Chapter 2

Analysis of UFO Photographic Evidence

William K. Hartmann

1. Introduction
2. Selection of Cases
3. Sources of Data
4. Hidden Data
5. Quality of UFO Photographic Data
6. Natural Phenomena Photographed as UFOs
7. Fabrications
8. Techniques of Analysis
9. Review and Summary
10. Conclusions
Plates 1 - 12
Barra de Tijuca Plates
BACK to Contents

1. Introduction

The first reported photograph of a UFO after the Arnold sighting of 24 June 1947, was made on 4 July 1947 in Seattle, Washington. (Ruppelt, 1956, p.32) The object was identified as a weather balloon. This first photograph is typical of the photographic evidence that has accrued since: It accompanied a "wave" of reports and was inconclusive in establishing the existence of any extraordinary aircraft.

Although photographic evidence, in contrast to verbal testimony, might be considered "hard" data, experience has indicated that one cannot assume that a photograph of an airborne disk is more credible than a verbal report. Even if it were true that cameras never lie, photographers sometimes do. A photograph may be more interesting than a verbal account; indeed, if we knew that "flying saucers" existed, the best documented photographs would be extremely valuable in establishing their properties. But in the absence of proof of the existence of such aircraft, we are concerned at this stage with the credibility of reports.

The most convincing case of photographic evidence would involve not only multiple photographs but multiple photographers, unrelated and unknown to each other, a considerable distance apart (preferably tens of miles), whose photographs demonstrably show the same UFO. No such case is known to the Colorado project.

The Colorado project studies of UFO photographs are based on this approach. The question that is central to the study is: does the report have any probative value in establishing the existence of flying saucers? A question definitely secondary in importance (and conducive to unproductive arguments) is: What is the final explanation of each photograph?


That is to say, our principal task is to examine UFO photographic evidence that is alleged to indicate the existence of "flying saucers," and make a judgment as to whether the evidence supports this assertion. Photographic evidence is peculiarly open to the contention that one must establish what is shown, before one can say that it is not a "flying saucer." This argument is invalid. It is not necessary to prove that an object is an orange before establishing that it is not a mushroom. Exhaustive attempts to establish the identity of each object or image recorded were therefore not made. Yet possible interpretations were suggested in many cases where it was concluded (for one reason or another) that there was no evidence of an unusual phenomenon.


2. Selection of Cases

Time and funds did not permit exhaustive investigation of all interesting cases. About 90% of the cases could be assigned second or third priority upon inspection or brief study. Such a priority rating was based on a judgment that the case had little potential value in establishing the existence of "flying saucers." The remaining 10% of the cases were of first priority and required intensive study, some as much as a month of full-time effort. A "residual" of about 2% to 5% of all cases remained unexplained after this process. It is such a residual that is the core of the UFO problem (both in photographic cases and more generally).

The O'Brien committee (see Appendix A) suggested that the proposed university study of UFOs give emphasis to current reports. However, certain older, "classic" cases from the last two decades contain the most significant photographic evidence. Neglect of them would justifiably be open to criticism. Hence, the present photographic study includes both new cases and independent reevaluations of older cases.



3. Sources of Data


Material on a number of older cases was obtained from the Aerial Phenomena Office (Project Blue Book) at Wright-Patterson Air Force Base, Ohio. In many cases, these files were not sufficiently organized or complete to permit an intelligent evaluation of the report. Further investigation was carried out in these instances.


Cordial relations were maintained with APRO, and through the kind assistance of Mr. and Mrs. J. Lorenzen much first-or second-generation photographic material was made available.


Contacts for the exchange of information on photographic cases were established with NICAP in the spring of 1967, and files on a number of cases were made available to us at that time.


The help of Dr. McDonald, Institute for Atmospheric Physics, University of Arizona, who conducted a study of UFO phenomena concurrently with this study, was invaluable in bringing a number of cases to our attention.


