E5 E14 E21 G5 V5

This sounds very much like E14!
Click for audio.
From the UFO book  "Mysteries of Time and Space" by Brad Steiger. 
Thanks to Helen D.

German TV programme. A reporter receives the E5 Counting Station signals outside the U.S. Frankfurt site. Thanks to James C.

US-Spionage in Deutschland (Phoenix 2003)
Part 1 VIDEO

US-Spionage in Deutschland (Phoenix 2003) Part 2 VIDEO

V5 "4F CIA Counting Station" - including footstep call for " 123"  9075Khz 0000 UTC 15 APR 1984

V5 CIA Counting Station 

G05 (The German Counting Station), von ca. 1980, A3E, QSO by OM Karl-Heinz 

Trip hop musician Moby has used my E5 Counting station clip "1234567890, 257" in his track of the same name, 257.zero. 
You can watch the clip on youtube here:

Moby - 257.zero  NEW

E5 American YL   
E21 "British" 5F YL*  (see below for more sound clips)
E21 "British" 5F YL (2)
G5 German 1 YL 
G5A German 2  YL  
G5 German 3 YL  
E21A American 4F YL  
V5A Spanish 4F YL  
E14 YL/EE "Count Control" with jamming.
E14 YL/EE "Count Control" 2
G5 Counting station mixing with S5 OLX  
V1 Skylark mixing with E5 Counting station
Clip from Peter H. in Germany from 1988.  

*Clip courtesy Daniel Stadermann
Rare 4F transmission wma
Another 4F transmission wma poor quality (Gary in Wales)
CIA German YL  (Leen in NL) via spook007
CIA Jammer  (Leen in NL) via spook007
CIA YL 4F (Leen in NL) via spook007

Following comments courtesy Billy in Ohio, USA.
I have picked up E5--the standard American English Cynthia--many times here in Ohio, so I'm quite familiar with the "sound" of the Counting Station (and it's foreign language variants via recordings). Not just the format, but the actual tonality of the voices and the beeps, and the timing of these elements. I've been intrigued with the so-called British variant of E5 ever since I first heard the recordings of it (i.e. on your website and on "The Conet Project") and a couple of things have occurred to me. First of all, the voice does not strike me as sounding British at all...or female for that matter! It has a very pronounced American accent to my ears, rather like the old Bell Telephone System recordings for disconnected numbers (the "four" and "five" especially). As for the gender, it is obviously a sped up or otherwise modified male voice. I slowed down your sound clip from 16000 Hz to 12000 Hz sample rate, and it sounds much more natural--and male--in speed and pitch. Although the format is the same as the standard version, the sound is almost completely different to my ears. The tonality and timing of the ten beeps, for example, are nothing like the usual x5 transmissions. The spaces between the group counts and the 4FGs or 3/2 FGs are much longer.  What I'm getting at is that this one variant, which I understand hasn't been heard in years, is unique among the Counting Station family that I do wonder if it might in fact have been an unrelated station copying the E5 format.

Thanks to Billy in Ohio for some more fantastic revelations regarding the "British" E21 version. He has finally laid to rest the myth that E21 is a British version of the E5 "Cynthia" station. Here is his conclusion including two superb recordings.
  I have come across something you're really going to like. But how it came into my possession is an interesting story. And YOU, Simon, are indirectly involved in the story. It started with a Wikipedia entry on languages and accents, which led me to a website containing many audio samples of people from around the globe speaking English.

I passed this website along to a friend of mine who himself has a remarkable speaking voice, knowing that he would be fascinated. So tonight we were discussing accents during an AOL IM chat session, and we got to talking about various British accents. I mentioned that I knew a fellow in England who speaks with a quite interesting English accent, which I wasn't able to pin down based on the samples on this website (note: my Hull accent! -SM). Since he was curious, I pointed him at a certain streaming audio file of a certain NPR radio program devoted to number stations, as that friendly fellow had been interviewed for that program.
See http://www.npr.org/programs/lnfsound/stories/000526.stories.html

Now, this friend of mine's ears perked up when he listened to this radio program about radio "programs" because he himself had once taped a number station, and he knew nothing about it except the voice. He was familiar with the voice. See, his main interest was telephones, and he had amassed a huge number of telephone recordings over his years of being a "phone phreak" during the 1970s. So he told me that this number station had used the same exact voice as a particular type of telephone-company automatic announcement system. I'd never heard any number station that sounded anything like a phone company recording I recognized, but I offered to try to identify it for him, being fairly well-versed in the sound of number stations. First he sent a sample reel of the voice in the telephone context, and then his recording of the number station (identified by him on the tape as being recorded April 20, 1981).

