APPENDIX A

ANALYSIS OF THE TECHNICAL REPORT ON THE 1812 MINUTE GAP

On November 21, 1973, Chief Judge Sirica was informed by the President's counsel that the tape of a June 20, 1972 conversation between the President and Haldeman contained an 1812 minute_buzz which obliterated the recorded conversation. Subsequently, Judge Sirica asked a panel of six technical experts, previously appointed by the Judge and endorsed by the Special Prosecutor and the counsel for the President, to determine and report on the nature and cause of the obliteration of that tape recording that had been supoenaed by the Watergate Grand Jury. (Book IX, 871) On January 15, 1974, the panel reported the conclusions of its study to Judge Sirica (Book IX, 926-28) and on May 31, 1974 the panel's final report on the EOB tape of June 20, 1972 was submitted to the Court. The key conclusions of the panel were:

(1) The Uher 5000 tape recorder used by the President's secretary, Rose Mary Woods, to transcribe tapes of Presidential conversations probably produced the 1812 minute erasure and buzz.

(2) The 1812 minutes of erasure and buzz were accomplished by at least five, and perhaps as many as nine, contiguous and separate operations.

(3) Erasure and recording of each segment of erasure and buzz required manual operation of keyboard controls on the Uher 5000 recorder. (May 1974 Tape Report, 35-36)

The Uher 5000 tape recorder, as it true of the Sony 800B tape recorder used to record the Presidential conversation, has two magnetic "heads," an erase head and a record head. (The record head performs both recording and playback functions.) When the "playback" button on the tape recorder is depressed, the erase head is inactive while the record head is activated to pick up electronic signals from the magnetic tape as the tape is drawn across it. The machine then translates the electronic signals into sound. When the "record" button is depressed, both the erase head and the record head are activated. The tape is drawn first over the erase head where the tape is cleansed of prior magnetic signals and then over the record head where new magnetic signals, representing the sounds being recorded, are imparted to the tape. To erase a tape, the "record" button is depressed but no new sounds are introduced into the recording machine; the tape passes over the erase head and is erased, and then over the activated but silent record head.

The Uher 5000 machine may be used in conjunction with a foot pedal. The pedal is capable only of moving the tape forward at recording speed or backward at the higher rewind speed. The foot pedal cannot, in effect, depress the "playback" or "record" button; it cannot activate or deactivate either the erase head or the record head. (Thomas Stockham testimony, In re Grand Jury, Misc. 47-73, 1/15/74, 16)

(227)

228

Whenever the record head is activated by depression of the "record" button, it leaves a distinctive "record-head-on" mark on the tape. (Richard Bolt testimony, In re Grand Jury, Misc. 47-73. 1/15/74, 2172) When the "record" button is released, and the erase and record heads are deactivated, the electronic pulses dying on those heads leave distinctive "erase-head-off" and "record-head-off" marks, respectively, on the tape. (Thomas Stockham testimony, In re Grand Jury, Misc. 47-73, 1/15/74, 12-13) The "record-head-on," "erase-head-off” and "record-head-off" marks vary from one type of machine to another, and may be used to help identify the machine on which tapes were recorded or erased.

The panel was able to identify five clear sets of "on" and "off" markings which enabled it to determine that erasure of 1812 minutes of the June 20 conversation was accomplished in at least five different segments. (Richard Bolt testimony, In re Grand Jury, Misc. 47-73, 1/15/74, 8)

When a segment of erasure is completed, and the machine is reversed and restarted, the "on" and "off" markings of previous erasures may themselves be erased. The panel found four additional markings that might have been part of segments of erasure where the matching "on" or "off" markings themselves had been erased; the panel could not be sure whether these marks were evidence of additional segments of erasure. (Thomas Stockham testimony, In re Grand Jury, Misc. 47-73, 1/15/74, 21-22)

The Advisory Panel conducted the following tests and analyses on the June 20 tape in reaching its conclusions:

1. Critical Listening

The panel played 67 minutes of the evidence tape, including the 182 minute buzz, through high quality back-play equipment. Their expertise enabled them to identify and clarify significant acoustic phenomena on the tape. (May, 1974 Tape Report, 8)

2. Magnetic Marks

The tape was treated with a liquid that "developed" the tape, that is, rendered visible the magnetic patterns and markings on the tape, such as "record-head-on," "record-head-off," "erase-head-off," and "K1-pulse" (see below) marks. (May, 1974 Tape Report, 8-11)

