistent with what one would expect if the UO traveled along a straight or nearly straight trajectory that was essentially perpendicular to the line of sight toward the building.

Then, during the last 2.8 seconds before the UO disappears (for the last time) behind the second building, the image shinks by about 20%.

The shrinkage is consistent with the assumption that during this time the UO was moving away from the camera, an assumption that is supported by the decrease in contrast of the UO image relative to the sky. (Under hazy conditions the contrast with the sky backgroud diminishes as distance increases.)

By assuming the acceleration occurred in much less than a second and by assuming a particular distance after the acceleration ended one can use the following 2 equations to calculate the component of velocity away from the camera, Va: (1) Si(t) = SaF/D(t), where Si(t) is the image size as a function of time, t, after acceleration ends, Sa is the actual size of the object (e.g., 20 ft), F is the effective focal length of the camera (unknown) and (2) D(t) = Do + tVa, where Do is the distance when the acceleration ended. The image size at t = 0 is 86 units (relative size) and at t = 2.8 seconds is 70 units. Hence equation (1) at t = 0 yields SaF/Do = 86. Then at t = 2.8 equation (2) is 70 = 86/(1 + 2.8Va/Do) which yields Va/Do = (86/70 - 1)/2.8 = 0.082. The actual value of Va depends upon the assumed distance, Do. If Do = 1,500 ft its speed was about 123 ft/sec (84 mph).

The clear implication of these calculations is that the UO traveled away much more rapidly than it moved to the right (at 12 mph).

The UO must have made a sharp left turn (at East Gate?) at the same time that it accelerated away. The turn angle is estimated at about 80 degrees.

So, here we have one of those (nearly) 90 degree turns people have reported for years!


The wobble motion has been carefully measured by Jeff Sainio.

Figure 4 shows the periodic up-down motion of one end of the UO image as a function of time with a sine wave superimposed.

Although the data are noisy it is nevertheless impressive to see that a single sine wave equation can fit the whole series of 18 3/4 cycles without being affected by the accelerations that occurred.

(There are several cycles missing in the middle when the UO was behind the building.)

The period of the oscillation is 1.065 sec (a frequency of about 1 Hz). To fit the whole curve the phase of the sine wave was set at 173 degrees.

The previous graph shows that the wobble motion is constant. It suggests continual rotation of the major symmetry axis of the UO about a vertical axis through its center.

The 7x blowup, which shows the UO image field-by-field, provides evidence of another, more basic sort of rotation. This is the rotation of the object about its major symmetry axis. The moving dark areas suggest a rotation speed of about 1/3 of a revolution per second (1/3 Hz) about the symmetry axis. The two rotations taken together make it appear that the UO was rotating about its symmetry axis at 1/3 Hz, while its symmetry axis was rotating around a vertical axis at 1 Hz, just like a spinning, tilted top that is precessing about a vertical axis drawn through the support point of the top (e.g., the point on the table the top spins upon).

The UO's appearance of being like a precessing top is seductive, but neither torque free precession nor torque driven precession of a rotating object seems consistent with the apparent precession rate (1 Hz) being larger than the spin rate (1/3 Hz). (Of course, there could be within the UO a rapidly spinning object - a source of high speed angular momentum - which is not visible from the outside, the precession might be caused by a torque due to magnetic fields and/or gravity or something else.)


In general there are three possibilities for any UO sighting: it is a hoax, a misidentification or the "real thing" (a True UFO - unexplainable as a conventional phenomenon).

The possibility of a misidentification seems unlikely here because of the shape.

It wasnt a bird or a plane or Superman (.....well, maybe Superman).

One can barely imagine that this was just a misidentified blimp or an oddly shaped motor-driven balloon of some kind that just happened to be passing by.

Yet, this is the only conventional (i.e., not a True UFO), non-hoax hypothesis that could be consistent with witness testimony about seeing a real object.

However, this hypothesis must be rejected because of the video evidence which shows anomalous accelerations of the UO.

What if it were a full scale hoax involving a large object, 20 or more feet in size, designed to look like a True UFO that could fool unsuspecting witnesses?

