From: Bruce Maccabee <brumac.nul> Date: Tue, 26 Nov 2002 16:09:40 -0500 Archived: Tue, 26 Nov 2002 09:44:54 -0500 Subject: Re: Roswell Had Victims? >From: Gary Anthony <garyant.nul> >To: <ufoupdates.nul> >Date: Mon, 25 Nov 2002 23:47:45 -0000 >Subject: Re: Roswell Had Victims? >>From: Richard Hall <hallrichard99.nul> >>To: ufoupdates.nul >>Date: Mon, 25 Nov 2002 16:50:12 +0000 >>Subject: Re: Roswell Had Victims? >>>From: Tom Bowden <tomrbowden.nul> >>>To: UFO UpDates - Toronto <ufoupdates.nul> >>>Date: Sat, 23 Nov 2002 17:14:17 -0800 (PST) >>>Subject: Re: Roswell Had Victims? >I agree with Dick on his point of faces in clouds. However, >amorphous images aside (as Neil Morris may corroborate to some >degree) it has not prevented us from sending photographic >representations of the Ramey Memo (both enhanced and >computerised samples and normal best photographic paper copies), >to two separate competent 'cryptanalysts' (who have no interest i>n UFOlogy) a few months ago. Status - currently awaiting >responses! Apparently the process takes time, especially on a >limited budget... Checking out whether the Ramey Memo can be >deciphered is a worthwhile project, it is one of the unanswered >questions in the Roswell arena that needs addressing properly, I >am encouraged others have thought of engaging in it, though for >my two-penneth's I am of the opinion this requires the input of >qualified experts outside of ufology before anything useful >(either way) can be learned. Whether this will happen is >questionable. <snip> >From an American perspective we are talking about guys who can >decipher a range of blurry texts from long range photography, >even in some cases from spyplane aerial reconnaissance. To >emphasise an important fact, this type of cryptanalysis is >worked from technique, application and pure brain power, there >are many limitations and this is no small task An area of >cryptology called cryptoeidography (which Bruce Macabee may be f>amiliar with?) is a good analogy for this problem. >Cryptoeidography is very effective in rendering either pictures >or text on an image secret, one employs cifax which modifies >electrical patterns to distort an image while the latter i>nvolves actual optical alterations to it. I am not suggesting >either has been done in the case of the Ramey Memo, but in >reference to the latter, an expert can confirm the second form >of altering optical images can be very effective in disguising >content. Here is the kicker, there are natural counterparts to >this type of optical concealment which sometimes crop up in >photographic images that possess text or similar patterns. This >is a major factor and consideration for the prescribed >undertaking. >If anyone can make head-way with an enquiry like this though, >t is in a specialised area of cryptanalysis, which should >also be capable of pointing out any limitations. SNIP >Suggestion of a simple 'scientific' method for Ramey Memo >analysis to substantiate or refute David Rudiak's assessment may >go something like this: >Encourage the participation of at least two independent .cryptanalysts and ask them to: >1. Obtain clearest photographic representations of Ramey Memo. >2. Allow a photographic expert (linked with cryptanalysis) to >enlarge, enhance and manipulate the images in a number of useful >ways to obtain workable sample/s. >3. Submit samples to any feasible analyses >4. Publish all conclusions and peer review. Of course it would be "nice" to have cryptanalysts ... or anyone with experience, analyze the Ramey message., However, as Dr. Rudiak has pointed out, understanding the message, which appears to be in plain text (not encrypted, we presume), comes down to correctly identifying individual letters, such as the V, I, and I in the word that Dr. Rudiak used to illustrate his technique. Identifying each letter is a problem of separating the "signal shape" from the "noise shape." The noise shape is a complex area density of dark spots... the photographic grains. Each grain has a probability of registering the correct amount of exposure level for its position in the image. One way of looking at it is that there are "holes and excesses" in the image. Consider the letter I, a straight vertical line (perhaps with a "cap" at the top and bottom). Suppose this letter is correctly focused onto the film. Since the background paper is white, the letter I appears as a dark image i.e., an unexposed (or at least less exposed) area of film which is surrounded by "white" or highly exposed areas. (Of course, the original picture is a negative, but I am talking about the printed format. IN the negative the white paper will look dark and the letter will look bright.) A correct photograpic representation would have a uniform density of _unexposed_ film grains running along the location on the film where the I is focused., That is, the I will appear as a straight dark line surrounded by uniform white. Would not require cryptanalysis to identify the I. Even "richer" for photographic information would be more complex letters like a, e, V, etc. Sounds nice, but sorry... things don't work that way. The focus is not perfect so the image on the (negative) film has brightness gradients where one would like to have sharp boundaries. The focused line might not even be straight because of aberrations in the lens (distortions of the image.... not noticeable for large images like Ramey's face, hand, hat, etc. but noticeable for small images). The film does not react perfectly: the creation of a bright spot on the film (print) is a result of high light exposure. Where the light is very bright the complete development of a film grain (a tiny region... 