Frequently Asked Questions|
Computer Reconstruction of the JFK Assassination
On November 20, 2003, ABC Television News broadcast the prime-time special Peter Jennings Reporting: The Kennedy Assassination - Beyond Conspiracy, which featured selected sequences from my computer animation project, Secrets of a Homicide: JFK Assassination. A re-edited version of the ABC special is periodically re-broadcast on the History Channel. A 2005 Discovery Channel program, Beyond the Magic Bullet also included sequences from the computer animation project.
While the television specials touched on some of the methods used to create the forensic animation, the amount of time devoted to exploring the geometric principles behind my work was necessarily limited. Unfortunately, these time constraints left many viewers with unanswered questions about my project.
The following is an additional effort to provide factual information about my work and address some of the allegations raised by uninformed skeptics.
Dale K. Myers
When will Secrets of a Homicide: JFK Assassination be available on DVD?
The DVD release of Secrets of a Homicide: JFK Assassination is planned for the near future, though no firm date has been set. The DVD will contain background information and animated sequences not seen in the ABC News special. Details will be posted here when available. CLICK HERE to see a preview of the forthcoming DVD. (Quicktime 7 or higher and a high-speed Internet connection is required to view this video. To get the free Quicktime plugin, CLICK HERE.
Meanwhile, the ABC News special Peter Jennings Reporting: The Kennedy Assassination - Beyond Conspiracy is currently available on DVD from abcnews.com.
Weren't you originally a conspiracy theorist?
Yes. I became interested in the Kennedy assassination in 1975 while working at a radio station in northern Michigan. After studying numerous books and periodicals on the subject, I began searching out original documents and photographs from the National Archives, Texas State Archives, and other institutions nationwide. By the early 1980's I had secured thousands of documents through the Freedom of Information and Privacy Act from the Federal Bureau of Investigation. I eventually conducted a series of interviews with key figures in the assassination story, with a particular emphasis on eyewitnesses and law enforcement officers connected with the murder of Dallas police officer J.D. Tippit. I later turned some of those interviews into a 1983 radio documentary which was honored by the Associated Press. Throughout a ten year career in radio, I was an outspoken conspiracy theorist, lecturing at local universities and community colleges and appearing on local radio talk-shows.
In 1985, I left radio and began working as a writer/producer at CBS/FOX Video. That fall, I authored The Detroit News' 25th anniversary coverage of the JFK slaying. In 1989, I began freelancing as a computer animator. Four years later, I served as an on-camera expert and technical consultant for the critically acclaimed 1993 BBC/Frontline documentary, Who Was Lee Harvey Oswald? That same year, I began working on two concurrent projects: a computer recreation of the JFK assassination and a book on the murder of J.D. Tippit. My extensive work in both areas changed my mind about the case, convincing me that Oswald acted alone.
I have made numerous trips to Dallas, Texas over the last twenty years and continue to visit the region and conduct ongoing research.
What are your qualifications for conducting this forensic study?
First and foremost, I have an extensive knowledge of the Kennedy assassination. My thirty-one year career in radio and television has also provided me the skills to produce high quality visuals with a particular attention to details.
In 1989, I taught myself computer animation using Sculpt 4D, and an Amiga 500 computer, whose specifications boasted 512KB of RAM and a single floppy drive. The following year, I purchased an Amiga 2000 which boasted a 286 motherboard, 16MB of RAM, a 500MB hard drive, and a copy of the first version of NewTek's LightWave 3D animation software. I organized a national Special Interest Group (SIG) of LightWave users with fellow animator and software developer Jon Tindall, produced a monthly newsletter, and was influential in the development of LightWave 3D.
I've been blessed with numerous broadcast and animation awards including 3 EMMY ® awards presented in 1994 by the Michigan Chapter of the National Academy of Television Arts and Sciences for my computer creation, Robo Jr. The six-minute short subject - which I produced, directed and animated - was one of the first entirely computer generated 3D cartoons and was featured at SIGGRAPH's annual computer convention in Los Angeles. Additional honors include the Gold Camera Award from the 27th annual U.S. International Film and Video Festival.
A seasoned writer and public speaker, I've served as spokesman on many subjects ranging from video production techniques to the art of computer animation. My most recent book, Computer Animation: Expert Advice on Breaking into the Business, distills experiences from industry leaders into a definitive street-smart guide.
Over the last 17 years, I've worked as lead animator on everything from M&M commercials for Will Vinton Studios to automobile commercials for Ford Motor Company. My knowledge of computer animation and LightWave 3D software in particular, is extensive.
Has your recreation been independently examined by forensic animation experts?
Yes. ABC News hired independent experts from Z-Axis Corporation, one of the nation's leading forensic animation companies, to assess my work and submit a report. Established in 1983, Z-Axis Corporation had been involved in the business of producing computer generated animations of events, processes and concepts for major litigations in the United States and Europe. They participated in most of major air crash litigations in the U.S. over the past 15 years including the crash of Delta Flight 191 in Dallas, the crash of USAir 427 in Pittsburgh, the crash of American 965 in Cali, Colombia and the crash of Korean Air 801 in Guam. They also performed work for the prosecution in the Oklahoma City bombing trials of Timothy McVeigh and Terry Nichols.
On October 9, 2003, Z-Axis Corporation CEO and co-founder Alan Treibitz and Senior Producer Gary Freed met with me at my studio near Detroit, Michigan. Over a six hour period, I answered questions and shared all of the materials and methods that went into the making of the assassination recreation. Treibitz and Freed reported that I did "an excellent job" of matching the Zapruder film which allowed the assassination sequence to be viewed "from any point of view with absolute geometric integrity." Regarding the use of error cones to describe potential inaccuracies in establishing the single bullet trajectory, Treibitz and Freed reported that "Mr. Myers' techniques for establishing this cone were well thought out and accurately created." Treibitz and Freed concluded, "Mr. Myers has taken a comprehensive and reasoned approach to animating this event and has successfully incorporated many diverse visual records into a unified and consistent recreation. We believe that the thoroughness and detail incorporated into his work is well beyond that required to present a fair and accurate depiction."
