The compression algorithm used in mp3 files is a lossy one, essentially what is being done is that a lot of the subtle parts of the music which we don't notice much, is stripped-off in order to reduce the amount of digital data, Marcel could probably explain it more clearly.
A typical audio CD (compact disc) has a dynamic range of about 105dB or so, that is, from the softest sound that can be heard to the loudest sound, there's a range of about 105dB, on the other hand, a vinyl disc has about a 30dB dynamic range, which is a lot less than a CD, reason being is that the fine grooves cut into the surface of the vinyl disc can only accommodate so much before the stylus starts jumping out of the groove (this is called skipping), the vinyl disc system also tends to produce a characteristic type of distortion which is similar to what the human ear naturally produces by itself, so as a result, a vinyl disc will tend to sound warmer than a CD even though the CD system is technically superior (supposedly).
In digital audio, there is something called the "Nyquist Theorem" which basically states that to accurately reproduce the highest frequency in the frequency response of a digital audio system, the sampling frequency has to be at least twice the highest frequency that is likely to be reproduced.
https://blog.telegeography.com/what-...does-it-matter
The thing to note is that there is no "perfect" audio media system, each one has it's good points and flaws, when CDs first became available to the paying public, they were regarded as perfect and indestructible, it wasn't too long after that when we started seeing the flaws that proved that CDs were far from indestructible, they were prone to skipping due to circumferential scratches on the surface of the CD, the digital encoding of the information on the CD also included an error-correction system that could cope with loss of data due to a 4mm hole drilled through the CD, or a radial scratch going from the centre of the CD outwards.