I have a some CDs that have hidden tracks that follow several minutes of silence. To give you an example, on Alanis Morissette's Jagged Little Pill: Track 13 contains an alternate take of "You Oughta Know" and, following silence, at 5:12 the a cappella "Your House."

I recently discovered that Nero Wave Editor has a feature that will detect and remove those long moments of silence, it will also insert track splits. This results in saving two files as opposed to one. So far I have had little problems with it, albeit I am not sure if I understand the detection settings.

I think I understand what "minimum pause length" and "minimum song length" are. The current (default?) settings are 2 seconds for pause length and 20 seconds for song length. I haven't really messed with those settings. What I am confused about is "threshold." I have noticed that, for the most part, I achieve the best results when I have it set at -80. According to an article that I read: If the silence threshold is too low and the track contains decoder artifacts, the software may not recognise [sic] some silences. Conversely, if the threshold is too high, the software may remove entire sections of quiet music at the beginning or end of a track. This, cf course, is unacceptable.

On the rare occasions I would actually use this feature, I would rather have the program not recognize silence than remove sections of quiet music. But I came across one CD which the software, for some reason, would not recognize the a lengthy silence. I'm not sure why, but I was able to resolve the problem by bumping up the threshold setting a few notches (-74, give or take).

I suppose the question is: What, in layman's terms, does threshold actually mean?

i wouldnt know the exact technical definition, but in this case this is simply a level where software will think that the signal should be treated as silence.

Here comes another definition about digital silence (is it called that way?), but you can google that as well.

Threshold is a loudness level that you choose, below which you consider that there is only silence. In this software, it's almost certainly measured in decibels below full scale (decibels are a logarithmic scale, describing a ratio of one power level to another). Full scale is 0 dB.

You'd probably be listening pretty loud to have sound at 83 dB SPL and louder still at 89 dB SPL, which are more than 80 dB above the threshold of hearing, so something around -80 dB from full scale is probably in the right ballpark.

If the sound dips below -80dB FS for less than 2 seconds (pause duration) it isn't cut out, but if it's longer than 2 seconds and the track is already at least 20 seconds in, it will be cut.

Some recordings, including tape and record transfers might well have noise above -80 dB FS.