Isn't your computer animation work, like all computer science, a product of its human creator and therefore subject to its creator's biases - whether conscious or unconscious? In short, aren't your conclusions nothing more than an example of the old axiom, Garbage In - Garbage Out?
No. This is one of the most common arguments offered by critics of my work. In an effort to call my integrity into question, some critics have charged that my own biases about the case (either intentional or unintentional) invalidate my conclusion. In critical thinking circles this is known as an abusive ad hominem argument (or poisoning the well), which is designed to convince an audience to stop listening to a particular point of view by hurling insults at the opponent. Often such dismissals appear prior to the opponent's presentation and include some innuendo about ulterior motives. Think of it as a pre-emptive strike. One critic took this tact with my work nearly one month before the ABC-TV special even aired, claiming that "small errors" in basic information, "whether unconsciously or deliberately introduced by the creator's bias," led to invalid conclusions. Of course, there is no basis for the allegation.
While some critics have been subtle in raising questions about my integrity, others have been more blatant, claiming I resorted to "junk science and chicanery" to make my points and that the results I achieved were "crudely fraudulent." One critic claimed my "entire model is a disingenuous facade of scientific inquiry." Still others called me "devious" and "pathetically dishonest," and my work "a flagrant and transparent piece of propaganda; obviously intended for those who know little or nothing about the case." As is the case in most fallacy arguments of this type, none of these critics has offered one ounce of believable evidence to support the charge that my integrity is, in fact, suspect.
Ultimately, the fatal flaw in the "Garbage In - Garbage Out" (GI-GO) argument is that while it may be true in some cases it is not true in all cases. This is known to critical thinkers as a form of fallacy of composition. It refers to arguments which characterize an opponent by lumping him/her into an all encompassing category. In this case, critics invoke the rather broad GI-GO axiom to dismiss my computer recreation. While the GI-GO axiom has some validity, its application to my work hinges on a key assumption - i.e., that my conclusions are, in fact, derived from faulty data, or worse yet, influenced by my own biases. Despite the charge, critics have yet to offer anything to support either assertion.
In order to draw a valid conclusion, doesn't a recreation of this kind require an extremely high level of precision; one that would be virtually impossible to obtain given all of the unknowns in the Kennedy case?
No. There is a common misconception that there are too many unknowns to be able to accurately recreate the assassination and that the cumulative effect of these unknowns invalidates any conclusions. One critic wrote, "I think the whole notion of using 'animation' to prove or disprove anything is a hopelessly inconclusive venture, unless everything is 100% correct, and how can anyone know that with all the variables involved?" Another critic, who claimed to have a background in computer engineering, wrote that in order to determine a bullet's trajectory a person would need to know "the exact dimensions" of Dealey Plaza "within several decimal points," including the height of the 6th floor window, the location and angle of the rifle barrel, the location of the Presidential limousine, its occupants and their wounds, and even the exact air pressure of the limousine tires.
This is nonsense, of course. Although the models used in the assassination recreation are highly accurate, the kind of precision suggested by this critic is unnecessary because it would not significantly alter the results. In other words, minor differences in position and accuracy (especially in the 1/100th of an inch range, as one critic suggests) would be negligible.
In fact, the degree of accuracy needed to deduce a valid trajectory is more than adequate in my recreation. Contrary to claims, the height of the sixth floor window is known and measurable. The location of the limousine and its occupants can be deduced from the Zapruder film and other films and photographs using the same mathematics employed in a multitude of scientific disciplines; everything from architecture to motion picture special effects. In addition, the recreation makes no assumptions about the location of the shooter. The idea is that if you know the location of the limousine and its occupants, you can determine the trajectory of the bullet that passed through their bodies. In this case, the results show that both men were struck by a single bullet. Of course there is a certain level of imprecision inherent in any reconstruction of this type, which is the reason that "error cones" are used to illustrate the potential margin of error in placing the trajectory. The results show that even when we take into account the margin of error, the answer is the same.
Isn't it true that you used an unreliable survey map of Dealey Plaza, created by Robert H. West for the Warren Commission, to produce your model of the crime scene?
No. One critic complained that I took "data such as surveyor's measurements and elevations of Dealey Plaza" from surveyor Robert H. West's plat of Dealey Plaza to build my model. "Problem is, (BIG problem)," according to this critic, "the surveyed measurements were changed by junior Senator Arlen Spector to align his awkward 'single bullet' or 'magic bullet' theory. The man who did the survey testified under oath to Congress in 1976 that all his measurements were changed. Some rather dramatically...So Dale does in fact prove that the Warren Commission's altered data works within its own boundaries, but his research sadly gets us no closer to what actually happened."
This criticism stems from Assassination Science, a conspiracy book which claims that "the survey measurements made for the Warren Commission by Robert H. West, Dallas County Surveyor, were altered" by Arlen Spector "to insure his single bullet theory would not be contradicted." [p.250] The book claims that various versions of West's survey map show discrepancies in the placement of X's representing rifle shots, the number and length of lane markers, and the alteration of a data block containing figures related to several re-enactments. None of these purported alterations, even if true, have any affect on my work since my only interest lay in the contours and elevations of the streets as marked on West's survey maps. West himself testified on February 13, 1969, during the Clay Shaw trial, that the "curves and contours" of all the streets were complete and accurate. (West never testified "under oath to Congress in 1976 that all his measurements were changed," as claimed.)
More importantly, my computer model of Dealey Plaza did not rest solely on the West survey. I used a number of resources to build my model of the crime scene: a 1939-40 dimension and grading plan prepared for the Dallas Park Board by landscape architects Hare & Hare; a survey map prepared by Robert West for the Warren Commission in 1964; a survey map prepared by Drommer & Associates for the House Select Committee on Assassinations in 1978; hundreds of personal photographs and measurements of modern-day Dealey Plaza; and computer software capable of constructing three dimensional models from 2D photographs. My ability to align the Zapruder film with the model created from these resources verifies their authenticity and accuracy.
