Before I explain how loudness works in mastering, let me first say that when I master an album, I always make sure that the average loudness of the mastering is competitive with other releases in the same genre, but I try not to over compress or limit tracks so the music retains the right amount of punch and sounds good when cranked up loud as well as being clear and high fidelity in headphones. If you have a preference in this regard, just let me know. And if you have heard recently that digital-only releases should be mastered at -14 LUFS for streaming services, see the second to last paragraph of this article.
The “loudness war” is over now due to the way streaming services treat average volume, which I discuss below, so it’s not necessary to make one’s album loud just for the sake of it. It’s more important to make it balanced, punchy and clear. Metal albums, for example, will always be mastered somewhat “loud” while folk and world music albums will not be treated in the same way.
When artists and engineers speak about the “loudness” of a song or an album, they are referring to the “perceived” or “average” loudness when compared to other albums. In other words, if you quickly switch between 5 different albums on your CD or music player, you will find that there is a difference in perceived loudness between those albums even if you don’t adjust the volume of your stereo. That’s because each album was mixed and mastered totally differently and this is normal.
The reason why mastering affects the loudness of a mix is because of the use of compression which limits the dynamic range of music and this gives the impression that the music is louder because the peaks are pushed down, allowing the average level to rise. The more these peaks are compressed, the higher the average loudness becomes.
There are very good practical reasons for limiting the dynamic range of music with compression (and limiting) during mastering and it is part of the way modern music sounds, but the loudness war was very damaging to music and had no practical purpose.
What was the loudness war? As recording and mastering technology progressed over the years, it became easier to compress music to make it “louder”. By the early 2000’s record labels started pushing mastering engineers to make their releases louder. The entire situation was always silly because no music fan spends their time switching between CDs to discover which one is louder. If a listener wants the music louder, they just turn up the volume knob! Making an album louder “out of the box” is not a selling point for music fans. They don’t prefer music that has been mastered louder. But the labels pushed this for a while and then artists picked up on the fashion and for about 15 years, pop, hip-hop, rock and metal bands wanted their albums mastered “loud!”
The problem with this is that after a certain point, the music starts to distort and becomes tiring on the ears. Modern mastering technology allows us to make albums loud without “obvious” distortion, but when we listen closely to a loud album all the way through on good speakers or headphones, we realize that it’s not as clean as it should be and its fatiguing to the ears. Often older albums are re-mastered by record labels and sometimes they master them too loud and the fans notice this. They often prefer the older versions that are not mastered as loud.
Then in the late 2010’s when streaming services like Spotify and iTunes became the most popular way to listen to music, things changed. These platforms apply an algorithm that lowers the volume of loud tracks and raises the volume of quieter tracks so that the listener doesn’t have to adjust their volume as tracks from different artists are played back to back, similarly to the way radio stations process music with an Orban Optimod processor. Because of this, many albums mastered today are actually turned down on streaming platforms. If you want to see how much a mastered track will be turned up or down on a streaming platform, upload a track on the Loudness Penalty website to find out.
Although this trend is dying, still some artists request loud masters and some mastering and mixing engineers still think people want this and they send out test masters that are far too loud (dynamics crushed). If you are evaluating test masters of your music created by multiple mastering engineers, try not to be flattered if one master is louder than another. Adjust the volume control so that they are each at the same level and then listen for fidelity. Try to listen in different stereos, at different volume levels and see which one has the attitude and tonal balance that suits you. Try to compare on headphones and listen for clarity.
Finally, there is a new rumor going around that mastering engineers should deliver masters set to -14 LUFS so that streaming services do not have to adjust the volume. This is not correct and is explained well in this article by Ozone. As is known by all seasoned mastering engineers, masters should be produced at the average level (LUFS) that suits the music best and the peak level (DBFS) should be set to -0.1—.1 to avoid digital “overs”. For metal music which is less dynamic than other styles, an average level of -6 to -8 LUFS is common. We master for archival purposes and let each playback service do what it does.
My advice is that if you are requesting a test master from a mastering engineer, tell them that you want your master to be relatively loud but that you prefer clarity and quality over max loudness. As for myself, I always deliver a master that is competitive in the genre it belongs to and the artist never has to worry about this subject or bring it up unless they want to. I got your back! :-)