It's all in the mind as this thread illustrates.
As an old hi-fi enthusiast I spent many years with the mindset that 'everything' matters, 'everything' makes a difference. Passing a music signal through a cable, a tone control circuit, anything really may (will?) have an effect on the sound so you really should avoid these possible issues.
The result, 'pure direct' switches, amps without tone controls and of course cables, the very stuff of being an 'audiophile'.
Then you learn a little about computers, data copied, moved from A to B all with 100% accuracy, it's how computers work. Compression can simply be a way of 'packaging' the data more carefully for storage, the data is untouched or it can be more complex with the data manipulated into even less space. But most importantly the reconstructed data is still 100% the same as the original and the computer checks that this is so, if it isn't it reports an error. This is lossless compression.
Unfortunatly, in the hi-fi world compression is also used to describe a process where data is compressed to a very large degree by deliberately throwing away some of the bits. What is actually thrown away is determined by the codec (be it mp3, aac or whatever), and they can and do sound different both from each other and from the uncompressed original.
Thus for hi-fi use, uncompressed or losslessly compressed files are the solution, lossy files may be very good in some cases, high bitrate aac for examble is extremely difficult to distinguish from the original, but as has been said before, why take the risk.