Isn't it true that you incorrectly modeled the presidential limousine, positioning Connally's jump seat six inches from the inside of the door rather than the actual distance of 2.5 inches?
No. One critic claimed that I "used the incorrect limo measurement of a 6 inches clearance between JBC jump seat and door. The actual measurement was 2.5 inches. So whatever trajectory [Myers] thought he proved was not what 'a single bullet' could have taken."
This particular criticism stems from a comment made during the ABC News broadcast. At one point in the program, a computer animated sequence compares a diagram of how conspiracy theorists believe Kennedy and Connally were seated in the limousine with how they actually were seated as seen in the Zapruder film. Peter Jennings notes in voiceover narration that Connally was not seated directly in front of Kennedy, as some conspiracy theorists believe, but was "six inches" to Kennedy's left. However, the six inch figure mentioned in narration did not refer to the distance between the jump seat and the inside of the limousine door, as presumed by this critic, but instead referred to the distance between the center of Kennedy and Connally's body. Kennedy was seated to the extreme right side of the limousine. Connally was turned to his right and had shifted left on the jump seat in front of Kennedy. Projecting an imaginary line forward from the center of the both men shows that the difference between their two center points is six inches. Connally's jump seat, which was about 20.5 inches wide, was correctly located 2.5 inches from the inside of the right-hand door.
Isn't it true that you failed to take into account the differences in the height of President Kennedy and Governor Connally and that this alone invalidates your single bullet alignment?
No. One critic wrote "I find it ironic that Dale Myers makes such a big deal out of conspiracy writers not taking into account the 3 inch difference in the height of the seats and then himself fails to take into account the 3-1/2 inch difference in height between the two men."
This critic fails to grasp that two separate issues are involved here. First, the fact that conspiracy writers often use diagrams to show both Kennedy and Connally seated at the same height (and hence, are not in alignment for a single bullet strike) was shown in the ABC News special to be false and misleading. Connally's jumpseat was in fact 3 inches lower than the backseat upon which Kennedy was seated. Second, and more to the point, the height of the computer models used to represent Kennedy and Connally in the computer reconstruction were aligned to the height of both men as seen in the Zapruder film. Consequently, any discrepancy between the height of the computer models and actual men they represent has no bearing on the results of the subsequent alignment. (Incidently, Kennedy stood 6 foot, 1/2 inches and Connally stood 6 foot, 2 inches - a difference of 1-1/2 inches, not 3-1/2 inches as this critic claims.) This is another case of conspiracy theorists grasping at straws in order to undermine the results of my computer reconstruction.
Isn't it true that you distorted the size and position of your models of President Kennedy and Governor Connally in order to fit your simulation to the single bullet theory?
No. One critic charged that "Myers depicts Governor Connally’s body as considerably smaller than JFK’s body" - 25% according to one of this critic's measurements, and later 15% smaller, after a second measurement - and that "by depicting Connally as a midget it allowed Myers to place Connally’s armpit, site of the bullet entry, at a point several inches further in from the side of the car, and more in line with the predetermined trajectory from the sniper’s nest through JFK." According to this critic I also distorted Kennedy's size and position, writing "In order to make the trajectory work, however, Myers had to distort Kennedy’s body shape. Kennedy suddenly had a crookneck, which curves forward and then up, like the neck of an old crone. This is not apparent in any photo of Kennedy. This Myers invention made it possible for a shot to hit Kennedy in the shoulder line and still come out his throat. When viewed from above, however, this distortion should have been obvious, with Kennedy’s head a half-foot forward of his shoulder. In the Discovery Channel program Beyond the Magic Bullet Myers’ animation was shown both from behind and above, and the two depictions of Kennedy’s shoulders and his back wound were completely at odds with one another. Clearly, Myers uses different measurements for his animated figures depending on how he wants them to be seen."
The critic also charged that "While the Zapruder film shows Kennedy’s arms inside the car, as there’s no shadow on the side of the car, Myers' animation always depicts Kennedy’s right elbow hanging over the side...Not surprising, this misrepresentation by Myers puts Kennedy more in line with the pre-determined [single bullet theory]. When JFK is put in his proper position the trajectory traces back to the Dal-Tex Building."
What does the critic conclude from all of this? "While Myers has done some valuable research...this deception regarding the single-bullet theory is inexcusable...He lures you into thinking that because he's using a computer the proportions and angles are the same from frame to frame and shot to shot when they're not...I've come across three different depictions by [Myers] of the single-bullet shot which move the President's position and the wound itself depending on what would look plausible to the viewer from that angle...It's all smoke and mirrors...[H]is animation is blatantly dishonest and demonstrably inaccurate...The bottom line: it's okay to misrepresent the evidence as long as you do it to PROTECT the government."
Hogwash. This is a common refrain from critics of my computer work. The truth, of course, is that my computer reconstruction of the Kennedy assassination is based on a single model put in motion. The mistakes this particular critic made in his analysis of my work are numerous.
First, his claim that the relative sizes of Kennedy and Connally change according to the angle at which they are presented is apparently based on the critic's measurement of the final rendered image. Performing an analysis in this manner fails to take into account photogrammetric effects as well as the size distortions produced by the computer's virtual camera. Photogrammetry describes how three-dimensional spatial relationships can be extracted from two-dimensional photographs or images. Without taking into account these relationships, accurate interpretations of two-dimensional images are impossible. In short, you cannot simply draw or overlay lines on a two-dimensional image (as this theorist has claimed) and extract three-dimensional measurements. This is a common amateur blunder. In addition, each rendered viewpoint is generated by a virtual camera whose focal length characteristics are akin to real-world cameras. For instance, a wide angle focal length in both virtual and real-world cameras will produce images in which identical-sized objects appear at different sizes depending on their relationship to the camera. In the case of my computer reconstruction, wide angle overhead-view renderings of Kennedy and Connally in the limousine will produce images in which the model of Kennedy appears slightly larger than the model of Connally if the virtual camera is positioned closer to the Kennedy model. Clearly, this was the case in the rendered images used by this critic for his "analysis".
