As audio engineers, it is essential for us to train our ears (and brain) to listen to music at the best possible quality. For this article, I've decided to convert, to different audio formats, a song I've recently mastered, and compare all the exports using iZotope RX.
NOTE: This article/shoutout doesn't want to be an extensive treatise about music quality that requires much more than a visual shoutout. This article intends to hint at what happens to the music when converted to different audio formats.
Listening to music at high quality is fundamental as it helps us to develop a healthy internal reference. I cannot stress aspiring engineers enough about this...please listen to music at the best possible quality or you risk training yourself to a misaligned quality reference.
For this audio format shoutout, I've used a track I've recently mastered. The snippet of this analysis is a transition from an ambient and relatively quiet part to a busier part of the track.
Let's start...
This is the original .WAV mastered at 16 bit - 44.1 KHz
Let's start with lossless formats. This is the .AIFF version and as you can notice nothing changed from the .WAV
Another lossless format .FLAC. Again RX doesn't show any relevant difference comparing it with the .WAV
From this point, I started analyzing lossy format files.
This is an .OGG. This is the kind of file used by Spotify.
Streaming services have to reduce the size of the file, and consequently the quality of the audio, in order to reduce the buffering time.
Please notice how the image shows a noticeable loss of frequencies in the area around 20 KHz.
Here is the .M4A file at 320 Kbps. This format is used by iTunes and Apple Music. This kind of encoding is known as AAC. Again we can notice a loss of information on the upper range at about 20 KHz. Although RX shows little differences, I would say that the .OGG and .M4A are pretty similar.
According to the current guidelines this it also the format used by Sound Cloud.
Keep talking about lossy formats from this point I'm going to analyze.MP3s at different Kbps. In this example, we have an MP3 at 320 Kbs (the highest resolution for this format). RX shows a considerable loss of the top end.
MP3 at 256 Kbps. Compared to the previous MP3 the even lower quality is quite visible. (Look at the range between 15 KHz and 20 KHz).
What follows are other MP3s exports at different Kbps. Respectively 192, 160, and 128 Kbps
As you can notice, the lower the quality of the MP3 the less are the information we can find in the top end. This corresponds to a lighter file size.
