Photo Analysis Report.
"The Case of the Speedy Celestial Dove"
A dove landing faster than a speeding bullet? Only a dove from outer space (hence, celestial) could have made the vertically oriented white image in the photo which is the subject of this analysis.
Figure 1 was intended to be a picture of a dove sitting on a dog's nose. Figure 2A, which preceded Figure 1 by an estimated 10 seconds, shows what should have appeared: the wall of the room, a birdcage with a white sheet and a bird on it, a desk and chair at the left and a man in a striped shirt at the right side, a dove at the lower right of the cage and a dog at the bottom. Figure 2B, a brightened version of Figure 2A shows these images clearly. Figure 3, taken a short time later, provides a close up of the cage and birds. Figure 2 A is presented with no brightening for comparison with Figures 1 and 3 (no brightening; Figure 3 is a magnified view). The dark areas in Figure 2B have been artificially brightened by increasing the effective "gamma." By increasing "gamma" one can brighten the whole picture, but the dark areas are brightened more relative to the bright areas. Note that in Figure 2A it is difficult to distinguish the image of the dog, whereas in Figure 2B it is very easy. The relative brightness's of the unbrightened pictures and the effects of the brightening are important to understand in the analysis that follows.
The "speedy celestial dove" photo, Figure 1 , is immediately notable for two features: the vertical white streak (speedy celestial dove in vertical takeoff
or landing?) and the very dark background as compared to the background in Figure 2. Consider first the background because this part of the photo should not be unusual or "unexplained," yet it is puzzling. These were flash photos taken with an ordinary (simple) flash camera (fixed settings). If all lighting conditions as determined by the flash were the same the reflected light from the wall, chair, man, dog, cage, etc. should be the same in Figure 1 as in Figure 2A, but obviously there is a great difference: the background in Figure 1 is so dark that all one can see of the background is a small portion of the back of the chair immediately adjacent to the vertical white image. If one looks carefully one can also see a bit of the sheet-covered cage at the left edge of the "speedy dove." Figures 4 and 5 are artificially brightened versions of Figure 1. In Figure 4 the effective gamma has been increased and in Figure 5 both the brightness and contrast were increased without changing gamma. The brightness increase was necessary to bring out background details. Even after brightening the background is not as clear as in Figure 2B, the brightened version of Figure 2A. So, the immediate question is, why is Figure 1 so dark?
No definitive answer on this can be given since the camera characteristics are not known. However it is possible to make a reasonable hypothesis based on an assumption about the operation of the camera. In some cameras there is a light integrating sensor (which creates a running total of the amount of light received by the camera) which keeps the shutter open until the camera "calculates" that there has been enough light for a good exposure (i.e., until the running total amount of light reaches or exceeds a predetermined value). Then the shutter closes. This calculation is done automatically and takes into account the "speed" of the film (ASA or ISO rating; this sets the predetermined value of the running total of light received by the camera). The higher the speed of the film, for a given aperture (lens opening diameter) and light reflection from the scene, the shorter the shutter time. In a flash picture with this sort of camera, if the reflected light gets brighter, the shutter time gets shorter. Hence, a very low brightness reflection will cause the shutter to stay open the longest it is allowed (with some cameras this could be several seconds). Conversely, a very bright reflection could return a lot of light, thereby causing the shutter to close quickly.
If we assume that the camera used to take these photos had this light integrating property, then we can understand the difference in brightness of the background images in the following way. For Figure 2 the shutter stayed open long enough for the reflection from the ordinary objects in the photo, and in particular from the wall, the man, etc., to make a useable picture. One would expect the same sort of brightness's of the wall, man, etc., in Figure 1 under ordinary conditions. However, the implication of the overexposed "speedy dove" image is that there was a bright reflector (or a source of light) directly in front of the camera which caused the light sensor to close the shutter before the light reflected from the background had time to make a useable image, i.e., the integration time was shortened by the bright reflection from the "dove." Of course, this explanation assumes explicitly that the cause of the white image was a "real light reflector," i.e., a real object in 3-D space capable of reflecting light.
If the camera did not have the light integrating property but instead the shutter stayed open a fixed time, then one must make other, more bizarre assumptions about the light received by the camera from the background objects (wall, man, cage, chair, etc.). For example, one could imagine that under the unusual condition which created the "dove", light from the brightness of the flash itself was suppressed in some way, that the light from the flash was largely prevented from reaching the background or that light reflected from the background was, in some way, reduced considerably. Light reduction in this case might be a result of increased opacity of the atmosphere...a condition totally at odds with normal physics. But then, it would appear that the "dove" is totally at odds with normal physics.
It has been stated that there was a witness to the very brief appearance of an "ethereal being" or whatever at the time of the flash. It apparently appeared that something was moving downward at a high speed. If this were true, in some way, and this something was lit by the flash, then that might explain the very dark background images IF the camera had the light-integrating property discussed above. If there was no light integrating property then we must attribute the dark background to some property of the alleged moving "entity." Whether or not such an entity could partially opaque and a reflector or light (or a brief source of light) is a subject for discussion that goes beyond this analysis.