3. Wave Forms

When the electrical output of a recorded tape is fed into an oscilloscope, each signal on the tape produces a distinctive wave form. Wave form analysis enabled the panel to make a detailed study of the significant events on the June 20 tape. The panel scrutinized the wave forms of the events that occurred during the 1812 minute erasure and buzz, and found that the wave form analysis corroborated the conclusions drawn from examination of the magnetic marks. (May, 1974 Tape Report, 11-13)

4. Spectra of Speech and Buzz

Through spectral analysis (analyzing the component frequencies and amplitudes of sound signals), the panel was able to study the differences, similarites, and time of the signals. Through use of a chart of the spectral analysis of the 1812 minute buzz (a spectrogram), the panel was able to examine "windows" (tiny fragments) of original speech, to conclude that the 60 cycles per second power line hum was the source of the buzzing sound, and to corroborate the

229

evidence of stops and starts indicated by the magnetic marks. (May, 1974 Tape Report, 13-16)

5. Phase Continuity and Speed Constancy

There is a discernible wave pattern in the power line hum on all recorded tape; this wave pattern will be of a continuous nature until the recording is stopped. Each uninterrupted portion is called a phase. The panel could determine where the recording mode has been stopped and restarted by noting the phase discontinuities. The phase discontinuities on the June 20 tape corroborated the "stop" and "start" conclusions drawn by the panel from their study of the magnetic marks and wave forms. (May, 1974 Tape Report, 16-18, 43)

6. Flutter Spectra

The mechanical irregularities in the rotating elements of every tape recorder are unique to that machine. These irregularities produce additional tones known as "flutter sidebands," distinct from the machine's original or "pure" tone.

The degree of "flutter" can be plotted, and this phenomenon will aid in the identification of a particular tape recorder.

The panel used this test to determine which machine was responsible for recording the 181⁄2 minute buzz on the tape. (May, 1974 Tape Report, 18-20)

7. Search for Physical Splices

The panel studied the June 20 tape with an instrument (an accelerometer) that could measure and detect any variances in tape thickness. The panel concluded as a result of their studies, that the tape contained no physical splices. (May, 1974 Tape Report, Technical Note 13.1)

8. The K-1 Switch

As further proof that the erasure was caused by manipulation of the keyboard, the panel studied evidence of K-1 pulses on the tape.

The K-1 switch is an internal mechanical switch. This switch only opens and closes as a result of pushing certain keys on the keyboard. It cannot be actuated by a malfunction in the electronics of the recorder. It cannot be actuated by the foot pedal. (May, 1974 Tape Report, 45) The switch opens and closes as a result of a physical latching and unlatching action that only occurs when one of the keys is pressed down manually. There are four keys that can close this switch: the recording key, the rewind key, the start key, and the forward key. (May, 1974 Tape Report, Technical Note 8.3)

K-1 switch activity is reflected on the tape by K-1 pulses. Because of the many other larger transient pulses that are generated by other electro-mechanical activity, K-1 pulses are difficult to discern. However, where a K-1 pulse is unambiguously identified, it is an unmistakable sign of manual activity of the keyboard. The expert panel was able to identify six distinct K-1 pulses. (May, 1974 Tape Report, Technical Notes 8.3-8.5)

ALTERNATE HYPOTHESES

A number of alternative hypotheses to the conclusions reached by the expert panel were considered and rejected by the panel in arriving at its conclusion, including the following:

51-908 98-9

230

Hypothesis No. 1

That the 1812 minute gap was produced on the June 20, 1972 tape at the same time that the tape was originally recorded. This hypothesis failed because the June 20, 1972 original tape was recorded on a Sony 800B tape recorder. The experts determined that the 1811⁄2 minute gap was produced by a Uher 5000 tape recorder. (May, 1974 Tape Report, Technical Notes 9.1-9.2)

Hypothesis No. 2

That the 1812 minute obliteration was caused by setting the Uher tape recorder in the record mode and operating it in fast rewind. This hypothesis was rejected because if the tape had been erased in rewind the obliterated section would have had an audible tone of 500 cycles when played back at its usual operating speed of 24 millimeters per second. However, the frequency that is on the 1812 minute gap is the normal 60-cycle frequency. This shows that the tape was erased at its standard operating speed of 24 millimeters per second. Additionally, if the 1812 minute buzz had been recorded in rewind, there would have been no record and erase-head-off marks left on the tape. More than 20 such marks were found in the obliterated section. (May, 1974 Tape Report, Technical Note 9.2)

