(EST PUB DATE) ENHANCING THE WHITE HOUSE TAPES THOSE MISSING 18 MINUTES

• Document Type: FOIA
• Collection: FOIA Collection
• Document Number (FOIA) /ESDN (CREST): 01474442
• Release Decision: RIFPUB
• Original Classification: U
• Document Page Count: 4
• Document Creation Date: December 28, 2022
• Document Release Date: August 7, 2017
• Case Number: F-2007-00094
• Publication Date: January 1, 1974

Approved for Release: 2017/01/18 C01474442.
- � . �
Enhancing the White House Tapes
'Lapse
Missing P. Minutes
by George Toole
The headline on the Washington Post article reacl,"Tape,
Experts Say It May 13e Possible To Restore Gaps." It
was carried in the November 29 edition under the. by-
line of John Hanrahan, and reported on his interviews �
with me and another former employee of the Central
Intelligence Agency on the subject of restoring the.18-
minute gap found on one of the White House tapes. I
wouldn't (and didn't) call myself a "tape expert," but
I am "a former CIA computer specialist," and it re-
ported with perfect accuracy what I told the Post re-
porter. During the next 48 hours, I received a number
of inquiries from various .news media asking for ad-
ditional' information on how the missing 18 minutes
might be restored, and several invitations from talk
show producers to come on their programs and discuss
'the matter. I declined because, for one thing, there
� are the Espionage Acts and that famous CIA secrecy
agreement; in my opinion, I didn't violate either in my
..,statements to the Post, but it's difficult to keep from
slipping over the line during repeated, detailed ques-
tioning about a matter that is closely related to national
intelligence procedures. Secondly I did not want to
' give the impression that what I said might be feasible
could certainly by done by good sound recording
specialists. Finally I didn't want anyone to conflise the
� statements of a former CIA employee (I've been gone
nearly five years) with an official pronouncement by
the CIA or any. other government agency.
Why then� did I say anything in the first place? Be-
cause I doubted that anyone else who knew very much
about this specialized subject would be likely to speak
up. I know technical people. They tend.only to answer
the question they're asked and never volunteer any-
thing more. Last May the Senate Watergate Committee
interrogated James McCord, a former intelligence.tech-
nician. McCord answered slowly and precisely, choos-
ing his words carefully and never straying beyond the
limits of the question he had been asked. After several
days of this Senator Baker declared with some frustra-
tion that he felt McCord could tell the committee a
great deal more than he had, if only its members knew
the righrquestions to ask.
� When I heard Senator Baker's comment, I remem-
Gnoncr: OrronT.1., formerly with the CIA, is the author of
the .current tiove.:, An Agent on the Other Side.
LT
-tLA C
4
VkAA rLLIA,
� ' AAA-e-LvUlak
Ill
Approved for Release: 2017/01/18 C01474442
Approved for Release: 2017/01/18 C01474442
bered a meeting! attended some years ago at a govern-
ment installation near Washington. A senior govern-
ment official was trying to nail down the manager of a
certain technical program and get a delivery date for a
particular item that was needed. After several not very
responsive responses from the technical man, the of-
ficial said, "Look, I'm just trying to get a ballpark
figure. WOuld I be far from wrong if I told the* director
this will be ready in six months?" The technocrat re-
plied, "No, you wouldn't," and the official departed. I
turned to the manager and expressed some amazement
that he had promised delivery in six months on an
item we both knew would take a minimum of 18
months to develop. The man looked at me evenly and
said, "I didn't say that it would be ready in six
months; I agreed that he wouldn't be far from wrong
if .he told the director that it will be ready in six
months. If he tells the director it will be ready in 'six.
months, he won't be far from wrong, he will be very
close to wrong. In fact, he will be wrong."
'There wasn't even the hint of a twinkle in his eyes.
,
All I know 'about the White House tapes is what I
read in the newspapers: that they are "of poor quality,"
that portions are "nearly inaudible," and that an im-
portant 18 minutes seem to have been erased. I also
know that Judge Sirica appointed a panel of sound
recording specialists agreed upon by the White House
and the special Watergate prosecutor to answer such.
questions as: Were the tapes edited? Are they originals
copies? Was any one tape produced by two or more
different recorders? Was the .18-minute gap the result
of an accident? in other words, is the present condition
of the tapes due to deliberate attempts to obstruct
justice? ,
These are appropriate questions. But there is a more
i.important question: can the present condition of the
tapes be enhanced to the point where the inaudible
portions are audible and, perhaps, even the erased 18
minutes restored?