Consider, now Figures 4 and 5. These are brightened versions of Figure 1. For Figure 4 the gamma was increased to 3.0 to bring out the background, whereas in Figure 5 the brightness was increased by 100% and contrast was increased by 12%. These brightened versions of Figure 1 show the following facts. First, comparing with Figure 2, we see that the camera has rotated slightly to the right, but the location of the camera has not moved (for example, the images of the vertical slats of the chair overlay the image of the wall socket in exactly the same place). Second, the background scene IS present, i.e., the "dove" has not caused a completely opaque blockage of the light from the background. This means a that either the presence of the "dove" caused the air to be "optically dense", that is, to absorb some (but not all) light at the left and right of the "dove," or that the "dove" caused the camera to integrate light for a much shorter time than for Figure 2. As a physicist, when offered the choice between two physically impossible conditions, I would choose what appears to be the least physically impossible, which in this case requires that the camera be of the light integrating type. In other words, I choose to accept the idea that the "dove" affected the "real world" only within the confines of its image and not outside the image. However, this is a debatable matter and someone might find a "better" explanation.
Regarding the "dove image" itself, consider Figures 6 and 7. In Figure 6 the gamma has been reduced to 0.1, the brightness by 20% and the contrast by 40% in order to investigate structure within the "dove" image. This is further elucidated in Figure 7, which is a blowup and indicates the nearly horizontal striations that appear throughout the image (these can also be seen at the top of the image in Figure 1) and other "linear features" which may or may not be real attributes of the "dove" image. That is, these features might be artifacts of the photo process.
So, the big question is, what could have made such an image? A reflective object close enough to the camera to be somewhat out of focus and which also moves very rapidly could, I imagine, make an image which has the basic characteristics observed here: a vertically smeared bright reflection that is fuzzy at the edges.
To further elucidate the "anomalousness" of this image, let us do a "gedanken experiment" ("thought experiment," a concept made famous by Einstein!). Assume a non-anomalous object were to fall in front of the camera. In particular, I choose to do my experiment with a cue ball (white). Let it fall at a distance of 2 or 3 feet from the lens so that it is not far enough away for good focus (i.e., it is close enough so that the edges are fuzzy). I know, from having done VIDEO experiments of this sort (falling ball) that the image, if shot in STEADY light (not a flash) would be a wide vertical line. The width of the line, I, would be given by the width of the ball, W, times the focal length of the lens, F, divided by the distance to the ball, I = WF/D (actually a bit bigger than this if not perfectly focused; for example, a 7 cm diameter ball at a distance of 100 cm from a camera lens with a focal length of 3 cm would make an image 0.21 cm wide if focused; a bit wider if not focused). The length of the vertical line would be given by L = vt + 1/2 gt^2, where v is the velocity when the shutter opens, t is the shutter time and g is the acceleration of gravity, 9.8 m/sec^2 or 32 ft/sec^2. To be "perfect" add the width of the ball image, I, to this length.) If the falling cue ball were photographed with a flash, the length of the image would be determined by the duration of the flash (the width would be the same as before). However, whereas with steady light illumination the brightness of the vertical line image would be (nearly) constant with height, when a flash is used the brightness of the image would vary considerably along its length. Consider that the brightness curve, that is the light output vs time, of a flash is shaped like a "mountain," perhaps with a nearly flat top. That is, the light increases rapidly
but not instantaneously from zero to some maximum and then the light brightness decays, but does not drop instantaneously to zero (the light curve is not a rectangle on a brightness-time graph). Returning to the falling cue ball we see that, starting at the top of the line image and moving downward (i.e., moving forward in time as the ball falls) one would find the image brightening from zero ("fading in") to a maximum value, then perhaps maintaining a maximum value for some distance (or time) if the flash bulb maintains a constant brightness for a short time, and then fading out as the flash dies. In other words, the brightness graph of the image would have the form of a "mountain" perhaps with a nearly flat top. Looking at Figures 6 and 7 we see that there is, in fact, this vertical variation in brightness, with the image fading in at the top and fading out at the bottom rather than being abrupt in its appearance and disappearance. (The fade-in at the top of the much brighter image in Figure 1 is apparent to the naked eye but the fade-out at the bottom is not as easily noticed.) Thus it appears that the overall structure of the "speedy dove" image is consistent with what one might expect for an object moving rapidly past the camera in a vertical direction although, of course, it is impossible to tell from the picture whether the object was moving up or down. (Note: the fact that an object falling under the force of gravity would increase in velocity as time goes on means that its IMAGE would get even dimmer at the bottom since the IMAGE would spend less time at any particular location near the bottom of the picture than it would near the top, thus creating less exposure near the bottom of the picture. This would be yet another reason to expect a decrease in image brightness with increasing time or distance measured downward on the image.)