Damned if it wasn't E21, the mysterious 4F "British" version of The Counting Station! As you may remember, I once remarked that the voice didn't sound British to my ears at all, but oddly, it DID sound like an American *male* voice sped up to sound female. That was on the familiar recording available on your site and on The Conet Project, which in truth isn't the clearest recording. Well, here are TWO reasonably high quality recordings that definitively put to rest the nationality and gender of the voice of E21. It is very clearly a woman speaking in an *American* accent, but obviously it is completely different than the familiar "Cynthia" voice.

My friend had some information on the voice machine used. It was a system manufactured by the Cognitronics company of New Haven, Connecticut and used mainly for telephone intercept messages (ie "the number you have reached xxx-xxxx has been disconnected," etc.) in the '60s and '70s. This machine used announcements recorded on optical soundtracks rather than magnetic media, which gives the Cognitronics machine a distinctive sound. Apparently they used the same woman to read all their announcements. (The latter half of the file cognicomparison2.rm demonstrates the optical version of the voice.)

This ancient technology may explain why E21 has such a different "sound" from the familiar E5. Besides the obvious differences in the voices and sounds, the cadences and timing of the digits, beeps, etc. and perhaps even the number of digits per figure, could be dependent on the design of the announcement machine, and it could be that E21 is actually an older version or even predecessor of E5, using an old-fashioned technology which "Cynthia" superceded or gradually phased out. The fact that my friend's recording dates all the way back to 1981 makes this quite possible, especially if the familiar "Cynthia" we all know began appearing later. (Of course, with all the radio noise I can't say with complete certainty that E21 uses the optical recordings, but it sure sounds like it).

And there you have it, Simon. I hope this message and accompanying listening material is as fascinating to you and the numbers readers as it was to me.

Btw, here is the site on accents that got this whole thing started:


One more note: I believe my friend was living somewhere in the southern US at the time, either North Carolina or Georgia, which of course would have put him very close to Warrenton and accounted for the excellent audio fidelity of the voice. According to his tape he was receiving E21 between 9.7 and 10.0 MHz around 1500 UTC. Maybe this could be a vintage sked entry!

Here are Billy's recordings:

1) Cognitronics company of New Haven, USA telephone announcing machine.  
 Announcer rm

2) Rare E21 clip recorded near Warrenton TX site, USA at 1017h 20 APR 1981.
 E21 recording wma 
 E21 recording rm

Written by P.S. in Saffron Walden with additional material from ENIGMA readers.

INTRODUCTION: The newcomer to the numbers scene soon comes to recognise certain voice stations which are heard on a regular basis. One of the most distinctive of these is the English language counting station, which has a female voice with a distinct American accent speaking five figure groups. She has been around certainly since the 1970s at least; she can he heard on any day of the week; she has been noted in the early morning, around midday, at various times in the afternoon, and during the evening

I first became aware of this particular lady towards the end of 1990; I was not all that interested in numbers stations then, being more a fan of the radioteletype transmissions from the various news agencies which were still using HF RTTY. As I tuned around looking for the characteristic frequency shift carrier of teletype in full flow I would often come across this American lady who seemed to speak only groups of numbers The transmissions were usually very strong, often almost as strong as and in a few cases stronger than many AM stations in the shortwave broadcast bands. Since she seemed to be on the air every time I tuned around in the evenings, and it was clear there was a definite schedule with regard to the day of the week and the time of day, so I decided to keep a log of her activities. I have continued to take an interest in her ever since and my observations form the basis of this article.

FORMAT: For those not familiar with the lady in question, a description of the format might be useful. Transmissions begin on the hour, her time-keeping is usually pretty good, starting within a second or two of the hour indicated by a watch set up against the Greenwich "pips", although in the past she has been noted on a few rare occasions starting up to half a minute or so early or late. She speaks a three figure call which is repeated three times followed by "1234567890"; this continues for ten minutes. At ten minutes past the hour ten one-second bursts of audio tone are sent and she says "count" followed by a two or three digit number; this is then spoken again as is the number of five figure groups containing the encoded message which will follow. The message is then spoken, and there is a distinct pause between the third and fourth figure 50 the groups might more correctly be described as "three plus two" figure groups rather than five figure. This done, she says "repeat", followed by 'count" and its number, again spoken twice, and the message is spoken a second time, finally finishing up with "end". (* nearly always - Ed)

