Мехико (Мексика). 6.08.1997. Эту видеозапись сделал Fernando Camacho в Мехико в середине дня 6 августа 1997 года. Использовалась цифровая камера "Sony". Общая продолжительность фильма - 25 секунд.   http://alienUFOs.com/     Этот фильм довольно долго сохранял репутацию неоспоримого доказательства реальности инопланетян, чему способствовали опрометчивые заявления некоторых уфологов.  Анализ, проведенный профессиональным экспертом по фото- и видеосъемкам Джеффом Сайнио, убедительно доказал, что видео из Мексики - подделка, выполненная с помощью современных компьютерных технологий. Эксперт обратил внимание, что движение камеры вызывает некоторое размазывание изображения пейзажа, но изображение "объекта" остается резким. Кроме того, подрагивание и покачивание камеры, и, как следствие, подрагивание изображения пейзажа плохо согласуется с подрагиванием "НЛО". Очевидно, что при реальных съемках такого быть не может.  Полный текст статьи Сайнио я привожу ниже:

A video of a wobbling disk hovering over Mexico City on Aug. 6, 1997, has generated much controversy and flooded the Internet with shots of the saucer. Opinions of the UFO's validity have ranged from "best evidence of alien presence" to "falling down laughing fake," but little hard analysis has supported these conclusions. I am a firm believer in Lord Kelvin's comment that "when you cannot measure it, when you cannot express it in numbers, your knowledge of it is of a meagre and unsatisfactory kind." Using machine-recognition measurement techniques involving no human bias, I found three indications that the video is a fake requiring thousands of measurements by the faker to create, but still not satisfying Kelvin enough to avoid detection.

These indications are: shakiness of the camera causes smearing of the image-except the UFO, which is always clear and sharp; rotation of the camera causes the view to spin-except the UFO, which remains level (the wobble of the UFO makes this difficult to notice); and the "hovering" of the UFO isn't quite perfect. Instead, the UFO altitude varies, depending on the bouncing of the camera. Imagine taping a 747 on takeoff and crashing it by shaking the camcorder. Absurd? This video shows the same effect.

Original quite bouncy

The original video is quite bouncy, typical of a hand-held shot. This makes analysis difficult, unless one is able to measure the bounciness and compensate for it. My custom software "freezes" the terrain, so the resultant video appears to be from a tripod-mounted camera. Subtle features are then easier to spot.

The illustration shows both stable and bouncy contrast-enhanced images (l/20th second apart in time). Note that in the lower bouncy image, the windows, terraces, and building edges are poorly defined due to the skakiness of the camera. In the upper stable image, all these features are more clearly focused. But in both images the UFO is equally sharp in focus. The gradient lines show that the bouncy buildings' edges require 4 video lines to traverse from light to dark; only 2 lines are needed to make the transition in the stable image. This problem is seen through the entire video; the UFO image is always sharp regardless of the camera bounce. This indicates the UFO wasn't in the video when the camera was shaking, but was added later.

Altitude changes

The illustration also shows that the UFO altitude changes as the camera bounces. The effect is not large; at most, it is about half the thickness of the UFO. The "hovering" UFO is seen to bounce slightly. Note that in the lower image the UFO is slightly higher above the reference line over to the building ledge. After a few false starts, I finally determined that the UFO bounce matches, with 72% reliability, the velocity of the camera shakiness. (The remaining 28% appears to be inaccurate measurement by the faker who pasted the UFO into the scene.) If the camera shake makes the scene move downward, the UFO altitude increases, and vice versa. The faster the camera is moving upward, the lower the UFO altitude. One could think this might be due to chance, and such a possibility can never be eliminated. But the possibility of this occurring randomly is about one in ten trillion trillion trillion trillion. Yes, worse than winning the PowerBall lottery four weeks in a row. This problem is what would be expected if the UFO were pasted into the video, but 1/150 second late.

The camera not only bounces, but also rotates slightly. Rotation is most obvious when the UFO passes over the second windsock atop the buildings. The rotation is not large, only 1.6 degrees at maximum. Since the UFO wobbles about seven degrees, the camera rotation has no obvious effect on the UFO wobble. But by removing the periodic component of the UFO wobble, the UFO can be seen to NOT rotate with the rest of the scene. Also at this point, the UFO will be seen to "dip" in altitude, getting closer to the windsock below it.

Both problems expected

Both of these problems are expected if the faker used the first windsock to measure the camera bounce. Since switching reference windsocks involves some tedious calculations, a reference point is often used as long as possible until it leaves the field of view. So these problems are expected for a lazy faker.

A chance wobble?

Did, by chance, the UFO happen to wobble at the same time the camera did, to create this effect by coincidence? The chances of this are about seven in a million. Better than PowerBall, but still poor odds. Given the noise inherent in video, what is the chance that this video is real, and that the problems are due to the imperfect nature of video? Imagine the whole Solar System packed solid with valid videotapes in every direction, a million times further than the orbit of Pluto. You would have to analyze all these videos to find one having these problems by coincidence. These characteristics show that the video is the result of an otherwise-real video shot as seen, but with the UFO added later. A "paste on lighter" command is used to add the UFO, so it is added to the lighter skies but disappears "behind" the darker buildings. Using this command, I easily put a whole armada of UFOs "behind" buildings.

Why did a fake contain these errors? Smearing the UFO image takes some work, and since the smearing is barely obvious, the faker apparently decided not to do the extra work (or simply didn't know of the problem). Perhaps the faker's software doesn't support this detail, or perhaps the UFO's small size and relatively poor contrast were designed to masquerade the problem.

Incredibly tedious

To add the UFO image to the video, the faker had to measure the position of some reference point, 60 times in every second of the video. (About a thousand pairs of measurements total.) This is incredibly tedious unless one has software to automatically recognize image position and rotation. (I wrote custom software to make these measurements; it is not publicly available.) To correct for rotation, one must measure TWO points, not one. For such slight rotation, the faker probably didn't think the error would be noticed. Perhaps the UFO wobble was added to disguise the fact that the UFO didn't rotate with the rest of the image.

The UFO altitude problem is what is expected if the tedious camera-bounce measurements were smoothed or averaged in a non-symmetrical way. (Ask your statistician if you don't understand this; expect a long-winded reply.) This video gives a good example of rules that good investigators should follow:

1. Be suspicious of anonymous reports.

2. Note oddities in the case; here, a daylight sighting of a huge craft in one of the most crowded cities in the world, but no witnesses until after televising the video.

3. "Expert's" opinions are meaningless unless they can give repeatable, measurable procedures by which they reached their opinions.

Jeff Sainio. Analysis of Mexico City video   (Three indication of a hoax) / MUFON UFO Journal, N 366 (October 1998).  Текст любезно предоставил М.Герштейн

Главная страница | Каталог ложных фотографий НЛО