Many individuals submitted reports directly to us and other recent cases were investigated by our field teams. Certain news organizations, in particular BBC, Time-Life, Inc., and United Press International were very helpful in obtaining material. Dr. R.M.L. Baker, Computer Sciences, Inc., kindly made available to us his files on the Great Falls, Tremonton, and Vandenberg AFB motion pictures. Dr. J. Allen Hynek, of Northwestern University also rendered valued assistance in providing materials for analysis.


4. Hidden Data

The problem of hidden data is characteristic of the study of UFO phenomena. Only about 12% of those persons who have seen flying objects they cannot identify actually report the sighting (Section III, chapter 7). The indication that we are aware of only a small fraction of all sightings of


UFOs and the experience of investigators in uncovering photographs suggest that we have considerably less than half the photographs considered by their owners to show UFOs. Of the photographs that may have a bearing on the existence of extraordinary aircraft we probably have a larger fraction, since they are more interesting to their owners. The distinction is that an UFO photo may show just a point source of light, or an amorphous blob, while an alleged "flying saucer" photo must exhibit some detail. But even in these cases, the fraction may well be less than half.

Reasons for the existence of hidden data include:

  1. apathy on the part of the photographer,

  2. ignorance of what to do with the photographs,

  3. fear of ridicule,

  4. fear of becoming involved with authorities in situations involving security or military restrictions (e.g. Ft. Belvoir case),

  5. fear of restrictions in JANAP-146.

It is also possible that data, generated by various technical recording equipment, such as all-sky auroral cameras, or the Prairie Network are another "hidden" source (Section VI, Chapter 9).

Finally, there is another class of "hidden data": sightings supposed to have occurred on various military bases but allegedly suppressed by military or intelligence authorities. We have heard many allegations of such cases. Usually they were not detailed enough to be fruitful, and in only one case was it possible for us, even with the cooperation of the Air Force, to locate any alleged photographs of UFOs. Such allegations of suppression may typically arise as a result of incidents like that described in Case 51 . In this instance a bright UFO was recorded by several tracking cameras at Vandenberg AFB. The UFO was described as "streaking up past" a rocket during a launch. Project investigators recovered the films in question without difficulty. Study of them conclusively identified the UFO as the planet Venus. Meanwhile, however, the story had reached the rumor stage, and it is likely that belief that an UFO had paced a rocket was widespread as a result.



5. Quality of UFO Photographic Data

The statistical properties or the quantity of photographic data are less important than the content of a single case that might strongly indicate the existence of a hitherto unrecognized phenomenon. Nonetheless, it is a part of the problem that most of the data are of very low quality. A glance through typical UFO periodicals and books illustrates this. Many of the photographs are blurred, usually due to poor focus. Many are badly processed or light-struck. Many, usually because they are fabrications made with small models too close to the camera, show, against sharp backgrounds, objects that are hopelessly out of focus. Many photographs do not give the subjective impression of a metallic or luminous entity flying through the air at some moderate distance from the observers.

More specifically a large part of the data is inappropriate for analysis. Night-time photographs that show either point sources or amorphous blobs with no background or foreground fall in this category. Daytime photographs of objects of very small angular size are also of little value. A large number of reports consist of only one photograph, and single photographs are of much less photogrammetric value than sets.

Damage to negatives frequently renders them valueless for investigative purposes. An investigator visiting one witness found a baby playing on the floor with the negatives. (McMinnville, Case 46) A crucial spot on another set of negatives was burned out by a dropped match, assertedly by accident. (North Eastern, Case 53) Loss of original negatives or prints is reported, as in Santa Ana (Case 52).

Accurate descriptive testimony, even in photographic cases is also difficult to obtain. For example, a witness described an UFO as "half as large as the moon"; his photograph and sketch show a disk having an angular diameter of about 15°.



6. Natural Phenomena Photographed as UFOs

A number of natural phenomena, well known in various branches of the scientific community,but little known to the general public, have been reported as UFOs. Three classes of these are meteorological, astronomical, and photographic.

Plate 1 shows an excellent example of a lenticular cloud. These thin clouds are usually related to irregularities in ground elevation (hence classified as "orographic" clouds), and sometimes appear stacked, one above the other, like a pile of saucers. A number have appeared in UFO reports.

