A file-identifying fingerprint or marker is associated with a media/audio file after waveform values are extracted therefrom via certain data extraction and/or audio analysis routines. The extracted waveform values are statistically manipulated so as to render an array of statistical values for the analyzed media/audio file, which are then compiled to form the custom marker, and embedded with the metadata of the media/audio file thereby marking the media/audio file. The custom marker or fingerprint can be matched against a set of reference fingerprints (e.g., in a database) to determine the identity of the signal or the similarity between two signals. Because of the nature of the fingerprint extraction algorithm, it does not suffer from many of the problems that plague existing solutions, and as compared to such solutions it is fast, efficient, highly accurate, scalable, and robust.