– The process involves mathematical models that are based on human hearing and largely, on the principle of masking. In essence, when an MP3 is created, sounds that you do not hear in the recording are thrown out of the file.

– This makes the file size considerably smaller than the original wav file from which it was generated. These processes are known as ‘lossy’ codecs, because once the compression takes place, the content cannot be recovered; it is impossible to fully reconstruct the original wav file working from the MP3.

– Lossless codecs, on the other hand, preserve much more of the fidelity, but afford lower compression rates (i.e. bigger file sizes); truly lossless codecs preserve all of the original content.