The falling cue ball analogy does not explain the brightness variations or structure internal to the image, as illustrated in Figure 7. An object of constant reflectivity, such as a cue ball, should make an image with no structure (or the structure smeared to "smoothness" by the continual motion). The striations suggest a slight periodic fluctuation in brightness of whatever moved in front of the camera. Unfortunately any real "shape" of the object or phenomenon has been blurred by the presumed (witness reported) motion. I see little chance of recovering, with any reasonable confidence, an actual shape from an image such as this. Even if one knew what the shape was and how it moved it would be difficult to create an effectively unsmeared image.
What about the "linear features?" The nearly vertical feature obvious near the bottom in Figure 7 (parallel "lines" down the center of the image) could appear even if there were vertical motion. However, the non-vertical linear features are contrary to the continual motion hypothesis discussed above. Whereas features with "vertical structure" would be elongated into further vertical structure, and thus appear as vertical structure, features with horizontal components would be blurred by motion. Thus if the features indicated on Figure 7 near the vertical center of the image are real attributes of the object or phenomenon as viewed with no motion, then it must have been stationary during at least the brightest part of the flash. But then, if these features indicate stationarity during the flash, then the nearly horizontal striations must also be essentially unblurred features of the object or phenomenon. On the other hand, if the witness had the impression of something moving downward, then the features near the vertical center must not be characteristic of the object but rather artifacts of the photography combined with the human tendency to look for order in the disorder (a Rorschack test for the photo analyst).
An amusing aspect of the striations is that they are not perfectly horizontal. Instead, the left end of each faint line is a bit lower than the right end. Furthermore, the left end of each line is very nearly (or may be exactly) at the same level as the right end of the next line down. (Note: there is a lot of "breakup" or randomness to the striations. The following discussion is applied mainly to the striations at the top of the image in Figure 7 where they seem most consistent.) This is the characteristic of a "scanning raster" such as one finds in a TV set. In that case one has a moving "object" (the illuminated dot on a TV screen) that travels to the right very rapidly while moving downward slightly so that when it jumps back to the left (scan retrace) it is down lower than the beginning of the line just above it. In other words, the scan lines in a TV are slightly slanted (although in the opposite direction to what is observed here). Hence the striation structure could, in principle, have been created by some light reflector (or source) that moved quickly to the left and then even more rapidly to the right and then again to the left and again more rapidly to the right, etc., as the object or phenomenon moved downward, also at a high rate. In this case the tilt of the striation "scan lines" would be related to the scan rate (time to go from right to left), the "retrace time" (very short compared to the scan time) and the downward velocity. Now, granted that it would seem unlikely some reflective phenomenon would be created as a "raster scan", let us consider another possible cause for tilted "scan line-like" striations. Imagine a rotating cylinder with reflective "dots" on it. Assume it spins about its axis as while moving along its axis. Furthermore, assume that for each dot at a location on the cylinder surface there is also one at the opposite side. Thus, as the cylinder rotates, just as one dot disappears at the left another appears at the right and at the same distance along the cylinder. Now, if the cylinder has moved along its axis at the same time as rotating, then the dot moving right to left will also move downward and just as it disappears at the left another will appear at the right at the same "altitude" (or distance along the motion track) as the one at the left. (Note: this sort of operation would create a perfect raster scan with "instantaneous" retrace.) (Note: this is not consistent with a "screw thread" in which the distance from one thread peak to another is 1 divided by the pitch. Starting at one side of the screw and following a thread around one finds that by the time one reaches the other side the tread has advanced 1/2 the pitch distance. Thus moving perpendicular to the axis from the peak of a thread at one side of a screw to the other side does not take one from the peak of one thread to the peak of the next thread, but rather from the peak of one thread to the valley between thread peaks.)
There is another possibility: that the object or phenomenon did not have a continuous motion but rather a jerky motion which allowed it to spend a "short" time at any location along its path and then in a much shorter time move to the next location. The approximate periodicity of the striations would then indicate an approximate periodicity in its jerky motion and, in particular, each striation might, therefore, actually be a brief "photo" of the top (or bottom) edge of the object or phenomenon.
Since there really is nothing to "go on" in terms of interpreting this photo from the point of view of conventional physics, one may as well take the witness' testimony into account, in which case the "moving downward" assumption of "speedy dove landing" hypothesis would lead one to the following conclusions:
*One of Americas foremost research photo analysts completed this
final report in the Fall of 2000. The researcher now requests to remain
anonymous and told me why in vivid detail.
I've honored the request and indeed, wished him well and blessed for the work he
had done for me..
- during the period of time of the camera flash there really was some reflective "object or phenomenon" in front of the camera and this either caused the shutter time to be shorter than normal or, in some way, it affected the opacity of the air in its vicinity, thereby causing the background images to be considerably underexposed. In other words, the photos appears to be consistent with the claim that something went zooming past the camera (downward) at the same time that the flash went off.
- whatever it was appears to have oscillated slightly in brightness or reflectivity. This oscillation might have been a result of rapid rotation of reflective portions of the object or phenomenon, as discussed above, or it could have been the result of actual changes in reflectivity without rotation, or it might be a result of jerky motion, or something else.