Extracting subliminals

[nakor]

I have noticed that many hypnosis files online have subliminals in them that are not easy heard simply by raising the volume. (EMG's are not hard to hear, but he has regular files anyway so there would be no point.) I am looking for tips on extracting subliminals, so I can know what they are saying. I have seen GoldWave suggested for this purpose. It probably can do it, but it is a complex program. Maybe all I need to know is which filters to be trying on GoldWave.

[Mallic]

I find that if you know the purpose of the file, it doesn't matter whether you here the words or not

[davelowe1977]

Nakor, some thoughts.

Extracting the voice component from a subliminal file is a non trivial task. Using filters in Goldwave, or some other program might help, but I doubt that you would be completely successful. The reason is that many sublims use a broad spectrum of noise (sometimes white noise) which mask the whole vocal spectrum - in this case what do you filter? You could try 'topping and tailing' the file by removing all sound below about 400Hz and above 4KHz. I still have my doubts however.

There is another case where you might strike lucky. This is when the noise is in stereo but the voice was recorded in mono identically on each track. By using noise reduction techniques (Cool Edit has this facility - not sure about Goldwave) you might be able to subtract the noise and leave only the vocals. It's not an easy technique, but is likely to be much more successful than filtering alone.

Hope this helps.