Second, the critic's claim that I distorted Kennedy's neck in order to produce a position favorable to the single bullet theory is equally invalid. The computer model of Kennedy was matched to the position dictated by the Zapruder film - the only complete filmed record of the event. This filmed record records the three-dimensional position of Kennedy's head, neck, shoulders, upper torso, arms and hands relative to his surroundings. The computer reconstruction tracks the dimensional changes of Kennedy's body as recorded by Zapruder's camera. Those positional changes are not "inventions" created by me in order to validate the single bullet theory. Rather, the Zapruder film, and consequently my computer reconstruction based upon it, are a definitive record of what actually occurred in Dealey Plaza.
Third, the critic's charge that my computer reconstruction falsely depicts Kennedy's right elbow hanging over the right side of the car, while the Zapruder film shows Kennedy's arm inside the car, and that this "misrepresentation" puts Kennedy more in line with the single bullet theory is also invalid. The critic apparently bases his claim on the fact that the Zapruder film doesn't show Kennedy's arm casting a shadow on the car. The critic fails to consider whether a shadow from Kennedy's arm would even be visible on the limousine's surface given the quality of the film and the highly reflective nature of the limousine body. More importantly, despite the critic's claim, the position of Kennedy's torso, as determined by the shoulder-line, and the length of his upper arm, make it a certainty that Kennedy's right arm would have extended over the side of the limousine. In fact, numerous photographs of the Kennedy motorcade show this to be true. Incidently, the critic never offers any facts to support his claim that positioning Kennedy properly in the limousine would result in a trajectory that traced back to the Dal-Tex building.
Finally, the overall charge that I re-positioned Kennedy and Connally and the location of their wounds from rendered sequence-to-sequence in order to validate the single bullet theory and/or hide the truth about their actual positions during the shooting is completely false and without foundation. I find it quite entertaining that critics of my computer work are perfectly comfortable embracing the portions of my work that support their own theories (this particular critic agrees with my reconstruction of the trajectory of the fatal head shot), while rejecting those portions they disagree with. Evidently, it's okay to use the work of a "blatantly dishonest" individual as long as you can pick and choose your own truth.
Isn't it true that you admitted a key flaw in your recreation; the incorrect positioning of the president's back wound?
No. One critic has repeatedly made this false claim on number of Internet newsgroups. The charge stems from a preliminary 1995 version of my computer recreation which employed generic humanoid figures to represent Kennedy and Connally and determine the locations of their bullet wounds. When a trajectory line was connected from the entrance wound at the back of Connally's right armpit with the exit wound on the front of Kennedy's throat, and projected rearward, that line passed a bit high of the presumed location of the entrance wound on Kennedy's upper-right back. I explained that the generic nature of the humanoid figures, which were not exact matches of either man's physique, was probably responsible for this visual effect. I also pointed out that because the medical evidence shows that Kennedy had only one bullet wound in his upper-right back, and the projected trajectory nearly intersected that exact location, there could be little doubt that the discrepancy was attributable to slight inaccuracies in the generic models.
In the recreation broadcast on ABC Television in the fall of 2003, I used new human skeleton models to more accurately pinpoint the wound locations in both men and also used new high resolution human models, though still generic in physique, to "skin" the underlying skeletons. Both of these model upgrades resulted in more accurate wound placement and improved trajectory analysis which effectively demonstrated that the earlier inaccuracies were in fact due to the generic models, as I surmised.
Isn't it true that you blatantly misrepresented the shape of Kennedy's back in order to get the single bullet to work?
No. Several critics have pointed out that Kennedy's posture appears distorted in the animation. One critic wrote, "I still cannot figure out why Myers made JFK a hunchback. Was that the only way to achieve a downward trajectory projection between the back injury and the neck wound?" Another critic wrote, "The blatant misrepresentation of the shape of the back is the sort of thing which makes me dismiss everything that Myers tries to 'prove' with his animation. If he's willing to lie to his audience to make his point, then he deserves no consideration whatsoever."
This criticism stems from some modeling and animation issues that were not fully resolved at the time of the ABC Television broadcast. The issue is ultimately a cosmetic one and has nothing to do with the trajectory analysis or its conclusions, as these critics falsely suggest.
Photographs and films taken during the course of the motorcade show that the president's suit jacket had ridden or bunched up, making his shoulder line appear to be higher than it actually was. When shot, the president's elbows rose dramatically, increasing the effect. While animating the shooting sequence, the shoulders and collar of the president's computer generated "clothing" was raised off the shoulder line of the human model beneath to mimic what is seen in the film. Due to modeling constraints, the collar could not be returned to its proper position without affecting the shoulder line. To get the clothing to "look right," the model would have to be redone, a luxury I did not have time to complete given the production schedule. That's showbiz. While the position of the collar was not a perfect match with the film (in fact, it is too high in its current position), the shoulders, as defined by the "clothing," did fit better with what is seen in the Zapruder film. In the end, because of time constraints, it was decided to leave the "clothing," including the collar, in the raised position throughout the animated sequence.
Contrary to the criticism levied by my detractors, the position of the president's clothing in the recreation has nothing to do with the validity of the single bullet theory. That's because the human model representing Kennedy, which is position beneath the "clothing," and therefore hidden from view, has not been moved. Only the "clothing" has been tugged around. As explained elsewhere on this page, it is the location of the wounds on the body, not the "clothing," that is the basis for defining the trajectory path of the bullets.
Isn't it true that you had to change the position of a number of wounds in both the president and governor to get the single bullet theory to work?