MODE: The mode of transmission is a little unusual; it is upper sideband with carrier, or to put it another way AM but with the lower sideband suppressed whichever way you want to look at it, it can be copied quite well on an AM receiver, provided the signal is reasonably strong, but sounds much better with the receiver in USB mode especially if the signal is weak or if there is co channel interference as is often the case these days. Once the transmission has ended, the carrier usually stays on for a few minutes at least; this has a characteristic background noise which is difficult to describe but is very distinctive and Soon comes to be associated with the American accented lady. The engine noise does not seem to be quite as common as was the case a few years ago hut carriers are still noisy. The use of USB is becoming more common.

FREQUENCIES; A variety of frequencies used by this lady over the years ranges from 4007 kHz at the lower end of the HF part of the spectrum to 29790 kHz almost VHF - noted in use on Sunday afternoons during the winter of 1992. At times the choice of some of the frequencies used is rather strange; for example at the time of writing a couple of her favourite channels are inside internationally agreed broadcast bands with consequent QRM from strong broadcast stations making copy extremely difficult for the numbers enthusiast, the use of such frequencies perhaps suggests that the intended recipient may be outside the European target area.

Perhaps the strangest choice of frequencies was also during late 1991 and early 1992 when she was noted using a couple of frequencies in the 27MHz hand resulting in mutual QRM with the Spanish and Italian kilowatt CBers which infest that part of the dial when conditions are right. In order to ensure the message gets through, usually two - and on a few rare occasions three or even four - frequencies running in parallel are employed, which are typically two or three MegaHertz apart, and even where we are only aware of one there is no doubt another one somewhere which we have not been clever enough to find ! A compilation of all the frequencies used by this lady gives the last seven years gives a total not far off 100.

AGENCY: So perhaps at this point we might take a few moments to consider on whose behalf this transatlantic temptress works so hard. Perhaps there is clue to the answer to this riddle in the nickname by which she is affectionately known to her followers; whereas she is referred to in the ENIGMA Newsletter as "English language Counting Station E5, she is known to her followers by the much more romantic sounding name of CYNTHIA Why ? Well, perhaps because her voice is sampled or synthesised by computer techniques, but if we take the first letter of her name together with the last two we have the initials of the intelligence organisation who are her employers, and who have their head office in Langley, Virginia, U.S.A this is not to say that when we hear her she comes to us directly from a site up the Potomac River from Washington D.C.; to be heard as strongly as she is on this side of the Atlantic she must be relayed from transmitters in Europe located in countries friendly to the United States.

LOCATIONS: The whole subject of transmitter locations is made more difficult with this particular station due to its world-wide coverage and reliance on host countries. We will start with the mainland United States and work towards Europe.
UNITED STATES - Warrenton, Virginia
- Japan
- Australia
- Diego Garcia
South Africa 
Ascension Is.
ENIGMA has learned that the CIA have now established facilities in South Africa which are 'classed' as their African HQ following the demise of the Liberian operations centre
EUROPE - England Barford St John, Oxfordshire
- Germany Frankfurt area
- Spain
- Portugal
- Greece - Tatoi and Nea Makri near Marathon
- Turkey
- Morocco

OTHER FORMATS ('CONTROL' TRANSMISSIONS): As well as the frequently heard 5 figure transmissions, our Cynthia has also appeared on the air in another form (E14). At one time she could be heard daily with transmissions of ten minutes duration in which she spoke a short message of four figure groups, usually two or three 4 figure groups, occasionally four or five - even six or seven have been noted. These 4 Figure transmissions went out several times a day, unlike standard messages these particular transmissions always started on the half hour The first one of which I became aware was in late 1992 which used to start at 17.30 UTC on two frequencies in parallel, 5205 arid 8560 kHz These were usually very strong signals, especially the 5215 outlet; the typically noisy carrier would be on some time before the start of the voice - I recall one Saturday afternoon when the carrier appeared on 5205 kHz a good two hours before transmission time *

At exactly 17.30 UTC ten short audio tones would be sent and Cynthia would go straight into her 4 figure groups; these would be repeated over and over for ten minutes and then she would stop. By the middle of 1994, a third frequency had been added to run in parallel with the existing two, 12285 kHz so we can assume these short 4 future messages must have had some particular importance to be given so much in the way of transmitting resources. Also at about this time there were 4 figure transmissions at 07.30 UTC and two separate 4 figures, each using two frequencies in parallel, at 13.30 UTC, with further 4 figures at 00.30 01.30, 06.30 & 07.30 UTC. Response from the United States indicated that the same format was also running in Spanish at 00.30 02.30, 10.30 & 18.30 UTC. however, both the English and Spanish language ("14) transmissions are no longer in existence having left the airwaves towards the end of 1995 - by which time the 17.30 UTC had moved to 11072//13465 kHz-no doubt having served their purpose, whatever that may have been.