I was gratified to read in the newspapers the day
after the Post article that one of the court-appointed
panel had announced that it might be possible td
recOver the missing minutes. I was disappointed to
learn, when the panel made its preliminary report to
Judge Sirica on December 13, that .it had concluded
that the missing portion probably could not be re-
covered (Although this was "yet to be confirmed").
If the panel, finally concludes that the gap is beyond
restoration, I would not doubt that they had made
their best efforts and failed [As web� to press the re-
port of the teChnical experts is said to be imminent�
The Editors]. However I would like to offer the lay-
man a nontechnical explanation of enhancing tape
recordings.
To begin with there is a security rule observed by
the Department of Defense and other government
agencies handling information relating to national
security, which says that .if a magnetic tape recording
has highly classified information, that tape cannot be
declassified simply by recording unclassified informa-
tion over the classified data. Implicitly then, it might
be possible to recover the classified information even
after it had been "erased" by the process of ye-.
recording.
:Security people have developed, to very rigid.
government specifications, "degaussers,"* which are
powerful electromagnets that can obliterate even the
faintest latent trace of recorded information from a
tape. The use of. degaussers on computer tapes �is
standard security practice in the defense and intelli-
gence communities.
Some years ago a'government computer installation
that processes highly classified data installed a com-
puter which stored information on magnetic disk
devices. The surfaces of these disks are coated.with the
same kind of material as magnetic recording tape and,
except for their geometry which makes them more ef-
ficient, these devices' are in effect magnetic tapes.
The computer system, including the disks, was
rented, by the' government from the manufacturer,
with a view toward later upgrading to a different com-
puter system. The rental on the disk devices was $600
per month per disk device, and there were many disk
devices in the system. After a year or two the computer �
installation manager decided to mike his move up to a
larger system,. and wanted to return these rented
disk devices to the Manufacturer along with the rest
of the old equipment. There was one hitch, however:,
the security people had degaussers (or computer tapes,
but because of the different geometry of the disk de-
vices, there was nothing available to erase all the
classified information on the disks to their satisfac-
tion. The manager had no choice but to buy the disk
devices that the government had rented for so many
mohths, and store them away .in a vault. A staggering
amount of money was involved, and I am sure that
this particular security rule was given very careful
review before the decision was made that the disk
devices could not be turned back to the manufacturer.
So much for the possibility of restoring erAsed infor-
mation... Let's take a look at the how of doing it. In
principle it is a very simple matter. The tape is noth-
ing more than a itrip of plastic coated on one side with
a finely powdered metal that can easily be magnetized
and demagnetized. The vibrations in the air, i.e., the
sound, are transformed into electrical vibrations by a
microphone, and leave a magnetic trail in the powdered
metal.coating on the strip of tape as it moves past an
electromagnet. Wlwn the tape is again moved past the
electromagnet; the magnetic trail� the signal� repro-
*A reference to the basic unit of magnetic field stiength; the
Gauss, named after Johann Karl Friedrich GausS, a German
mathematician and, physicist.
oontinued
Approved for Release: 2017/01/18 C01474442
Approved for Release: 2017/01/18 C01474442
duces those electrical vibrations, an-d they' in- turn-
are amplified and transformed back into sound by a
loudspeaker.
The situation .becomes a little more complicated'
when we want to reuse the tape�to replace the
magnetic trail that's already there with a new one. �
If we simply repeated the process, we then would have
two trails on the tape, and when we played it; we
'would hear both the new and the original sound.
Therefore a tape recorder must first pass the tape over
another electromagnet to remove the original signal
before it puts a new magnetic trail on the tape. This
electromagnet is called an erase head, and the other �
electromagnet� the . one that actually puts the sound
� signal on the tape, and also picks up the sound signal
again when the tape is played back�is called the
record/playback head.
� Ir
Fo reasons that are too technical to go into here,'
the erase head does its job by recording a very high
frequency signal on the tape�something on the order
of 80,000 cycles per secOnd, which if the tape recorder
could transform into sound (and it can't), only bats
and dolphins could hear. Once the tape is past the
erase head, this new signal has rendered he old signal.
inaudible, i.e., so weak that were it moved past the
, record/playback head, it would not produce an electric
current that could be amplified into sound again by
conventional tape equipment. .
