Friday, October 13, 2017
Health attack? Or just scrambler?
Somebody has published a wave that is supposedly the Cuba "sonic attack" or hate-hoax. Not much info about how this was recorded, if in fact it was recorded.
It seems to have a 7100 cps peak, amplitude modulated randomly at a much slower near-syllabic rate. It doesn't bother my ears.
Spectrum of the original:
Piece of the waveform, showing the modulation in speech-fundamental range:
The MP3 sampling rate is 44k, so 7k is getting up into 'aliased' territory which makes the spectrum sort of dubious.
Slowing it down drastically so that the carrier is about 300 cps:
Slowed-down segment
This does bother my ears.
In this form you can hear that the modulation is basically a beat between two nearly identical carriers, shifting ever so slightly in freq.
Without any context I'd guess that this was a scrambler, using something like FSK plus a random component to create the audio equivalent of a one-time pad cipher.
Since we don't have any context, that's as far as I can go. This could have been a VLF radio signal emitted by the embassy, which was rectified and resonated accidentally by some metal surfaces. Or it could be an intentional audio signal.
7k is not an especially meaningful freq in terms of cochlear function or auditory processing. It's well above all the freqs we need or analyze for speech. Landline telephones cut off at 3k, and AM radio cuts off at 5k, without losing intelligibility. The only nearby landmark is 8k, which resonates at the first curve of the cochlea. Long-term exposure to 8k damages the hair cells more than other freqs, and a typical noise-induced hearing problem has a peak loss at 8k.
But this would be counterproductive if you were trying to drive the enemy crazy. The damage produced by long-term exposure to 8k would gradually make the victims more immune to the signal, not more sensitive to it.
If I had the job of driving Yanks crazy with audio I'd go the other way, down to 20 or 30 cps. A subsonic rumble makes you jumpy and edgy without being consciously heard. You don't know it's there until it stops, when you suddenly feel relieved and relaxed.
Piece of the waveform, showing the modulation in speech-fundamental range:
The MP3 sampling rate is 44k, so 7k is getting up into 'aliased' territory which makes the spectrum sort of dubious.
Slowing it down drastically so that the carrier is about 300 cps:
Slowed-down segment
This does bother my ears.
In this form you can hear that the modulation is basically a beat between two nearly identical carriers, shifting ever so slightly in freq.
Without any context I'd guess that this was a scrambler, using something like FSK plus a random component to create the audio equivalent of a one-time pad cipher.
Since we don't have any context, that's as far as I can go. This could have been a VLF radio signal emitted by the embassy, which was rectified and resonated accidentally by some metal surfaces. Or it could be an intentional audio signal.
7k is not an especially meaningful freq in terms of cochlear function or auditory processing. It's well above all the freqs we need or analyze for speech. Landline telephones cut off at 3k, and AM radio cuts off at 5k, without losing intelligibility. The only nearby landmark is 8k, which resonates at the first curve of the cochlea. Long-term exposure to 8k damages the hair cells more than other freqs, and a typical noise-induced hearing problem has a peak loss at 8k.
But this would be counterproductive if you were trying to drive the enemy crazy. The damage produced by long-term exposure to 8k would gradually make the victims more immune to the signal, not more sensitive to it.
If I had the job of driving Yanks crazy with audio I'd go the other way, down to 20 or 30 cps. A subsonic rumble makes you jumpy and edgy without being consciously heard. You don't know it's there until it stops, when you suddenly feel relieved and relaxed.Labels: skill-estate
¶ 10:46 AM
