Let There Be Sound!: The Acoustics Evidence in the JFK Assassination

[NOTE: This article is essentially an almagamation of material culled from my two recent CTKA book reviews, with minor changes and additions, compiled here as much as anything else for my own easy reference.]

On November 22, 1963, the day of President Kennedy's assassination, the microphone on a police motorcycle travelling in the Presidential motorcade had become stuck in the “on” position and the sounds had been recorded on a dictabelt machine at Dallas police headquarters. When the dictabelt was brought to the attention of the HSCA in 1978, it asked the top acoustics experts in the country to analyze the recording to see if it had captured the sounds of the assassination gunfire. James Barger and his colleagues at Bolt, Baranek & Newman (BBN) discovered six suspect impulses on the tape that occurred at approximately 12:30 p.m.—the time of the assassination—and reported that on-site testing needed to be conducted at Dealey Plaza. There, microphones were placed along the parade route on Houston and Elm Streets and test shots were fired from the two locations witnesses had reported hearing shots; the Texas School Book Depository and the grassy knoll. BBN found that five of the impulses on the dictabelt were found to acoustically match the echo patterns of test shots fired in Dealey Plaza. (8HSCA101) One of these, the fourth in sequence, matched to a shot fired from the grassy knoll. (8HSCA10) The fact that the suspect sounds had matched to some of the 423 test patterns is not, by itself, amazingly significant. However, the order and spacing of the matching microphone positions followed the same order as the sounds on the police tape.

If the sounds on the dictabelt were not truly the sounds of the assassination gunfire and instead represented some type of random static, a match would be as likely to occur at at the first microphone as the last and would most likely fall in some random order—there being, of course, 125 different ways to sequence five events. But far from being random, the matches fell in the exact same 1-2-3-4-5 topographic order as the impulses appear on the dictabelt recording.

• The first impulse matched to a test shot recorded on a microphone on Houston Street near the intersection with Elm.
• The second to a microphone 18 ft north on Houston.
• The third to a microphone at the intersection.
• The fourth to a microphone on Elm.
• And the fifth to the next microphone to the west.

Additionally, the distance from the first matching microphone to the last was 143 feet and the time between the first and last suspect impulse on the tape was 8.3 seconds. In order for the motorcycle with the stuck microphone to cover 143 feet in 8.3 seconds it would need to be travelling at a speed of approximately 11.7 mph which fits almost perfectly with the FBI's conclusion that the Presidential limousine was averaging 11.3 mph on Elm Street. (Warren Report, p. 49)

Finally, the gunshots on the dictabelt synchronize perfectly with the visual evidence of the other all-important record of the shooting, the Zapruder film. There are two visible reactions to gunshots on the Zapruder film. One of these occurs at Z-frame 313 with the blatantly obvious explosion of President Kennedy's head. The other occurs between frames 225 and 230 when the Stetson hat in Connally's hand flips up and down, most likely as a result of the missile passing through his wrist. When the fourth shot on the dictabelt, the grassy knoll shot, is aligned with Z-frame 313, the third shot falls at —yes, you guessed it—frame Z-225. This means that the exact same 4.8 second gap between shots is found on both the audio and visual evidence. These perfect correlations between the acoustics and all other known data provide the most convincing reasons to believe that the dictabelt is a genuine recording of the assassination gunfire. Unfortunately, this remarkable concordance was hidden from the public when HSCA chief counsel, Robert Blakey, in a “socially constructive” move, convinced the experts to label the third shot as a “false alarm.”

To understand Blakey's decision it is necessary to understand Blakey. Former HSCA staff investigator, Gaeton Fonzi, wrote in his brilliant book The Last Investigation, that, “Chief Counsel Blakey was an experienced Capitol Hill man. He had worked not only at Justice but on previous Congressional committees as well. So he knew exactly what the priorities of his job were by Washington standards, even before he stepped in.” (Fonzi, p. 8) Blakey, who later admitted that before he took the job he had found the idea of a conspiracy in the JFK case “highly unlikely,” (ibid. p. 259) was destined not to stray too far from the Warren Commission's conclusion that only three shots were fired and all were fired by Lee Harvey Oswald. As such, the acoustics evidence presented him with a big problem. As Dr. Donald Thomas puts it, “The acoustical evidence simply did not mesh well with the Warren Report...Blakey's problem was not just that a total of five putative gunshots were detected by BBN's test procedures, but that these shots came too close together.” (Hear No Evil, p. 584) In 1964, the FBI established that “Oswald's” rifle required 2.3 seconds between shots and, as Special Agent Robert Frazier testified, this was “firing [the] weapon as fast as the bolt could be operated.” (3H407) But the first three shots on the dictabelt had all come from the general vicinity of the book depository and came only 1.65 and 1.1 seconds apart. To “solve” the problem, Blakey acquired a Mannlicher Carcano similar to the one found on the sixth floor and, together with a group of Washington police officers, practised firing the rifle as fast as possible. Apparently, by “point aiming” - which means not really aiming at all - Blakey and HSCA counsel Gary Cornwell were able to squeeze off two rounds in 1.5 and 1.2 seconds respectively. (8HSCA185) This farcical display was enough to satisfy Blakey about the “probability” that Oswald fired the first two shots on the tape. He then told the acoustics experts that the third shot, coming only 1.1 seconds after the second, could not be what their analysis told them it was. And in another socially constructive move, the scientists played along.