Plate 2 illustrates a sub-sun, produced by reflection of the sun off a laminar arrangement of flat ice crystals (Minnaert, 1954, p. 203). The Gulfstream aircraft case is tentatively attributed to a sub-sun (see Case 54).

Plate 3 is a time exposure of the moon, showing trailing due to the earth's rotation. The explanation of such a photograph of the moon is obvious to anyone familiar with astronomical photographs. Yet a similar picture showing the trails of the moon and Venus was widely printed in newspapers across the country in March 1966. The trails were described as two UFOs.

Although aurora displays can produce colored, fast-moving arcs of light of various shapes and brightnesses, it does not appear that auroras are involved in a substantial number of UFO reports. No UFO photographs were attributed to auroras in this study.

A number of purely photographic effects can result in UFO-like images. Two classes are very common. The first is film damage. Creases or unusual pressure produce dark images on negatives and bright spots on prints made from them. Chemical damage during development can produce either bright or dark spots on negatives or prints. The second class is internal reflections, or lens flares produced by unwanted light paths through the camera optics. Many widely circulated UFO photographs are unquestionably the result of lens flares. Symmetry about a line connecting the flare to a bright light source in the photograph is usually the clue to identification of a lens flare photograph.


Plates 4 and 5 show examples of reported "UFOs" identified as film defects, and Plate 6 shows an example of a lens flare (see also Menzel and Boyd, 1963).

Manmade objects such as balloons and rocket exhaust trails especially illuminated by a low sun during twilight have also produced many UFO reports (N.M. aircraft Case 55). A number of photographs of bright, nearly stationary point sources in a day light or twilight sky may be balloons.


7. Fabrications

Fabrications represent a delicate problem. Nowhere in the discussion of photographic cases have I conclusively labeled one as a hoax, although I have shown that this hypothesis is entirely satisfactory in a number of cases.

Hoaxes are not new in UFO investigations. The Maury Island (Wash.) incident of 1947 has been called "the first, possibly the second-best, and the dirtiest hoax in UFO history." (Ruppelt, 1956). Photographs allegedly taken by one of the witnesses to the incident had been "misplaced," he said. Eventually, he, a companion, and an "investigator" hired by a magazine publisher admitted that the incident was a fabrication. Before the case was closed, much money and time had been spent, and two Air Force investigating officers had been killed when their Air Force B-25 crashed during the inquiry into the "sighting." According to Ruppelt, the federal government considered prosecuting the hoaxers, but later abandoned the idea.

Often a photograph apparently fabricated to amuse friends results in a full-blown UFO report. The friends take the photograph seriously and tell others. Eventually a local newspaper prints both picture and story. From there it may be distributed nationally by the press wire services, or one of the private UFO investigating organizations such as APRO or NICAP. In view of the demonstrable avocational interest of some persons, especially young persons, in producing "flying saucer photos," one must be especially wary of any alleged UFO photo that could have been easily fabricated under the circumstances.

Fabrications may be thought of in two broad categories: "physical," of a real object, which is then alleged to be an UFO; or "optical,"


the producing by optical and other means of an image falsely alleged to be a real physical entity at the scene. Retouched negatives, double exposures, and superimposed images are examples of the latter. Generally, physical fabrications meet tests of consistence in lighting and shadow but fail tests of size or distance. Most commonly, photographs of models are out of focus, or have inconsistent focus between the "UFO" and other objects at its alleged distance. Optical fabrications, on the other hand, may show inconsistencies in lighting between background and UFO details, or in the case of montages, image flaws.

Plate 7 is an example of the simplest and most common type of physical fabrication - a disk-shaped model thrown into the air by hand. Plates 8 and 9 are examples of more complex fabrications a model suspended from a string and a night-time photograph of a hand-held model illuminated by flashlight. These three photographs were made by the writer. Plates 8 and 9 were made for comparison with the Santa Ana and North East UFO photographs (Cases 52 and 53). Plates 10, 11, and 12 are examples of optical fabrications made by the writer.


8. Techniques of Analysis

Photographic evidence acquires probative value only when known natural phenomena can be ruled out and it can be shown that a fabrication was not easy or convenient.