No. A number of critics have claimed that I "back solved" the single bullet theory by simply predetermining the desired outcome and setting up the variables to achieve that result. As "proof" of the claim, one critic charged that I placed the throat wound "too low" on the front of the president's neck, the entrance wound "too far to the right" on the governor's back, and the exit wound "too high" on the front of the governor's chest; stating that apparently "the single bullet theory only 'works' when you fudge the data."
I believe that this, and other criticisms surrounding wound placement in the computer model, are based on the position of a series of red, cross-hair markers used during the ABC Television broadcast to show the locations of the various wounds in Kennedy and Connally. The producers of the program decided that the figures should be fully clothed when the position of each of the wounds was located and marked for the audience. In order for the markers to be visible, they had to be placed on the outside of the clothing at the point where the trajectory path intersected the outside surface. Otherwise, they would be hidden by the clothing. Therefore, because of the angles involved, it was necessary to slightly offset the markers from their actual wound locations in order to maintain their visibility. This was true in all cases. Regardless of the position of the markers, however, the wounds themselves remain fixed at their proper locations beneath the clothing. Remember, it is the wounds, not the markers that serve as the basis for all trajectory projections.
The charge that I "back solved" the results of the recreation has no basis in fact and is nothing more than a poor attempt to undermine my credibility.
Isn't it true that you refuse to release information about the angles you used to plot the single bullet trajectory because you would have to admit that you repositioned Kennedy and Connally to get the trajectory to work?
No. One critic has repeatedly claimed that I continue to "stonewall the question of what angles [Myers] used" to plot the single bullet trajectory. "Of course, [Myers] has to," this critic charged, "because the real numbers are easily calculated and well known. He would then have to admit that he repositioned the victims, in order to make his viewers think that shot came from Oswald."
This accusation is completely false. The angles pertaining to the positions of Kennedy and Connally at Zapruder frame 223-224, as well as the angles of the single bullet trajectory have always been available on this website. (See: Conclusion #2) The margin of error in deducing an accurate trajectory using this method has also always been available on this website. (See: Potential Errors & Trajectory Cones)
Other critics have begun repeating this false accusation adding their own brand of distortions. One such critic wrote: "Myers has refused to release his source data and admits that he made assumptions to correct the missing data to make the images fit other data."
Nonsense. My original source materials are readily available to the general public through the National Archives and other public institutions. There is nothing magic about obtaining these source materials. Every one of the source materials I used in the reconstruction of the assassination are described within this website. I suspect that what this critic meant to say is that I haven't released the original computer data files that describe the objects and scenes of my reconstruction in mathematical terms. This is true.
There are two primary reasons for this: (1.) The computer data files were created over a ten year period at great personal expense and consequently are a valued intellectual property asset; and (2.) to ensure the integrity of the original data, I have not permitted the files to leave my possession.
In order to address legitimate questions surrounding the integrity and validity of my work, the original data files were personally examined in their original form by qualified forensic animation experts from the nation's leading reconstruction firm, Z-Axis Corporation. I believe their report puts to rest any questions regarding my methods and conclusions. [Click HERE to read the complete Z-Axis report.]
The critic's other claim that I admitted making "assumptions to correct the missing data to make the images fit other data" is another completely false and unsubstantiated charge designed to sway the uninformed. My work, methods, and procedures are detailed throughout this website. Nothing this critic claims has any basis in reality as any sensible reader can see.
Your single bullet trajectory seems to pass through the center of JFK's tie, even though the tie only had a slit-like notch on one side. Doesn't this invalidate your trajectory?
No. The diameter of the red line representing the single bullet trajectory path was sized for the ABC News special in order to be visible in the various views selected for broadcast. In some cases, the line was considerably larger than the diameter of the ammunition fired from Oswald's rifle and appeared to encompass more of the knot of the tie than would have actually occurred. The varying size of the trajectory line was, of course, illustrative in nature and has no bearing on the issue of the validity of the single bullet theory as some conspiracy theorists have postulated. In fact, none of the clothing, as illustrated in the computer recreation has any bearing on the validity of the single bullet theory, since no attempt was made to match the clothing exactly nor was the trajectory path based on the location of defects in the clothing of either President Kennedy or Governor Connally. The trajectory path was based solely on the locations of the wounds as deduced by the House Select Committee on Assassinations in 1978.
How is it that you located the governor's front chest wound higher than the bullet hole in the front of his suit jacket?
As with the president's clothing, the location of holes in the governor's clothing, both front and back, were not used to position his wounds or calculate any trajectories for the simple reason that clothing can, and often does, move in relation to the location of wounds. Instead, I based the location of the governor's chest wound on the 1963 post-operative report of Doctor Robert Shaw and the 1978 personal examination of Connally's wound scars by Dr. Michael Baden, both of which, I found, correlate to the same locations in geometric space. It is the medical evidence that is used to position the wounds, not the clothing. In addition, there is no direct correlation between the computer model's "clothing" and the governor's actual suit jacket. No attempt was made to replicate the style of clothing of either man to any degree of accuracy, beyond basic cosmetic attributes.
Assuming you are correct in your analysis that a single bullet fired from the Texas School Book Depository passed through Kennedy's neck and struck Connally in the right armpit, why didn't you extend that trajectory line forward to explain Connally's wrist and knee wounds on the ABC News special?
The producer's of the ABC News special were satisfied that the single bullet theory was amply demonstrated by showing that both men were aligned to be struck by a single bullet. Obviously, as that single bullet entered the area of Connally's right armpit it had to continue forward. Doctor Robert Shaw, who performed the operation on the governor, told the press immediately after the surgery that it was his belief that all of the governor's wounds were caused by a single bullet. In my opinion, all of the evidence surrounding the governor's wounding supports this conclusion rather than refutes it. Obviously, the exact path that the bullet would have to take to connect all five wounds in the governor is a matter of conjecture since Connally's right wrist (the location of two of his wounds) and left thigh (the location of one wound) are not visible in the Zapruder film, and therefore cannot be positioned as they were at the moment of impact with any degree of accuracy.