� Irthe designer of the tape recorder wanted to ensure
that the old signal was totally erased, he could do so
� by making the erase signal so powerful that'it removed
even the faintest ;atent trace of the old signal. De-
signers don't do this because the design requirement
� of an audio tape recorder's erase head is merely that it
render the original signal inaudible. Designing an
audio recorder with an erase head that removed every
trace of the original signal would be "overdesigning."
It would make the recorder more expensive to manu-
, facture, but no more valuable to the consumer. The
average consumer, that is.
If all this talk of latent magnetic signals and erase
heads has put off the nontechnically inclined, let me�
offer an analogy. Imagine the kind of long blackboard
that spans the front wall of a large classroom. Imagine
also that someone has written one long sentence from
the leftmost end of the blackboard, extending all the
way to the rightmost end. Call this sentence the orig-
inal signal. Now let's suppose that I wish to replace
this sentence with another one. To save time, I have
.an assistant with a chalk eraser walking a few feet
ahead of me; erasing the original signal from .the
blackboard. Then I follow along and write, the new
signal. Functionally, this is how 4 tape recorder erases
one sound from a tape and records another sound in
its place. But the chalk eraser does not always com-
pletely remove the writing from the blackboard. Often
ye can read the "latent signal" afterwards, and some-
times-we can make it out even after Someone has written
over it. (As a matter of fact, a photographer could
probably recover such a latent signal from a blackboard
by using infrared film, or' even ordinary black and
white film printed .on high contrast paper.) The
chances. of recovery would, depend on: 1) How hard
the person who wrote the original sentence-was press-
ing on the chalk (strength of original signal), 2) How �
hard my assistant was pressing on the chalk .eraser
.(strength of erase signal), and 3) The relationship of
the size and shape of my handwriting to the size and-
shape of the handwriting of the original sentence.
- This third factor needs some explanation. Visualize
� this classroom situation and imagine that the original
writing was in a small hand, With the words and let-
ters close together, and further imagine that my own
handwriting�which is covering the erased trace of
the first sentence� is large, with large spaces between
words and letters. Obviously the portions of the latent
signal lying in those large spaces have not been further
obliterated by the new handwriting.
Now to draw the 18-minute gap in the White House
tape into an analogy, we have the original. hand-
writing (the conVersation between President Nixon
and Mr. Haldeman regarding Watergate matters), the
assistant with the chalk eraser (the erase head on Rose
Mary Woods' Uher 5000 tape recorder), :Ind the new
handwriting (the strange hum or buzz that was re-
corded over the original conversation). Apart. from the
strength of the Uher erase signal (which, as previously
noted, is probably insufficient to completely obliterate,
the latent traces of all previously recorded signals), the
factors affecting the chance of recovering the missing
18 minutes are not yet known, i.e., how strong the
original signal was (some reports say that President
Nixon and Mr. Haldeman can be plainly heard just be-
fore the hum begins), and the shape of the mysterious
hum. We cannot, therefore, say how likely it may be
that the presidential conversation can be recovered,
but we can engage in some informed speculation.
The strength of the original signal would depend
ort how close Haldeman' and the President were stand-
ing to the microphones during the 18 minutes. They
may have walked around the room during that in-
terval, but we would expect the strength of the signal
at the beginning of the gap to be about the same 'as
it is in the unerased portion immediately preceding
it, and vice versa for the last minute or two of the
missing 18.. . .
The shape of the new signal�the mysterious buzz
or hum�is more difficult to guess at, but the use of
those words, rather. than "screech" or "whistle" to
describe it, is encouraging. A hum sometimes gets into
equipment by one route or another, and the source is
often the 60 cycles of the alternating current .that
powers the equipment. If the mysterious hum is a 60-
cycle a.c. then there is no problem. It is below
the 'normal range of human voices, which goes from
continued
Approved for Release: 2017/01/18 C01474442
Approved for Release: 2017/01/18 C01474442
85 cycles up to 1100 cycle's per second. In other words
the new handwriting left a lot of blank spaces for the
latent image of the old handwriting to show through.
Better yet, the mathematically simple shape of the 60-
cycle ac. signal makes it possible simply to strip it from
�the tape by Means of electronic filtering techniques.
