I looked over some of Mr. Gentile's previous recordings under some digital audio editing equipment, and found a strange digital signature.
When looking at a word, or sequence of words digitally, we can actually pick out where a vowel or consonant begin and end. A consonant requires a stoppage of air, and is therefore a sharper, stronger signal (like a spike), where vowels are generally the continuation of air, and get a decreasing signal, or a much slower, smoother signal when beginning a word. So, we can easily detect where a consonant begins after a vowel. We can also interpret vowels by how far apart the consonants are without silence, and when a word begins with a vowel.
What I've found so far in Mr. Gentile's recordings is the near absence of consonants. Instead of spikes, the words in question are simply showing solid blocks of audio data. The only consonant found is at the begining of the sound, and there are no spikes thereafter to signify consonants, and no dropage in signal to signify a following vowel. This points to the explanation given by skoob earlier about the technology of the recorder.
Considering that there are no consonants or vowels to detect, I don't understand how he is interpreting words (as they would be unintelligible) other than to interpret the pitch, rhythm, and tone of the sound. Can anybody find a phrase he has detected that begins with a vowel so I can compare them?
Also finding a result of 1 in 50, especially on reel to reel (about the only non-digital recording device still used) does not seem particularly significant. A reel tape may have defects, a kink in the tape, or if the tape had been previously edited, that would all cause certain strange sounds. This all causes me to question why he gets the best results with lowest quality recorder? It seems very counter-intuitive.