Hypothesis No.3

The tape was erased through use of the foot pedal. This hypothesis was rejected because of the record and erase head signatures that were found on the tape; signatures that cannot be made by the foot pedal. Second, a distinctive set of magnetic marks is made by the Uher tape recorder when stopped and restarted by the foot pedal. None of these marks was found on the 18 minute buzz section. Furthermore, six K-1 pulses were found in the obliterated section. K-1 pulses also cannot be made by the foot pedal. (May, 1974 Tape Report, Technical Notes 9.2-9.3)

Hypothesis No. 4

The distinctive magnetic marks found on the 1812 minute gap came from a power supply failure within the Uher 5000 machine, i.e., a defective diode caused the power supply to sputter on and off, thus putting the distinctive marks on the tape while the tape was still moving. The experts rejected this hypothesis because they were able to determine that the wave forms that would have been produced by this sort of activity were not present on the evidence tape. Furthermore, if this "sputter" activity had taken place, there would be no phase discontinuity following the record-head-on marks. The evidence tape shows phase discontinuity and erase head signatures associated with the record-head-on marks. Additionally, there are K-1 pulses found on the tape that could only be caused manually. (May, 1974 Tape Report, Technical Notes 9.3-9.5)

Hypothesis No. 5

Voltage irregularities on the AC power line working in conjunction with the failing diode of the bridge rectifier caused the distinctive magnetic marks. A voltage drop sufficient to put these marks in the tape would have caused a drop in motor speed with a resulting differen

231

tial in tone frequency. There was no evidence of this on the evidence tape. Moreover, a drop in voltage could not cause the recording of K-1 pulses. (May, 1974 Tape Report, Technical Notes 9.6-9.8)

THE STANFORD RESEARCH INSTITUTE REPORT OF MAY 31, 1974

Dr. Michael Hecker of the Stanford Research Institute conducted experiments for the Special Counsel to the President with regard to the June 20, 1972 tape. It should be noted that while Dr. Hecker reviewed experiments and held a number of conferences with the expert panel, he never studied the June 20, 1972 tape directly. (SRI Report) Dr. Hecker reviewed the findings of the expert panel and stated that he agreed with the panel's approach and agreed with the panel's expertise. Dr. Hecker stated further that he was in substantial agreement with the panel's final report. (SRI Report 3) The Stanford Research Institute found evidence that there had been manual manipulation of the keyboard controls of the Uher 5000 tape recorder in order to cause some portions of the 181/2 minute gap. The Stanford Research Institute studied and rejected all the alternative hypotheses that were considered by the panel. (SRI Report, 4)

Dr. Hecker was less willing to commit himself to a finding of at least five manual erasures than the expert panel had been. (Michael Hecker 'testimony, In re Grand Jury, Misc. 47-73, 5/13/74, 18-19; SRI Report, 3) The panel rejected the hypothesis that any of the magnetic marks suggesting manual operation could have been caused by a malfunctioning machine. (SRI Report, 3-4) Dr. Hecker was of the opinion that it was wrong to rule out conclusively the chance that the malfunctioning machine could have caused some of the indicia of manual operation. (SRI Report, 4; Michael Hecker testimony, In re Grand Jury, Misc. 47-73, 5/13/74, 18-19) Dr. Hecker stated that because the machine had broken down once during testing; and after a defective diode bridge rectifier was replaced, the distinctive buzz could no longer be reproduced. Dr. Hecker did not state that any of the indicia of manual operation were caused by the defect on the machine; he merely said that, in his opinion, this possibility could not be ruled out completely. (SRI Report, 4-5) However, Dr. Hecker remained convinced that some of the marks of the operations were caused by manual manipulation of the keyboard controls. Dr. Hecker stated that he was absolutely sure that three events associated with the 1812 minute gap were caused by manual operation of the keyboard controls and that he was practically certain that two other marks had been caused by manual operation of the keyboard controls. He testified on May 13, 1974 that he was willing to agree with the panel that at least five of the events on the 1812 minute buzz had been caused by manual operation of the machine. (Michael Hecker testimony, In re Grand Jury, Misc. 47-73, 5/13/74, 18–21)1

1 The Court received two reports obtained by Miss Woods' attorney that questioned the conclusions of the Panel, whose conclusions in substance had been confirmed by the Stanford Research Institute, expert for the counsel to the President. The Committee staff has obtained copies of these reports. The organizations submitting the reports are Home Service, Inc., a Magnavox sales and service center in Cleveland Heights, Ohio, dated May 24, 1974, and Dektor Counterintelligence and Security, Inc. in Springfield, Virginia, dated May 30, 1974. Neither organization examined the evidence tape or Uher 5000 recorder, or reviewed the experiments with the expert panel.