4. Comment from Priyom

The researcher at Priyom.org who traced the Interférences article and contacted its author has submitted the following comments:

We at priyom.org have acquired a copy of the Interférences magazine issue that you mention; we were also able to interview the author of the article in it mentioning the Tyrolean Music Station:

http://priyom.org/blog/long-lost-evidence-shows-g01-run-by-french-sdece

There are also a number of points that you bring up about which I'd like to put forward some comments. You mention that the amateurish nature of the transmission suggests it wouldn't be a professional intelligence agency; I disagree with this. Most numbers stations fall to occasional operator errors, but this is a recurrent characteristic for some of them - the worst being E25 and HM01.

E25 is very erratic and rife with mistakes. They used musical introductions played on tape, and once the tape was played in fast forward on the air - something the TMS was known for too:

http://priyom.org/media/70260/e25-9450am-20140213-1215z-msg-bygwraspe.mp3

HM01 starts at the wrong time, forgets re-tuning to the new frequency between transmissions, and crosses its several audio feeds, sometimes even accidentally splicing Radio Habana Cuba onto the air; while G06 "splices" mains hum to the transmitter.

In the modern age, many if not most numbers stations have leaked Windows system sounds onto the air, E25 and HM01 being again the worst offenders with regular mistakes. G06 very regularly leaks Windows XP shutdown sounds after transmissions, but it has on occasion had sessions completely filled with error sounds. Even E11 and V24 have also leaked Windows XP shutdown sounds in the past. And sometimes the operator is just bored and plays Windows games at the operations computer, as can be heard on the air: E25 did this, and G06 once launched the Space Cadet pinball.

Some stations have occasionally resorted to a live voice operator as fallback instead of their pre-recorded voice samples; E25 while apparently dealing with technical issues, or E17z as training transmissions in preparation for such a contingency. Others like S06s have leaked unintended live voice operator chatter while setting up the transmitter.

Some stations have had transmissions using the wrong set of voice samples, in another language than was supposed to, like E06/G06, E11/S11a or again S06s; presumably the operator mistakenly choosing the wrong language at the console. Even digital numbers stations aren't spared by this: for example FSK stations F01 and F06 using the wrong modem or wrong FSK shift.

So the theory that numbers stations operations rife with technical mistakes would be incompatible with a professional intelligence operation is simply contrary to my experience. You also have to consider that bigger agencies transmit from several facilities, in different locations, staffed by different crews; and there can be a significant difference in the characteristics and quality of operations depending on this.

You also make observations about noise audible during silence due automatic gain control; AGC artifacts in fact a very frequent phenomenon of some numbers stations and shortwave oddities. I can in fact recognize some numbers stations solely by their characteristic transmitter noise audible during silence. This kind of artifact can also be heard between buzzes of the buzzer on 4625 kHz, and is at the root of the "XC crackle oddity". So that's another thing that wouldn't really be significant one way or another.

As for crosstalk between the TMS and East German MW broadcasts from Burg, obviously I can't comment on those particular reported instances, but I can say that this idea is not implausible. I mentioned above Radio Habana Cuba being mistakenly switched to the Cuban HM01's transmitter:

http://priyom.org/media/188475/hm01_rhc_overlay.ogg

Although in that case, it's more a switching mistake than crosstalk. However I can also put forward a genuine case of crosstalk, with a broadcast station positively embedded into V07's transmitted signal rather than a receiver artifact:

http://priyom.org/blog/priyom-pinpoints-v07-transmissions-to-khabarovsk

Finally, you mention as supporting arguments that: "The transmissions were heard only at weekends and for short periods, suggesting limited time or resources, or a need to avoid detection, while most numbers stations were on air constantly." And: "The other unusual feature common to both the TMS and the G20 station is the extensive use of music." Are you sure about these? 20th-century stations are not the ones I have most experience with, but short and occasional transmissions don't strike me as outstanding, neither does the use of a musical introduction - especially in that era, nor the use of various pieces of music: V24 does this for example.

Your article is a great show of critical work though, it's amazing and was very helpful to me in my part of the research! Thanks!

==========

About https://tyrolean-music-station.blogspot.com/p/themysterious-mihaela-enasoiu.html

Mihaela Enasoiu seems to be indeed quite knowledgeable about some details of past broadcasts, including some that I haven't seen in many other sources apart from the article in Interférences, which she apparently has't read. But when it comes to theories behind those broadcasts, I can't share the same appreciation.

Her assertion that the Swedish Rhapsody was not run by Polish intelligence is contradicted by evidence in the form of declassified documents from the Polish Institute of National Remembrance where the Polish Security Service describes it as their station:

http://numbersoddities.nl/N&O-200.pdf#page=3

But this is not my find, so it could always be argued that this document was sent to N&O by the CIA to plant disinformation.

As for the SDECE Morse transmissions sent from Les Alluets-le-Roi, I know of two French stations sending that kind of Morse messages: M51 and M23. M51 was operated by the French military, and is tied to its Morse lessons network FAV22. It is transmitted from Vernon, west of Paris, from an antenna field on top of the hill on the northern shore of the city - the sister control and RX site being in Favières.

The location she refers to hosts a well-known SDECE then DGSE SIGINT site, also west of Paris. There are actually two sister sites there again, on two sides of a field along a road connecting the neighboring towns of Les Alluets-le-Roi and Feucherolles, a few kilometers away from each other. And assuming that one is for TX and the other for RX, well the clues don't necessarily point to Les Alluets-le-Roi as the TX one.

Either way, M23 was triangulated to the west of Paris. Now I live in this area, and my personal research turned unable to confirm that either Vernon or Les Alluets-le-Roi / Feucherolles was the source of its transmissions. So there, unless she's referring to another station than M51 or M23, this claim sounds to me like a mere educated guess that was asserted without evidence and against mismatching observations.

She also makes mention of "the German-voice equivalent to M10c", G18 I presume (S10 family), "run by the StB (Czech)". I would also question that. We know from the Czech state themselves that the StB ran S05 / OLX:

http://priyom.org/blog/czech-intel-confirms-it-used-to-run-a-numbers-station
http://priyom.org/blog/priyom-releases-a-top-secret-document-about-numbers

We also know from Swedish intelligence that S10 was also from Czechoslovakia:

http://priyom.org/blog/swedish-security-service-tells-us-about-their-numbers-stations-experience

Since the StB was already running S05, it would make more sense if S10, and so M10 and G18, were rather run by another organization, like Czechoslovak military intelligence.

When it comes to the TMS, her comment seems to misunderstand the reference to the Interférences magazine, so she's apparently not aware of all the elements on this topic that she's making adamant claims about (and the article in Interférences does make a mention of the 1830z evening schedule, although nobody would have known that without reading the actual article). It seems that her main argument about the origins of the TMS is the announcer voice allegedly shared with G20 and/or G06. I already gave my opinion on this in my blog post:

http://priyom.org/blog/long-lost-evidence-shows-g01-run-by-french-sdece

In short, we have very little actual recording material to analyze - too little to draw relevant conclusions.

Mihaela Enasoiu is interesting, but rather than chasing claims that remain basically unrefutable without hearing the full story behind them, perhaps it would be best if you contacted her to directly discuss this and confront theories.




-- 
(name redacted)
with priyom.org 



Comment received by e-mail 15 March 2020