How can you possibly interpolate what happened behind the Stemmons freeway sign while Kennedy and Connally were hidden from Zapruder's camera?
One critic wrote, "...the trajectory of a bullet may be accurately interpolated when starting and ending positions are known. The positions of relatively motionless humans absolutely cannot. You have no inertia to establish a path. Perhaps you could interpolate the motions of say, a quarterback throwing a pass or other highly predictable patterns, but certainly not the actions of individuals randomly responding to being shot."
Wrong. Categorizing "highly predictable patterns" as numeric interpolation is a complete misunderstanding of the concept. Allow me to explain. Numeric interpolation involves three elements: a start point, an end point, and a time factor. A motionless object has the same start and end point. An interpolation of the position of a motionless object over a given period would show that the object remained motionless. So, motionless objects can in fact be interpolated. But, of course, that would be a rather silly exercise.
In the case of the JFK assassination, both Kennedy and Connally are seen moving throughout the Zapruder film. Between any two points in the film we have the necessary ingredients to numerically interpolate motion (i.e., start and end frames, and a time variable). The ability to correctly interpolate motion between any two points in the film is directly related to the time variable between those two points. Extended time variables create less accurate interpolations; short time variables create more accurate interpolations. Why? Because interpolating motion over a long period of time tends to soften or smooth out motion between two points, causing some inaccuracies. By contrast, interpolating motion over short periods of time (like the 1.25 second period of time Kennedy and Connally were behind the Stemmons freeway sign) tends to be more accurate, especially when the objects being interpolated have significant mass. Why? The reason is elementary. Objects of mass (i.e., large objects) move slowly. Objects of little mass (i.e., smaller objects) can move quickly. Small movements could pass without detection using interpolation. For instance, it is likely that moving a pinky finger would escape detection through interpolation, while the flexing of an arm would not.
To interpolate the motion of Kennedy and Connally behind the Stemmons sign, three-dimensional models representing the two men were posed to match their last known position before disappearing from view (this became the start point). Next, they were posed to match their earliest known position as they emerged from behind the sign (this became the end point). The time period in which they were hidden from Zapruder's camera was then interpolated to produce a motion sequence that bridged the start and end points. Because the time period being interpolated is of a short duration and the sizes of the objects (upper torso, arms, and head) are of considerable mass, the interpolated results are relatively error free.
After interpolating the motion of the two men behind the sign, I examined the resulting sequence in the context of the motion known to have occurred both before and after the interpolated sequence. This step is the key to understanding how interpolation is used in forensic animation. In my experience, if anything unusual had occurred behind the sign, the interpolated motion paths would be seen as a sudden "jump" or "acceleration" when placed in the context of the surrounding action. In other words, the interpolated sequence would interrupt the flow of motion between the sequence that preceded it and the one that followed. That's because the interpolated motion sequences are nothing more than linear projections and offer nothing when viewed in isolation. To be meaningful, the interpolated motion must be viewed in context. Think of it this way, the Zapruder frames preceding the sign show the president and governor moving at a certain speed and manner. The frames after they emerge from behind the sign also show the men moving at a certain speed and manner. If there were no large, dramatic change in the position of either man, the interpolated motion would bridge the two known segments in a contiguous and believable manner. If some dramatic change in posture did occur behind the sign, then the interpolated motion path, when seen in context, would seem interruptive, destroying any continuity between the two known segments.
One of the advantages of using computer graphics to analyze the assassination is that we can alter the surfaces of selected objects. In this case, we can make the Stemmons Freeway sign semi-transparent, which allows us to view the interpolated segment in context without the interference of the sign.
How does the interpolated motion of Kennedy and Connally fit into the context of the surrounding motion? When the sequence is viewed without the sign's interference, it is clear that the interpolated motion forms a contiguous bridge between the known actions of the two men as seen both before and after the Stemmons freeway sign. In short, Kennedy and Connally made no large, dramatic movements while hidden from Zapruder's camera.
Didn't you mislocate the entrance wound in the back of the president's head?
No. One critic claimed that I placed the entrance wound in the back of the president's head in a location that matches neither the report of the autopsy doctors (which located the wound low in the back of the head) nor the report of the House Select Committee on Assassinations (which located the wound high in the cowlick area).
This claim is false. As detailed in the conclusions section of this website, I did in fact base the location of the president's head wounds on the 1978 report of the HSCA Forensic Pathology Panel which located the entrance wound on the back of the head about 3.9 inches above the external occipital protuberance, and 0.7 inches to the right of the midline. I confirmed this location by constructing the only known three-dimensional model of the president's skull as it appears in two X-rays made on the night of November 22, 1963. These X-rays were authenticated by the HSCA Forensic Pathology Panel in 1978. The model confirms the entrance wound location reported by the HSCA panel. This model was sized to match the actual size of the president's skull, then "skinned" using a dimensional model of Kennedy's head, created from a digitized clay bust. The location of the entrance and exit wounds were located on the underlying skull model and transferred to the appropriate locations on the surface of the "skin."
Again, the criticism surrounding the placement of the president's head wound is probably due to the position of a series of red, cross-hair markers used during the ABC Television broadcast to show the locations of the various wounds in Kennedy and Connally. For reasons of brevity, the producers of the program decided to mark the wound locations on the fully clothed figures of Kennedy and Connally rather than on the surface of the skin. In the case of the entrance wound in the back of the head, the marker had to be placed on the outside surface of the "hair." In order to be visible to audience, the marker was placed at the point where the trajectory path intersected the outside surface of the hair. Because of the angle involved, it was necessary to slightly offset the marker from the actual wound location. This is why there appears to be a discrepancy between the location of the entrance wound on the computer model and the actual wound. But, appearances can be deceiving. And in this case, there is no discrepancy. The marker has merely been positioned along the projected trajectory path at the point at which the path intersects the surface of the hair. The wound itself remains properly fixed at the appropriate location, though invisible to the viewing audience.