In any event there is some.solid reason to hope that
the latent magnetic record of the President's chat with
Mr. Haldeman may yet be on the tape or, to use my
classroom analogy, the handwriting may still be on the
wall. But assuming the latent signal is found, how can
it be transformed into audible sound?. '
� If you take a fresh reel of tape from its carton, put it
on your recorder and play it, you will hear nothing�
except a hiss. The technical term for the hiss is "noise."
To an engineer any electronic communication medium
(including tape) contains just two things: signal and
noise. Signal is what you want to hear and noise is
what you don't. Nonetheless the' noise is always there,
random electrical impulses that carry no information.
Tape recorders (and other electronic communication
instruments) woric because it is possible to make the
:signal very much stronger than the noise, so When
you listen to a recording you don't usually hear the
noise .because the signal�the music or speech�is so
much louder.
The function of the erase head, as described earlier,
is to drive this signal down to the point where it is
almost no stronger than the noise. If a latent trace of
the presidential conversation remains, some means
will be necessary to "pull" the signal back up out of
.the noise. - .
The 'problem will, in fact, be very similar to that
encountered by NASA in receiving video signals from
deep space probes such as the Mariner series. Many
of us have seen the television pictures of the surface
of the planet Mars sent back by the Mariner spacecraft. �
A few of us have also seen some of these same pictures
in the form they were first received, before they had
been subjected to a computer process known as "image
enhancement." The unprocessed pictures were ex-
tremely faint, as though they had been taken through
a thick fog. Ohly vague outlines of Mars' surface
features could be distinguished, and some of the
pictures appeared at first glance to be completely blank.
After image enhancement, craters, chasms, mountain
ranges and a variety of other details were visible in
nearly perfect clarity. �
The television signal from Mariner, after &oSsing
millions of miles of interplanetary space, wa's faint. It
was not very much stronger than the noise that ac-
companied it as it was picked up by those giant
antennas in the Southwestern desert. In order to
transform it into a useful picture, some means� image
enhancement � had to be used to "pull" the signal
back up out of the noise. And those means were
dramatically effective.
How the trick was done is a little difficult to explain,
but I shall attempt to do so through another analogy:
the color vision test. The test consists of some 20-odd
black cards. Eaeh card bears a circle, withyt which
there are many dots of different colors. The subject is
shown the cards and is asked if he can see the outline
Of a number. The numeral is formed by a series of
dots which seem to stand out from the background .
and can be connected up by the eye. These dots are
"signal," and the other surrounding dots are "noise."
� Someone who is slightly color-blind will fail to
make out a few of these numbers because the retinas
of his eyes fail to distinguish among slightly dif-
ferent hues of the same color. He is not able to separate
the "signal" from the "noise." However if someone
else with perfect color vision marked each of the dots
forming the number with an "x," .the color-blind
person would soon be able to see the pattern for him-
self. The signal would have been enhanced and sep-
arated from the noise. This analogy takes some liber-
ties with the color vision test in order to convey only a
partial understanding of the signal enhancement proc-
ess, but it serves to illustrate the basic principle
involved.
Now, an audio signal and a video signal are alike in
that both are electronic signals carrying information.
But they are different in that the video signal is much
more complicated than the audio signal (the inside of
your television set is so much more complicated than
the inside of your radio). It would follow, then, that the
problem of enhancing a faint audio signal must be
a great deal simpler than that of enhancing a faint
video signal.. -
Where does all this leave us? In the matter of re-
.
covenng those lost 18 minutes we come up with a big
"maybe:" There is, at least, some substantial reason to
hope the missing conversation can be restored, al-
though it is an iffy proposition. Regarding the rest of
the tapes, which are "of poor quality," or "nearly in-
audible," I think there is every reason to expect that
the technical resources of our government are adequate
to the task of enhancing them to the point where they
are of good quality and completely audible.
When the Secret Service "bugged" the White House,
it did so with the full Ynowledge and cooperation of
the occupant, an unusual circumstance for an audio
surveillance operation. And yet the results were "of
poor quality," and "nearly inaudible." Yet this must
be an old, familiar problem to the intelligence com-
munity. And since they continue to use audio surveil-
lance so � much, one .might conclude that they have.
come up with some pretty good solutions to it. I am an
expert in neither sound recording nor signal en-
hancement, and if someone dumped the White House
tapes in my lap and said, "All right, enhance them!" I
couldn't do it. Not by myself. But I know the people I
would ask for assistance.
They work for the government.
Approved for Release: 2017/01/18 C01474442