The thing is, dynamic range compression and audio file compression are two entirely separate things. People often conflate the two by thinking that going from wav or flac to a lossy file format like mp3 or m4a means the track becomes more compressed dynamically, but that’s not the case at all. Essentially, an mp3 and a flac version of the same track will have the same dynamic range.
And yes, while audible artifacts can be a thing with very low bitrate lossy compression, once you get to128kbps with a modern lossy codec it becomes pretty much impossible to hear in a blind test. Hell, even 96kbps opus is pretty much audibly perfect for the vast majority of listeners.
In a distant past I liked to compare hires tracks with the normal ones. It turned out that they often used a different master with more dynamic range for the hires release, tricking the listener into thinking it sounded different because of the high bitrate and sampling frequency. The second step was to convert the high resolution track to standard 16 bit 44.1 kHz and do a/b testing to prove my point to friends.
The thing is, dynamic range compression and audio file compression are two entirely separate things. People often conflate the two by thinking that going from wav or flac to a lossy file format like mp3 or m4a means the track becomes more compressed dynamically, but that’s not the case at all. Essentially, an mp3 and a flac version of the same track will have the same dynamic range.
And yes, while audible artifacts can be a thing with very low bitrate lossy compression, once you get to128kbps with a modern lossy codec it becomes pretty much impossible to hear in a blind test. Hell, even 96kbps opus is pretty much audibly perfect for the vast majority of listeners.
In a distant past I liked to compare hires tracks with the normal ones. It turned out that they often used a different master with more dynamic range for the hires release, tricking the listener into thinking it sounded different because of the high bitrate and sampling frequency. The second step was to convert the high resolution track to standard 16 bit 44.1 kHz and do a/b testing to prove my point to friends.