Long story short: the Mp3 is perfectly fine for music listening – as long as you don’t abuse its compression.
Wanna know the scientific details about it? Keep reading.
The basics: the concept behind the Mp3 standard
Human senses far from the perfect system everybody seems to assume are.
First of all: they’re strongly limited by our brain’s capacity to interpret signals brought by our “sensors” – meaning that, although we might be receiving a given data, it is not assured that we will be able to elaborate it successfully. To be aware of it.
In this case: sounds captured by our ears.
And second: our “sensors”, for evolved they might be, are of organic nature – therefore: imperfect. Fragile. Prone to irreversible wear (aging, for example).
Mp3 based its concept upon these 2 factors: why record everything, when we’re able to perceive only a part of what we hear?
In other words: Mp3 used the studies upon human hearing limits in order to cut away everything we are not able to hear – not much differently than when, long time ago, the 44.1 Khz frequency become the golden standard for audio sampling (we can hear till 20Khz. And, due to Nyquist theorem, the right sampling frequency is at least the double of the one we need to record).
By the way: the “studies upon human hearing limits” is a science, and it is called psychoacustics – a wildly interesting argument, about which you can read more in this well crafted Wiki article: https://en.wikipedia.org/wiki/Psychoacoustics
At this point, we already have answered the question about why Mp3 is very good: the lost data is something we wouldn’t have heard anyway – therefore, although lost, it won’t impair audio quality.
The technical part: how Mp3 works
The Mp3 was created with a complete list of all of our “auditory system’s weaknesses”, in order to spot what we are likely to not to hear in an audio track.
The list is quite long and complex, and contains pretty interesting stuff like the auditory masking (about which you can read in this Wiki article: https://en.wikipedia.org/wiki/Auditory_masking). Who, in a very few words, is our inability to listen to two very similar sounds, both in frequency or time of happening.
Depending how hard we’d like to compress an audio track, the algorithm will erase more and more content. And, with very high compression rates, it will start erasing clearly audible data – although in a way that will optimize the most its loss.
The practical part: how to use Mp3 for music listening
First of all: the audio format is just a part of a longer “audio listening system” – in technical terms: an electroacoustic chain. All the elements involved in audio reproduction. Elements like cables, amplifiers, and speakers. And, long story short, with modern day’s excellent equipment, the only part of the electroacoustic chain who is still likely to ruin your listening experience is the final one: the speakers/headphones.
If you want professional quality, for high end listening purpose, 300KBPS is a good choice – although it’s something outside of 99,99999% Earth’s population possibilites: to hear music with this quality, you’re going to need a reference listening system (don’t know what it is? Read more about it here: http://lmkprod.com/guide-how-to-recognise-the-quality-of-an-audio-recording/ )
If you want maximum listening quality, for demanding genres like symphonic or Metal music, it is advisable to not go below 200KBPS.
If you need average listening quality, for less spectrally complex music (english for “music with quieter, less colorful instruments”) or casual listening situations (car, train, parties…), 150 KBPS is enough for your daily music listening.
The even more practical part: what’s better than Mp3
For cool it might be, Mp3 is still a lossy format: a little part of the auditory content will be irremediably lost – problem that we’ve solved with lossless algorithms. Nearly magical systems who, at the trade of higher CPU demand, will shrink your file with no quality loss.
I present you, dear reader, to the present and future of audio compression: the FLAC format.
FLAC compression format has quality level who matches an uncompressed format such as WAVE, but with a compression ratio up to 50% of its original size. Way bigger than the average Mp3, but with no quality loss whatsoever.
The price to pay for this? WAY more CPU demand. But not a big deal, with modern availability of calculation power.