 GERMAN LANGUAGE: It is worth recalling that at one time there was also a German language (G5) counting station which used the Cynthia format, and all the evidence was that this was operated by the same organisation. This was a rarer find than the American English version, and the German language version was regularly noted from the mid-1980's. The style was exactly the same, a three digit call spoken three times followed by 'eins zwo drei vier funf sechs seben acht neun null the German for '5" was pronounced "funf' in contrast to the 'funef' of number stations whose origins are more to the east, after ten minutes of this, ten audio tones and the word "gruppen' followed by the number of 5 figure groups to follow, this repeated and then into the message. In the ENIGMA sound archives we have tapes of a very short-lived version which used the word 'zwei' instead of 'zwo' and ended 'zero' instead of 'null'. The German language version of this station seemed to cease operations during 1995, another victim of the redundancies brought about by the end of the Cold War; my own last logging was in March of 1995 on 6780/9219 kHz, two frequencies which interestingly, are still used by Cynthia. Before leaving the subject of the German version it is worth recalling an event from early in 1995; on Sunday afternoons there was a well established slot for the German language version on 9070 kHz at 16.00 UTC; I checked it out on a couple of Sundays during January, however when I tuned in again on 19 February, the German YL had been replaced by Cynthia, and she turned up on Sundays at 16:00 ever since.

 SPANISH LANGUAGE: the CIA have not restricted themselves to just American English and German, a well established Spanish language network also exists, and is still active, (V5), although primarily these transmissions are not well received in Europe and are most likely sent over transmitters situated in North America for agents in the Central and South American region, an area in which the USA maintains considerable interest. We are primarily looking at European operations in this feature so I will not expand on the Spanish language activities at this time.

ENGLISH ACCENT: You will have also noted that we have used the term American English so far puzzled ? Well, there is also an English accent version of Cynthia, a recording of which exists in the ENIGMA sound archives. The station used the standard (E5) format but the voice used had a distinct English accent She was very rare catch indeed and was last heard calling '250' with a 102 group message on March 1995.

 DIGITAL TRANSMISSIONS: Voices are not the only sounds emitted by Cynthia's transmitters, A number of monitors have noted the presence of data signals close to, or borne on the same carrier as that transmitted by the Counting Stations. The two data signals heard thus far are as follows:

2400 bps PSK (Phase Shift Keyed) System 109.8 bd synchronous FSK System

Neither of these systems can be decoded with even high-grade amateur telegraphy analysis equipment. In general, the signals can be heard as follows. The Counting Station appears on frequency (often up to an hour before the voice transmission time) and sends carrier plus the usual noise associated with the station when idle. At about 20 or 40 minutes before the voice transmission, the noise will stop to be replaced by the fast P5K signal, which has a very harsh, wideband, rushing white noise" sound. The P5K signal is only present for a few minutes and is offset from the Counting Station's carrier by about 2kHz.

About five minutes before voice transmission time, the PSK signal stops, to be replaced by a plain carrier. The voice transmission then follows. Often, at the conclusion of the voice transmission, the 1 09.Bbd FSK signal will appear, again only on-air for a few minutes.

What can we infer from this behaviour? Almost certainly, the Counting Stations transmit messages (perhaps even the same message) to different grades of users. Those with the highest security risk resorting to pen and paper using the voice transmissions. The middle grade users have the decoder for the 109.8 bd signal and the lowest grade using the P8K signal - such as embassies who need to monitor messages to certain agents.