10's of microns in size...) is 100% or nearly so. However, where there is less brightness (more darkness!) or where there are are brightness gradients in the image (regions between full light and full dark) the development of a grain is less probable. That's the problem. Development of film grains is a probability problem! The edges of the letter I, as focused on the film plane are actually regions of brigthtness gradient, and in such regions the probability of development of film grains is less than 100%. The result is that the letter I will have jagged edges... in some places the film grains will develop correctly and in some places they won't. If the focused image of the letter I is so narrow as to be a few tens of microns wide, that is, if the width is comparable to the size of a film grain, then the developed image may have "bumps" along the vertical sides, that is, dark areas which should be white. This is the "bleeding" of the darkness of the letter I into the white surround, Conversely, there can be "bleeding" of the bright surround into the dark area (because of the aberrations in the lens or slightly imperfect focus or veiling glare - a phenomenon that became important in the Trent photo analysis). The net result is that the nice vertical straight line I becomes a distorted line. And the relative amount of distortion increases the smaller the image gets. In the Ramey case one gets a large amount of distortion in some cases, making it very difficult or impossible to determine the exact shape of the image of a letter. With more complex letters it requires even more accuracy to determine. Consider the difference between e and c. Suppose the letter is an e. If, in recording on film the image of the letter e, the image of the horizontal line is slightly blurred so that brightness of the surrounding white are "bleeds" into the thin area that should be the horizontal line, then some film grains that should not develop actually do develop making the center of the e quite bright. The analyst might look at the image and conclude it was the letter c, not e. Conversely, suppose the letter is a c. If in creating the letter c (assumed to be a small image, not much bigger than grain size) a film grain at the middle of the c did not develop (i.e. did not correctly represent the brightness at the center of the c) then one might think the letter was an e. This is the "salt and pepper" noise problem one has when analyzing tiny images (such as the face on Mars!). (Here "tiny" means not much larger than the pixel or film grain size.) One has speckles of "salt" (bright spots) where one should have dark and conversely, speckles of dark "pepper" spots where one should have white. Under these circumstances the best one can do is assign probabilities to the accuracy of detection. If the letter appears to be a c then one can assign a probability that is based on the physical size of the letter (image) compared to the grain (or pixel) size. If the size of the letter is only a few times grain size the probability of c vs e might be only 50%. If the letter is ten times grain size the probability might be 90%. As the size increases the probability of accuracy increases. For small letters one could say that if the image _looks_ clearly like that of a particular letter, then the probability is high that it is that letter. In any other situation one should assign a probability even though this is an "unscientific" probability. Using only letter images that seem to have high probability of being what they look like is a good way to proceed with the Ramey memo. Use only the highest confidence letters at first and then work "downward" to the least confident. OF course, since this is a language there are other constraints besides ust identifying the letters. The letters come in groups... words... so if a few letters of a word are readable then one can search the entire vocabulary for words with those few readable letters and that many letters in the word. The Ramey memo has another problem, which is distortion of the letters, not b the camera optics, but resulting from curvature of the paper as Ramey was holding it. (Why didn't he hold it upstraight and flat... that would have made things so much easier!!) That means one must try to identify letters from images that are distorted in shape as well as being photographically noisy. Hence it is not a pretty sight, that of a diligent investigator sweating over the interpretation of the Ramey memo, working his/her way through the film-grainy, blobby, bent images of letters that, when properly interpreted, could change our view of humanity's place in the universe. My complements to those who have slugged it out with the data, those who have dug in the muddy ditches of noisy photography and who, it appears, have struck some gold!
[ Next Message | Previous Message | This Day's Messages ]
This Month's Index |
UFO UpDates - Toronto - Operated by Errol Bruce-Knapp