Early in the study, it was decided not to select or analyze each case by a predetermined routine. Rather, cases were studied in terms of their individual characteristics. Diagnostic characteristics included such properties as

  1. potential stereoscopy,

  2. reports by multiple visual witnesses,

  3. cloud motions,

  4. use of haze to define distance,

  5. accurate altitude and azimuth data,

  6. structure and shape of object,

  7. geometry of motion, and

  8. geometry of lighting and shadows.

Initial selection of cases to be studied was also influenced by the degree to which other students of UFO phenomena regarded them as significant.

In the course of the investigation, analysis of the foregoing characteristics of UFO photographs resulted in our developing a set of protocols useful in the assigning priorities to UFO photographs


for study. These results are described in section 10 of this chapter.

The cases selected for investigation were analyzed as completely as possible. The techniques are demonstrated in the case reports themselves (Part IV, Chapter 3).


9. Review and Summary

The project gathered information on 35 photographic cases that occurred in 1966-68. These may be assumed to be a more or less representative cross-section of photographic cases. Of this 35-case current cross-section only two, Calgary and North Pacific (Cases 57 and 56), were initially selected as first priority cases. On investigation, neither case yielded data deemed to be of probative value. Second priority cases among the 1966-68 group were Camarillo (identified probably as airborne debris), Gulfstream Aircraft (sub-sun), and Sonora (airborne debris). Many of the remaining 1966-68 cases of lower priority had low strangeness or insufficient data for analysis.

The final disposition of the 35 cases is summarized in Table 1. The figures are thought to be representative of UFO photographic cases. That is, roughly one quarter are fabrications, one quarter are misidentifications, a quarter have such low information content as to be unfit for analysis, another quarter are clearly recorded but lack sufficient data for analysis. The residual cases that are genuinely puzzling constitute at most a very small percentage.

In addition to these current cases, 18 older reports, including some by advocates of the existence of "flying saucers," were also studied.

Of the 35 cases only those in which the nature of the evidence or the credentials of the witness were judged to have the highest a priori probability of producing evidence for an unknown phenomenon were assigned first priority for study. Table 2 shows the classifications finally assigned to these first priority cases. Of them some 60% were found to be identifiable or to lack probative value. Two cases (continued on p. 119)



Classification of 35 Current Photographic Cases

Evidence for probable fabrication 9
Misidentified natural or man-made phenomena 7
Insufficient data for analysis (night-time shots, point sources, amorphous blobs, etc.) 12
Inconclusive data (unidentified unusual objects shown, but little or no analysis possible; possible fabrications) 7
Unidentified after analysis (real objects with high strangeness) 0




Classification of 11 First-Priority Cases

Inconsistencies between testimony and photos, internal inconsistencies in photos, or evidence for fabrication
Barra da Tijuca
North Eastern
North Pacific
Santa Ana
Identified natural or man-made phenomena
Fort Belvoir
Vandenburg AFB
Not amenable to analysis Calgary
Unidentified after analysis (indication of real objects with high strangeness), conceivable but unlikely misidentification of birds, aircraft, etc. Great Falls
Clearly either a fabrication or an extraordinary object ("flying saucer") McMinnville


survived analysis: Great Falls (motion pictures of two bright light sources difficult to reconcile with known aircraft) and McMinnville (two photographs of a saucer-shaped craft).

Since the selection of older, "classic" cases was limited, it is probable that the "residual" of unexplained photographic cases could be increased well beyond these three cases if there were additional research. Whether or not anything of probative value would be found is a matter of speculation.



10. Conclusions

Our experience also leads us to conclude that UFO photographic cases can best be selected for study and analyzed on the basis of the following criteria:

  1. Subjective evaluation: Do various photographic factors (focus, clarity, sharpness, contrast) and the testimony combine to make the case appear credible? Does it have potential in providing probative evidence for the reality of an unusual phenomenon?

  2. Known phenomena: Is any known phenomenon rationally acceptable as an explanation of the observation? Phenomena considered must be based on a wide experience with meteorological, astronomical, optical, and photographic effects. Can the report be a case of mistaken interpretation?

