An Update on Light Based EVP Recordings

May 22, 2015

Last time I’ve written about ITC in general (and EVP in particular), I was thrilled by the possibilities that light-based recordings could open to paranormal research.

But some facts came to my attention meanwhile: waveforms provided as background noise seem to be more important than the carriers used. This is so because EVPs are probably modulated at the target device(s), and the transmission “medium” doesn’t seem to matter (i.e., background noise, unfortunately, must be present for good and intelligible trans-audio captures). Moreover, the modulations identified don’t look stable enough for our analog signals to carry the digital data needed.

Right now, these facts almost rule out the chance of employing digital protocols.

I’ve taken other wrong turns trying to improve trans-audio. It’s widely known that the problem of blind source separation (BSS) is very hard for computers to solve in every scenario (PCA/ICA methods don’t translate very well to EVP reality [0]). And BSS also got me stuck (it’s easy to fall into some traps when doing paranormal research; this is a field that is as complex as it is driven by sad hoaxes [1]).

I wrote in Paranormal Going Digital - An Intriguing Possibility:

background noise is a controversial subject (…) a) its proper “crafting” helps to improve trans-audio intelligibility, and b) its scientific neutrality is questionable (…) in my limited experience, even if adequate noises are carefully used, resulting recordings can become almost impossible to understand

Light offers nice possibilities, of course (such as more controlled experiments). But we still couple an annoying signal “behind” EVPs, making later analysis much harder. In this way, it seems that algorithmically handling paranormal audio is more important than replacing sound as carrier.

[0] - multi-channels help to leverage phase information in “cocktail party” problem solving; unfortunately, microphone-arrays are used as a control form in ITC experiments (e.g., it’s common for one mic to register an EVP, and another one, few inches apart, to capture nothing);
[1] - major instance debunked here?; others abound?;