There is an interesting new site (Homepage DD1US) run by Matthias Bopp with all sorts of satellite signals. It contains a real sensation: a recording of "bleeps" from Luna-3, presumably on 39.986 MHz. The signals were recorded from Czech radio broadcasts by Mattias' friend Alois DL3PD and probably the recording has its source in Russia - an officially distributed recording. When I put it through my Adobe Audition software running in the "spectral view" mode, it struck me as looking very much like the signals described in the official Tass announcement above, but that it was running at half the speed given by Tass. If you speed it up by a factor of two it is very similar to the Tass description. The modulation is similar to that of the Voskhod beacons where the "off" period length is respiration and the intervals between the "off" periods is the pulse rate. In this case the "on" duration is one channel (multiplexed, it seems) and the interval between "on" pulses is another channel (also multiplexed). You may call these two channels the the "pulse duration channel" and the "pulse interval channel".
You can listen to the short 10 second recording (mp3 file here), and view it below:
I wrote to Matthias Bopp and asked "Do you think the "bleeps" from Lunik 3 could have been recorded at twice the speed than with which it was played back?" Matthias responded: "This is an interesting question which I cannot really answer. As the sound file also includes the voice of the Czech speaker and this sounds "normal" the change of the speed must have been changed already before recording the file which was in the 1960s ... the speakers voice on the record is not Alois but the original speaker from the national broadcast service ... "
Soviet sources also gave the total message length as 50.9 seconds, perhaps some kind of frame length. How this can be reconciled with a varying pulse interval is hard to fathom. If we use the maximum pulse interval as the worst case the frame could only contain 43 values in each "channel".