This could be a motorized blimp or simply a wind borne or tethered balloon of an odd shape.

However, even with the most advantageous combination of assumptions about mechanical capabilities of this blimp/balloon one cannot explain the initial "instantaneous" acceleration to a constant velocity at the beginning of the video, the sudden stop in its vertical ascent, the rapid right angle turn with large acceleration just before it went away and the lack of effect of these accelerations on the continuous rocking motion.

Finally, it is hard to imagine that the unsuspecting witnesses would not realize they were just seeing (and hearing?) some motorized blimp or a balloon dragged by the wind.

Alternative hoax hypotheses that do not involve a full scale model are based on the assumption that the witnesses dont count. Either they were all delusional or part of the hoax conspiracy.

One possibility is that a several inch sized, motorized MUFO (model UFO) was videotaped inside a room in real time while the MUFO was moved past a window with a camera looking through the window at the background scene.

In order to make the MUFO appear to go behind the real building out there it would be necessary to have inside the room a flat model of the building which would be slightly closer to the camera than the model itself so the model could move behind it.

Of course, the edges of the model building would have to be perfectly aligned with the edges of the real building as seen from the camera and this alignment would have to be maintained throughout the video despite the camera jiggle, something that would be difficult to do.

There are many other details of this hypothetical MUFO method which would make it both time consuming and expensive although probably not impossible.

An alternative is a rear screen projection method in which a stationary view of the whole background scene from farthest left to farthest right is projected onto the front side of a large screen (i.e., the projector is on the same side as the video camera that records the scene). This screen is assumed to be translucent rather than opaque. There is also a projector on the opposite side of the screen (rear projection) which creates a movable and dynamic image of a UO on the other side of the screen. Because the screen is translucent some of the UO image leaks through the screen. Thus the UO image can be videotaped along with the front projection of the real scene. The UO image is then moved sideways to create periodic tilting, rotation, acceleration, steady motion to the right, etc., while the hand held video camera follows the image, thereby recording (a photograph of) the real scene with the supposed UO moving against the sky background. It would be necessary to place an opaque mask with the shape (outline) of the buildings at the rear of the screen to block the UO image from the rear projector at the locations where it seems to be behind the building. This method, while perhaps not as difficult as the MUFO method described above, would still require considerable effort and equipment.

Finally one must consider the all-out electronic fake.

In this case one might imagine that the cameraman videotaped the background scene with a handheld camera to make it look as if he were following the path of a True UFO that hovered for a few seconds and then moved across the scene. Then the computer graphics geniuses would create a classic UFO shape and make it appear to rotate and oscillate (tilt) at a steady rate. Then they would have to analyze the random camera jiggle motion of the background scene on a field by field basis. (Note: an alternative would be to have the camerman videotape the scene with the camera on a tripod and then the computer geniuses would add a synthetic camera jiggle later on.) Then on a field by field basis they would have to place the EUFO image on top of the background scene (hazy sky) in such a way that the EUFO would appear to move in a manner consistent with the motion of True UFOs. Finally, after about 1440 fields of EUFO video had been generated in this way the geniuses would be finished, since the last 11 seconds of the video dont show the EUFO.

Considering the number of possibilities for faking this video in a studio I would have to say that if Stephen Spielberg had given me this video I would have asked if he had been talking to George Lucas recently!

On the other hand, if the witnesses are credible these methods can be tossed out the window.


As of the time of this writing (early February) there is no firm conclusion.

There may be other hoax hypotheses that should be considered.

Investigation of the video and of the witnesses continues with a major effort to determine which hypotheses can be rejected.

If it turns out that all hoax hypotheses can be rejected then this will become one of the most spectacular True UFO sightings ever recorded on video.

Article ID: 869


  FAIR USE NOTICE: This page may contain copyrighted material the use of which has not been specifically authorized by the copyright owner. This website distributes this material without profit to those who have expressed a prior interest in receiving the included information for scientific, research and educational purposes. We believe this constitutes a fair use of any such copyrighted material as provided for in 17 U.S.C § 107.