The truth is that all three matches were as valid as each other and what the acoustics evidence actually showed was that there may have been a second rearward assassin and a triangulation of crossfire—just as critics like Josiah Thompson had been saying since 1967. But a Washington man like Blakey was not about to admit that the “buffs” had been right all along. In a conversation with Thomas in 1999, “Blakey confided that he knew he would take a lot of heat for the grassy knoll shot and he didn't want to dilute his case with the weak evidence for a fifth shot.” (Thomas, p. 590) By putting political considerations before the evidence, Robert Blakey did history a huge disservice and helped obscure the truth about the assassination. By cutting out the crucial third shot, he had essentially hidden the perfect synchronization between the dictabelt and the Zapruder film and it was for this very reason that many JFK researchers rejected the validity of the acoustics evidence. One can only wonder what reception the Dallas police dictabelt would have received had Blakey had the courage to stand up for the truth.

Predictably, the conclusions of the HSCA scientists have been attacked by numerous individuals and organizations seeking to uphold the conclusions of the Warren Report; usually on non-ballistic acoustic grounds. For example, a National Research Council panel commissioned by the Justice Department claimed that an instance of “cross-talk” on the dictabelt recording placed the suspect impulses a full minute after the assassination. What the panel chose not to report is that there are five instances of cross-talk on the dictabelt, none of which even synchronize with one another. Therefore, it is not possible using cross-talk alone to prove that the gunshots on the tape are not synchronous with the assassination. A more recent challenge came from Dale Myers (the Walt Disney of JFK research) who claimed that his analysis of the available assassination films and photographs “proved” that police officer H.B. McClain's motorcycle was not where it needed to be. This despite the fact that there is no film or photograph currently known to exist that shows the acoustically required position at the time the shots were fired.

In his recent book, Head Shot: The Science Behind the JFK Assassination, Dr. G. Paul Chambers, an experimental physicist for the US Navy and a contractor with the NASA Goddard Optics Branch, has calculated the odds against the impulses on the dictabelt being caused by random static. As mentioned above, the order in the data is by itself hugely compelling. The last in the sequence of test shot matches occurred at a microphone 143 feet from the first, and the time between the first and last suspected shots on the dictabelt was 8.3 seconds. In order for the Police motorcycle officer whose stuck microphone was suspected of recording the gunfire to travel 143 feet in 8.3 seconds he would need to be traveling at approximately 11 mph—almost the exact speed at which the FBI estimated the Presidential limousine was moving on Elm street. As Chambers asks, “What are the odds of that happening randomly?...One could certainly insert a big number for the total number of possibilities, leaving a very small probability that this would happen randomly. But it isn't necessary.” (Chambers, p. 142) On top of this, we have the fact that the timing of the shots fits so perfectly with the reactions seen on the Zapruder film. “Syncing the final head shot from the grassy knoll to frame 312...” Chambers explains, “the probability of finding the shot that hit Connally to within five frames...is about one in a hundred...Matching up the first shot to the frames before Kennedy reaches the Stemmons Freeway sign and the second shot to a strike of Kennedy behind the sign is another one chance in a hundred times one chance in a hundred for a one-in-ten-thousand chance for an accidental match.” Multiplying all this by the probability of all shot origins falling in the correct order is another one chance in sixteen, “yielding a one-in-sixteen-million chance that the acoustic analysis could match up the timing and shot sequence in the Zapruder film by chance.” Multiplying the probability of both the order in the data and the synchronization of the audio film being random together, “it is readily established that there is only one chance in eleven billion that both correlations could occur as the result of random noise.” (ibid pgs. 142-143)

Is Chambers math correct? I'll leave that to those better qualified than me to test. What I do know is that Dr. Don Thomas, who is a widely published expert statistician, has calculated the odds of a random impulse having the acoustic fingerprint of a rifle shot from the grassy knoll as 100,000 to one against. (Thomas, p. 632) And anyone with an even vaguely scientific mind can see that this calculation alone is enough to establish the validity of the acoustics evidence beyond any real reasonable doubt.