20/12/2023
"What's the big deal with mastering anyways? All you did was make it louder, right?"
Well, yes... but also, no... not exactly...
While loudness is in fact one of the central and most important aspects of mastering, it's not exactly the defining trait of a professional sounding record. Moreover, there is such a thing as taking it too far and harming the quality of your master rather than improving it.
Some other important characteristics of a quality master are things like the tonal balance/frequency distribution, micro-dynamics, stereo imaging, and the media format you select to containerize all of the audio information (as well as metadata for digital formats). Often times these characteristics can go unnoticed due to how subtle they are by nature, but do not mistake their ambiguity for insignificance, as they are crucial to providing a pleasant and consistent experience to consumers of music and other digital audio.
Now that we've touched on the subtleties that contribute to a quality master, let's circle back to the elephant in the room – loudness. Why does it even matter how loud my song is? Can't people just turn up their volume?
Well, yes... they can... and in fact, this was a very annoying reality for listeners up until relatively recently. However, this is not an effective solution by any means; which is why we've found better ways to combat loudness inconsistency in today's music industry. It was not that long ago when you could be listening to a song which needed to be amplified by a significant amount to reach a comfortable volume, just to have your ear drums blown out by the next song that played. If you've never had the misfortune of experiencing this for yourself, I envy you.
So, how did we fix it? In August of 2010, the European Broadcasting Union (EBU) introduced the LUFS system (Loudness Units relative to Full Scale) as part of their R128 recommendation which, in a nut shell, is a standard for measuring audio loudness that was specifically designed to reflect the way human beings perceive it. The LUFS system was introduced with the intention of addressing issues related to varying loudness levels in broadcast and multimedia content.
The development of this new measuring system was revolutionary. Finally came the ability for streaming services to reliably compare the loudness of individual songs with one another in order to identify and correct inconsistencies, leading to a more pleasant experience for the listener. Advancements like this do not come without consequence however. In this case, audio engineers were forced to alter their mindset on what is too quiet and what is too loud.
You see, the late 80s and early 90s saw the advent of digital recording and mastering technologies; and with the rise of CDs, there was a perceived need to make recordings louder to stand out on the radio and in commercial environments. This practice intensified over time, creating an era that we now refer to as the "Loudness Wars." Awareness of the importance of dynamic range and overall sound quality was not widespread until more recent years.
With the introduction of the LUFS system and streaming service loudness normalization, audio engineers can now focus on making their masters 'sufficiently' loud while preserving the dynamics that give a good master it's punch and depth, rather than just making records as loud as they possibly can in order to be competitive. Awesome! But how do I, the audio engineer, know when something is "loud enough" and when it needs to be louder?
Different streaming services have different answers to this question, and different rules they follow to fix audio that doesn't align with their standards. Some services will only turn audio up if it's too quiet, some services will only turn it down if it's too loud, and some will do whichever of the two is necessary for a given master. This can be problematic if your master is overall too quiet, but has peak levels that reach the loudness limit of digital audio.
Let's use the song in this video as an example. The original version has an integrated loudness of -18.38LUFS, but the audio signal peaks at 0dBFS. If the original was uploaded to Spotify, which allows users to normalize audio up to -11LUFS, the peaks of the audio would exceed 0dBFS resulting in the use of limiting which can cause audible distortion.
On the other hand, the remastered version of the song has an integrated loudness of -11LUFS, which perfectly aligns with the maximum loudness standard Spotify allows it's users to select. What this means is that if Spotfiy's normalization algorithm is going to do anything to my master at all, it's going to turn it down rather than up. This is incredibly important, because it means that Spotify doesn't need to alter the dynamics of my master in order for it to align with their standards. What I upload is exactly what listeners hear instead of an altered, distorted version.
You may be thinking "oh great, I have to have separate masters for each service I want to distribute to? This doesn't seem like a good solution at all!" and if that is what you're thinking then you've missed the point! You do not need multiple masters. As long as your master is at least as loud as the standard for the loudest service you want to distribute it to, your music will never need to be altered in order for it to be normalized!
Mastering is a nuanced art, and loudness, though crucial, is just one piece of the puzzle. As we embrace advancements like the LUFS system, we find a balance between ensuring a consistent listening experience and preserving the dynamic richness of our music. So, the next time you listen to your favorite song, remember the intricate work behind the scenes that goes far beyond simply "making it loud."