  3. Fabrications: Can the case be accepted as having been made in good faith? Are there any signs of tampering with the negative? (Are the negatives or original prints available?) Do the negatives represent a continuous sequence? Are focus, sharpness and other characteristics quantitatively in accord with the alleged sightings? Are light and shadows internally consistent on each photo?


  1. Consistency with testimony: In addition to the internal evidence of the photographs themselves, are the photographs consistent with the witness testimony? Is lighting consistent with alleged time and direction of sighting? Are time intervals between photos consistent with testimony?

  2. Physical and geometric tests: What peculiar characteristics suggest tests? Is the object in front of or behind any landscape features? Is contrast and focus consistent with alleged distance? What can be learned from motions and time intervals? Can the flight path be estimated from the sequence of positions and angular sizes?

The Colorado study of UFO photographic evidence failed to disclose conclusive evidence of the existence of "flying saucers." Nor did it, of course, establish that such objects do not exist. I believe that it is significant, however, that a number of the most widely heralded "classic" cases were either identified or were shown to be of little probative value in the present study. This finding suggests that much of the case for the reality of "flying saucers" has been built on very inadequate research into widely publicized reports. Some examples of such cases, the reality of which has been rejected after intensive study by the project, are summarized briefly below:

Barra da Tijuca, Brazil, (Case 48): A magazine photographer and a reporter allegedly saw and made five photographs of a large disk that passed overhead. The photographic sequence shows the disk approaching (edge on) in the distance, and passing by in a credible series. A report on the case by O.T. Fontes, of Brazil, (APRO, 1961) "pronounce(s) them authentic" and purports to establish


their authenticity with "top-secret documents" from Brazilian Air Force files kept since 1951. The documents purport to demonstrate "the absolute impossibility of a hoax." Study of photographs enlarged from the APRO copies shows that the disk in the fourth photograph (Plate 30) clearly illuminated from the left, with bold shadows, hut a palm tree as well as other confused foliage on the hillside below appear to be illuminated from the right. The discrepancy was first pointed out by Menzel and Boyd (1963).

North Eastern (Case 53): Two photographs show a bright, amorphous object that reportedly swept past four boys who were photographing the moon at night. The image on the photographs is strikingly suggestive of an out-of-focus plate-like object supported by a human arm and hand photographed by time-exposure. According to the original report, (NICAP, 1965) the "arm" was an invisible gaseous discharge from the UFO. A photograph (Plate 9) that demonstrates how such an image can be fabricated was made by taping a plate to a small handle. The apparent transparency of the "gaseous discharge" was simulated by moving the arm during the time exposure. In the light of such simple reproduction of these photographs, I have concluded that this case is of no probative value.

Fort Belvoir, Va., (Case 50): Six exposures made on this Army base show a ring-shaped object being enveloped in a white, puffy cloud. The photographs were proclaimed as "First Published Photos of the Amazing Ring-Shaped UFO" (Rankow, 1967). Aides of the commanding officer at Fort Belvoir demonstrated to a project investigator that this was a vortex cloud generated by atomic bomb simulation demonstrations that were frequently carried out at the base some years ago. Positive identification was obtained.

North Pacific (Case 57): Three boys in their back yard photographed a disk that allegedly passed overhead. The object was not reported by any other witnesses. The incident was given considerable publicity and the two photographs were published by APRO. In an


interview the boys stressed that they had accurately re-enacted the event and that the time interval between the two photographs was very short, about eight seconds; however, the cloud patterns were markedly different. Separately confronted with the marked discrepancy in cloud structure between the two photographs, the boys each said they could not account for it, though they reaffirmed the story of the sighting. The photographs cannot therefore be considered as satisfactory evidence for the existence of "flying saucers."

Santa Ana, Calif., (Case 52): A traffic engineer, of good reputation, with excellent references, and with experience as a former policeman, allegedly saw and made three photographs of a metallic disk and a fourth photograph of a vortex smoke ring allegedly left by the departing disk. Interruption of radio transmissions from his vehicle, reportedly associated with the presence of the disk, was confirmed by the engineer's supervisor. The series of photographs has been widely published and widely regarded as one of the best cases. Detailed investigation revealed several serious discrepancies. For example, a study of the weather data at surrounding stations indicates that an early morning cloud cover had entirely dissipated well before the report was made, yet the fourth photograph shows a background of moderately dense, gray clouds. Other circumstances surrounding these photographs reduce further their probative value.