Why didn't you deal with the backward head snap of Kennedy's head in his recreation?
While the ABC News special didn't focus on the backward head snap, due to the complexities involved and the time constraints of the program, my computer recreation project, Secrets of a Homicide, did focus on issues surrounding the fatal head shot. (See: Conclusion #3 and Conclusion #4)
Despite popular beliefs, the medical evidence clearly shows the president was struck in the head from behind by one bullet. My dimensional analysis of that moment shows that the president's head was driven forward approximately two inches in 1/18th of a second by the impact of the bullet. Yet, the backward snap of the head took four times as long (1/2 second) to recover that same distance. This simply means that the force causing the backward head snap was not as powerful as the force of the bullet that struck him from behind. This fact alone is compelling evidence that the backward head snap was not due to a bullet fired from in front of the president, as some conspiracy theorists contend.
It is also worth noting that my dimensional analysis shows that the president was not thrown "back and to the left," as many believe. President Kennedy was already leaning left at the time of the fatal head shot. An aerial view shows the President's upper torso moving straight back until it makes contact with the rear seat cushion, his head tilted to his left. Neither movement is consistent with any bullet trajectory originating from the grassy knoll and driving Kennedy back and to his left.
Finally, a dimensional analysis of possible firing sources for a bullet striking the right front of Kennedy's head and exiting at the lower right rear - a popular conspiracy theory - proves that no such shot could have originated from the grassy knoll (see Conclusion #4).
Why didn't you attempt to simulate any other shooting scenarios?
The computer recreation is not a scenario, a theory, or an illustration of what I believe happened in Dealey Plaza. The computer recreation is a three dimensional animated model of the shooting as recorded by Abraham Zapruder and other photographers, both amateur and professional. Therefore, there cannot be several versions of what happened.
Isn't it true that your animation fails the vanishing point test?
No. This charge was made by one theorist in a conspiracy-oriented newsgroup shortly after the animation aired on ABC-TV and is demonstratively false. The theorist wrote, "One of the things that caused me to question the accuracy of Dale's animation was that if he put in the same data as seen on the Zapruder film frames, then why did I not see angle changes on some of the points along the car? I thought a 3D image is supposed to show perspective?" The theorist then offered readers a test, overlaying lines drawn on a frame from the animation. The theorist concluded, "I didn't find the angle to change accordingly. If this data is missing from his 3D plaza model, then how can it be accurate when it uses that same data to build the images needed to show bullet trajectories...?"
Contrary to the theorists' claim, the computer recreation does show a changing perspective throughout the animation, a fact easily demonstrated by simply viewing any of the sequences. The false claim stems from the theorist choosing a close-up frame from the animation which contains little information on which to base a vanishing point analysis (i.e., in perspective drawing, a point on the horizon at which receding parallel lines seem to converge). Any one of thousands of other frames from the animated sequences, particularly wide angle views that provide a significant number of points on which to base a proper analysis, clearly prove this charge to be false.
Isn't it true that individual frames from your animation fail to match the Zapruder film frames upon which they are based? Doesn't this prove that your animation is inaccurate?
No. One theorist wrote, "I had taken one of Myers' animation frames (Z222) and I overlaid it onto an actual Z222 film frame. If the data that he used was as accurate as the show let on, then everything seen in the animation concerning the limo and its occupants should match up. The result was that they did not! I think Dale had an interesting approach and I am not yet prepared to say that he purposely fudged the points to get a said result, but I do think his method and reference points should have been tested or at least observed by someone else so to try and get a common ground for accuracy. Maybe in the future someone will do this again, maybe even check it by way of another film such as Bronson's or Muchmore's."
In fact, this false charge assumes a number of things. First, it assumes that the computer animated sequences broadcast on ABC-TV are a frame-for-frame match of the Zapruder film. They are not. As discussed on this website, the computer recreation was key-framed to the 18.3 frame-per-second Zapruder film frame-by-frame (the methodology utilized in achieving this alignment was investigated and found to be highly accurate by forensic animation experts from Z-Axis Corporation), then, re-photographed at 30 frames-per-second. This process creates a frame sequence that is slightly offset in time (i.e., in fractions of a second) from the original film. Consequently, none of the computer animated frames will exactly match the Zapruder film frames. This fact does not invalidate the computer recreation. The animated sequence remains a model of the original event, re-photographed at a higher resolution.
Second, the false charge assumes that the virtual camera used to re-photograph the assassination sequence occupies the exact same position, frame-by-frame, as Zapruder's actual camera. It does not. In order to eliminate the hand-held camera movements that make viewing the original Zapruder film difficult, the erratic camera movements in three-dimensional space (i.e., the six degrees of freedom in the camera's movement) were removed from the virtual camera's motion track. The result provides a smooth, unfettered view of the assassination. Of course, this would also make it impossible to precisely line-up any of the resulting rendered images with the original film.
It should be noted that even if the rendered images had matched the original film, the theorist making the false charge would still have not been able to achieve an alignment since he failed to use the correct aspect ratio when attempting to make the alignment. This is a common error found in the work of many conspiracy theorists critical of my computer animation work.
Finally, as this website points out, many other films and photographs of the assassination were used to ensure the accuracy and proper alignment of my animated model to the Zapruder film.
Isn't it true that your single bullet trajectory doesn't align with the sniper's nest window in the Texas School Book Depository when your animation is superimposed over the Charles L. Bronson slide, an actual photograph taken on the day of the shooting?
No. One theorist wrote that he tested my trajectory analysis by superimposing a frame from my animated sequence over a color slide taken by Charles L. Bronson, a manufacturing engineer who was present on the day of the shooting. The theorist concluded, "By applying Dale Myers' angle of trajectory against the real world showing the actual [Book Depository] behind the trees, we can then trace the trajectory east to see if it actually reaches the 6th floor alleged assassin's window." The theorist reported that "...the [trajectory] DOES NOT come close to the 6th floor at all." The theorist postulated that "...the limo would have to of been passing Jean Hill's position..." (i.e., considerably further westward down Elm Street) to achieve the trajectory I projected.