 JAMMING. If there are those amongst us who like to take an interest in Cynthia's activities, then there have been times when others have put a lot of effort into trying to prevent her from being heard at all, by which I mean the situation which existed a few years ago when many of the transmissions were subjected to very effective jamming. This was a regular occurrence when I first became an enthusiast during the early nineties. It took the form of a carrier which was swept slightly either side of the channel which produced a characteristic kind of interference which usually made copy pretty well impossible - something similar can still be heard on some of the "Lincolnshire Poacher (E3) transmissions. Sometimes those controlling the jammers would allow the ten-minute "call-up" to progress and then switch the jammers on just before the 5 figures were due to start. Often during a long transmission, it was quite common to hear the jammers go oft after around 25 minutes or so and then come back on again, no doubt the operators having taken time out to listen to see if Cynthia was still there and on finding that she was, turning the jammer back on again.

The jammers never quite got the hang of call-up '383'. This was a most peculiar call which had the standard 10 minute preamble, but as the last tone sounded, abruptly went off air and never sent any messages. Jamming would commence during the call-up and continue until the Cynthia was checked and found to be absent. The (E14) 4 figure 'control' messages also attracted the attention of the jammers who did seem to realise that these particular transmissions were indeed only 10 minutes in length.

The occasions on which jamming was observed gradually became less frequent and had ceased entirely by the middle of 1994. However, after having been absent for some time the jammers returned briefly during the late summer of 1996. On Saturday 3rd August at 20.00 UTC. on 7746/19160 kHz a jammer came on just after the ten minute warm up had finished, and again on Monday 5th August at 20.00 on the same frequencies. This 20.00 Saturday and Monday slot was subjected to the jammers a couple more times during August 1996 since when. happily nothing further has been heard of them. As far as I am aware, no other Cynthia transmissions were given the jammer treatment at that time. Who was spending so much time and transmitter power in operating the jammers ? It must be a country at odds with the foreign policy aims of the USA: with the demise of the Soviet Union, that leaves a small number of Middle Eastern countries, and in particular Iraq. Perhaps Saddam Hussein's merry men had a sneaking suspicion that the messages were directed at CIA agents operating inside Iraq and decided to try and block the route by which they received their instructions.

MISTAKES: Cynthia is noted for her reliability: mistakes and failures during a transmission are few and far between, but they do rarely occur. When something goes wrong with the audio content-with the transmitter still putting out RF - the voice is replaced by an audio pulsed tone to give a "rapid-dash" effect at the rate of about three per second: this remains on until the voice returns, the problem no doubt having been fixed. Such an occurrence was noted on Saturday 22nd June 1996 at 20.00 UTC on 7746/19160 kHz; the transmission failed during the tar minute call-up while Cynthia was calling "514" with about a minute to go before the 5 figures were due. The tones came on and remained on for about three minutes, the voice returning at 20.12 with "count 212" and the transmission proceeding as normal. On Saturday 9th November, a transmission at 16.00 UTC on 10529 kHz went off at 16.04: again tones came on with the voice returning at 16.06. Here the procedure used was "Go back to the beginning and start again" because the call-up continued until 16.16 UTC. Strangest of all was the transmission observed on Wednesday 24th April 1996 at 13.03 UTC on 8116 kHz where there was a Cynthia transmission already into her 5 figures - when she would be expected to be still in the call-up until 13.10. The "repeat' followed by count 196" came at 13.05 UTC, with "end" at 13.23 which means that if the transmissions followed the usual format it must have started at about 12.47. This timing has only been noted very occasionally.

 TEST TRANSMISSIONS: On rare occasions transmitter tests have been noted on up to 13 frequencies in parallel, these consist of a test tone sent for several hours on a wide spread of Cynthia frequencies. The purpose of these tests is unknown, but certainly gives an indication of the resources available.

SCHEDULES As mentioned earlier, Cynthia works to a definite schedule, that is to say she turns up at the same time on the same day of the weak on the same frequency these schedules change from time to time. Some of them such as the Sunday 16.00 UTC on 9070 kHz have been around for years while others will only exist for a few weeks. One of the shortest observed schedules was on Saturday and Sundays at 0900 UTC which is a somewhat unusual time of day for Cynthia put in an appearance- in the summer of 1996 on 9070-16790  kHz. This was first noted On Saturday 2Oth July and last heard on Sunday 18th August, so lasted a month. On May 1995 it was noticed that there was a 5 figure transmission at the same time of day on every day of the week, at 19.00 UTC on 5153 kHz; this is the only time a Cynthia 5 figure has been observed turning up at the same time and the same place every day. This daily airing took place all through the summer of 1995 until the end of August. Other patterns have included transmissions which are heard to repeat on the same day and on the same frequencies one hour after the test airing.