In the course of my study I was able to simulate effectively the first three photographs by suspending a model by a thread attached to a rod resting on the roof of a truck and photographing it (Plate 8 ). Without assuming the truth or untruth of the witness' story, this has led me to conclude that the case is of little probative value.

Vandenberg AFB, Calif., (Case 51): Tracking films from a rocket launch show a bright object apparently rushing up past the rocket just after second stage ignition. The films were first described in a textbook (Baker, 1967). The film sequence was taken very seriously because several cameras in different locations simultaneously recorded the object. Interest in the case was heightened by its resemblance to a number of apocryphal accounts of UFOs pacing rockets. The Colorado project at once obtained the films through official channels. Tracking data showed that the rocket was moving toward the horizon past the calculated position of Venus at the time.

To summarize conclusions relating to UFO photographs:

  1. About half of the photographic reports are clearly identifiable as known phenomena or can be demonstrated to contain internal geometric or other inconsistencies.

  2. About half can be ultimately classified as being inconclusive or presenting insufficient data to furnish probative evidence of an unknown phenomenon. Most single-witness cases must fall in the latter category. Most night-time photographs, point-source objects, and amorphous objects without background or foreground must be relegated to this category for lack of satisfactory quantitative tests that can be performed on them.

  3. A number of cases initially described publicly by UFO enthusiasts as representative of the strongest evidence for the reality of extraordinary aircraft were either conclusively identified as ordinary phenomena or shown to have serious internal inconsistencies.


  1. The number of identified or fraudulent cases is irrelevant to the existence or non-existence of extraordinary objects or "flying saucers."

  2. A very small fraction of potentially identifiable and interesting photographic cases remain unidentified.

Some conclusions relating to these residual photographic cases are:

  1. None of them conclusively establishes the existence of "flying saucers," or any extraordinary aircraft, or hitherto unknown phenomenon. For any of these cases, no matter how strange or intriguing, it is always possible to "explain" the observations, either by hypothesizing some extraordinary circumstance or by alleging a hoax. That is to say, none of the residual photographic cases investigated here is compelling enough to be conclusive on its own.

  2. Some of the cases are sufficiently explicit that the choice is limited to the existence of an extraordinary aircraft or to a hoax.

  3. The residual group of unidentifieds is not inconsistent with the hypothesis that unknown and extraordinary aircraft have penetrated the airspace of the United States, but none yields sufficient evidence to establish this hypothesis.

In summary, about 10% of the photographic cases can initially be selected as "first priority" cases, i.e. interesting and detailed enough to investigate. After investigation, there remains a small residual, of the order of 2% of all cases, that appears to represent well recorded but unidentified or unidentifiable objects that are airborne - i.e. UFOs. Yet there is insufficient evidence to assert that any one of these represents an unusual or extraordinary phenomenon. We find no conclusive evidence of unidentified aircraft or "flying saucers." The photographic data has been poorly presented in the past, and the frequency of hypothetical "flying saucers" appears much smaller than has been popularly assumed; it may be zero. The present data are compatible with, but do not establish either the hypothesis



  1. the entire UFO phenomenon is a product of misidentification, poor reporting, and fabrication or that

  2. a very small part of the UFO phenomenon involves extraordinary events.




    Baker, R. M., Jr. and Maud W. Makemson, An Introduction to Astrodynamics, N. Y.: Academic Press, 1967.

    Menzei, D. H. and Lyle G. Boyd, The World of Flying Saucers, Garden City, N. Y.: Doubleday, 1963.

    Rankow, Ralph. "The Ring-Shaped UFO," Flying Saucers, No. 4, Fall, (1967).

    Ruppelt, F. J., The Report on Unidentified Flying Objects, Garden City, N. Y.: Doubleday, 1956.