In fact, these false charges stem from a general lack of knowledge about three-dimensional animation, photographic technique, and the science of photogrammetry. The theorist demonstrated his lack of knowledge about dimensional animation by writing, "Myers' animation is called a 3D animation, but it isn't really. To be 3 dimensional it has to show a field-of-depth (sic) and it doesn't. Not on the stills - or on the animations does it do this." The theorist is confusing dimensional animation with a 3D image (i.e., an image that can be viewed stereoscopically). While the rendered images I produced for television are indeed two dimensional, the geometry used to produce the images is in fact three dimensional (i.e., containing width, breadth, and depth). In fact, three dimensional images can easily be created from the existing geometry by re-rendering the sequences stereoscopically (i.e., rendering right and left eye perspectives for stereo viewing.)
The theorist further demonstrated his lack of knowledge about photographic technique by taking a rendered frame from my animated re-creation and superimposing it over a photograph taken by Robert Croft on the day of the assassination. While both images are clearly taken from two different angles, the theorist claimed that his photographic overlay proved that the single bullet trajectory I computed would have hit Governor Connally much lower in the back than where his wound was actually located. This is, of course, silly and misleading at face value.
Finally, the theorist demonstrated his lack of knowledge regarding the science of photogrammetry by writing, "...you can do [photographic] overlays that allow measurements to be taken..." and "Any software program that allows one to place angled lines on an image can test my Bronson slide overlay. The angled line Myers uses is the exact same pitch as the one coming down from the alleged snipers lair. So what other way is there to interpret Myers line of trajectory when all you have to do is overlay it - and then slide it over to the 6th floor east most window? Myers has his occupants in a vertical sitting posture. I took his vertical alignment and matched it against other vertical objects in the Bronson slide. The degree of error cannot be seen with the naked eye, yet the trajectory going way over the top of the limo is easily seen. So enlighten me as to how I misinterpreted Dale Myers animation models?"
This particular theorist, a self-described "photographic expert," made a classic mistake reminiscent of conspiracy theorist Jack White's blunder during the 1978 House Select Committee on Assassinations (HSCA) hearings. Testifying before the congressional committee, White presented photographic "evidence" that the Mannlicher-Carcano rifle alleged by police to be the murder weapon was in fact multiple weapons of varying lengths. White's technique to uncover this "evidence" was to simply measure the length of the subject rifle as it appeared in various two-dimensional photographs. Acclaimed photographic experts on the HSCA's photographic panel pointed out that White failed to take into account the angle at which the rifle was seen in each photograph relative to the camera's line of sight. This relationship is the basis of the science of photogrammetry (of which White admitted he had no knowledge of or practical experience in).
Photogrammetry describes how three-dimensional spatial relationships can be extracted from two-dimensional photographs or images. Without taking into account these relationships, accurate interpretations of two-dimensional images are impossible. In short, you cannot simply draw or overlay lines on a two-dimensional image (as White and the subject theorist have claimed) and extract three-dimensional information.
In this case, the theorist used rendered two-dimensional images from my computer recreation, which were taken from angles unknown to the theorist, overlaid them unto a second set of images, taken from yet another set of unknown angles, and then proposed that he was able to project hand-drawn, two dimensional lines into three-dimensional space. Frankly, this would be impossible.
For a primer on photogrammetry, click HERE.
Isn't it true that you have has been peddling your computer recreation for years, changing the results until you got ABC News interested?
No. I began working on the concept for Secrets of a Homicide: The JFK Assassination in 1993, completing a preliminary Research Edition in 1995. The Discovery Channel expressed interest in the project in 1996 but no agreement was ever reached. I continued to improve the visual look of the recreation and in early 2003 producer Mark Obenhaus learned of the computer project and approached me about participating in a fortieth anniversary television special for ABC News. The conclusions I reached and presented in 1995 are identical to those depicted in the 2003 ABC News special.
Wasn't your computer recreation featured in Gerald Posner's 1993 book, Case Closed?
No. The graphics featured in Posner's book were created by Failure Analysis Associates (FAA) for a mock trial conducted for the American Bar Association. While FAA studied the positions of Kennedy and Connally at key positions in the Zapruder film, they did not attempt to re-create the entire film nor did they use the same method I used to achieve an accurate alignment. A few black and white still frames from my earliest work did appear in Gus Russo's Live By The Sword.
How can you claim to have removed human fallibility from your recreation?
While a number of critics have made this claim, I have never suggested that the recreation is without flaws. In fact, I have pointed out that despite the use of highly accurate models in my recreation, flaws are inevitable. What is at issue is whether those flaws are significant enough to alter the conclusions. In order to answer the question, I used error cones to calculate any potential errors in constructing the models, positioning the figures, and projecting possible trajectories. The results show that even when potential errors are considered, the conclusions are the same.
Why should I believe your conclusions?
As the old saying goes, "The proof is in the pudding." In this case, the information sources that formed the basis of the computer recreation are readily available to the general public. Software programs capable of generating valid geometry for such an analysis are also available. And the results are repeatable, as my own work demonstrates. In the end, my conclusions regarding the single bullet theory are not radically different than those reached by the Warren Commission in 1964 and the House Select Committee on Assassinations in 1978.
It is often pointed out by critics that 70% of Americans believe there was a conspiracy to assassinate President Kennedy, as if the percentage itself is enough to change the reality of what really happened. Six hundred years ago, the vast majority of humanity believed the world was flat. That still didn't change the reality of a spherical world. We know that many Americans base their opinion of the Kennedy assassination on a mixture of fact and fantasy derived from a variety of often suspect sources of information. While opinion polls can reveal cultural trends, they don't change facts. In the final analysis, the truth doesn't require anyone's belief.
Published by Oak Cliff Press, Inc., P.O. Box 608, Milford, MI 48381-0608