Mohammed,

This is probably a question for signal processing specialists. My background is in audio engineering and subjective codec testing, not signal processing, but I have some knowledge of signal processing so I'll try to tell you what I can.

To start with, audio encryption is not a term I am familiar with. Encryption usually indicates some form of security encoding to restrict access to information. I have not seen this done with audio. Usually if someone wants to retrict access to an audio file, they put it through a ZIP-type program and add password access.

When digitised audio is processed to reduce the data volume, it is said to be encoded rather than encrypted. In everyday English, encoding and encryption have very similar meanings. In signal processing however, they have quite different meanings. I am guessing that your question is really about audio encoding rather than audio encryption. If not, please ignore the rest of my answer.

The reason for measuring signal properties prior to encoding is to optimise the encoding parameters to get the best possible perceptual quality at a given bit rate. There are several types of encoding systems in use. Some work in the time domain. Some work in the frequency domain. The audio parameters needed to optimise algorithms will depend on the algorithm in use. For a detailed answer to your question, you will need to consult a text on signal processing. Introduction to Digital Audio Coding and Standards by Marina Bosi, Richard E. Goldberg 2003 is a good place to start. It is now 20 years old, and I don't know if there are updated editions, but it covers several codecs that are still widely used. If you want information on specific codecs, try the IEEE Journal of Speech and Signal Processing or the Journal of the Audio Engineering Society.

I hope this is of use.