Sound quality

Some audiophiles and sound technicians in the image of Neil Young denigrate this format that distorts significantly the sound 35 to such an extent that they never use 36 . Conversely, certain experiences including blind listening in the studio tend to show the difficulty for the human ear to distinguish a CD listening from a weakly compressed MP3 listening. Even more surprising, experiments like that of MusiClassics in 2009 37 classified the CD format in front of the MP3 320 kbit / s but behind the AAC 192 kbit / s and WMA 320 bit / s . The debate therefore remains open, and inthe majority of high compression uses , the loss of quality is often masked by the low quality of audio equipment: portable music players with headsets (also including mobile phones, audio players and tablets) as well as the on-board speakers of portable computers.

One of the causes of the a priori negative towards MP3 comes from the fact that at the beginning of its use by the general public (end of the 1990s), users encoded their audio CDs in MP3 most often with a bit rate of 128 kbit / s (the maximum in MP3 in Microsoft Windows XP). Nowadays (2015), it is generally accepted that an acceptable sound quality is obtained from 192 kbit / s. The speed at 192 kbit / s is often considered to be of good quality, that at 256 kbit / s of very good quality and that at 320 kbit / s of excellent quality 38 .

The use of a variable bit rate (VBR) allows better taking into account the complexity of the sound signal and significantly improves the quality of the result. A constant speed is only necessary for very specific use cases (streaming for example). In all cases, the sound quality of an MP3 file varies significantly depending on the encoding used (type of algorithm, compression rate, etc. ). Strong compressions (file at 64 or 128 kbit / s for example) strongly distort the audio signal and are more suited to voice (audio books) than to music.

The MP3 format has the drawback of not being suitable for encoding chained pieces, for example concert recordings or classical music. When playing, switching from one MP3 file to the next necessarily cuts off the sound for a few tenths of a second. This is due to the empty data ( padding ) located at the end of the file. The only way to really remove this silence is to modify the LAME MusicLength 39 tag which indicates the exact duration of the song by excluding ID3 tags at the end of the file.

Leave a comment