Last week I discussed the theories behind gain structuring, talking about how to measure sound level and calibrating your equipment to get the best signal to noise ratio. But measuring the loudness of something actually is a whole different story.
When scrolling and listening through a big variety of music, you’ll certainly notice that one tune isn’t sounding as loud as the other. But when measuring the level, both tracks peak at the same level (which digitally should be at 0 dBFS). The thing you are experiencing is called loudness perception, in other words, how loud your brain perceives the track.
How is this possible?
To understand loudness we need to look at how a track has been made “louder”. As you can see on a level meter of a pop production, there is a difference between the average (RMS) and peak level. Reducing this difference is done by extensive limiting “chopping the peaks off”. The smaller the difference between the average and the peak level, the louder the the track is perceived. But this comes at a price.
The peaks I’m talking about are also known as transients. Transients are in fact short high level peaks in the audio, for instance, the *tick* of a bass drum. Transients are fundamental to sounds, and something the human brain is used to hear. Chopping of the peaks of a track will take away these transients, and thus take away part of the feel. One of the studies I did on my high school involved testing the difference between dynamic (not limited) and loud (heavily compressed/limited and distorted) music. The result was simple. The human brain gets tired a lot sooner when listening to heavily compressed music, because it is a constant stream of music, as a result, the people got faster distracted then while they where listening to dynamic music.
That doesn’t sound good, why are we still doing this?
This is a question I ask myself a lot of times. I personally always give my clients the choice of having a little bit less loudness, but the number one comment I get back (and I think most other studio’s will agree with this one) is: “When I compare my track with track X, track X is louder!”. And when we’re going to look at the psychology behind this, we’ll see that our brains automatically judge louder as “better”, when comparing. But actually, it isn’t any better.
This needs to stop!
I cannot agree more to this, and I’m delighted to say that it is going to stop! Thanks to something called “loudness normalization”. To normalize tracks on their loudness (and not on their peak, like what normally happens during a normalization) we needed first to be able to measure it. Luckily, the norm has been made quite recently. It is called the EBU R128 (also known as ITU-R BS.1770). Made by the European Broadcasting Union to deal with the big loudness differences in between commercial breaks.
All you need is LUFS (bwaapwawawawaaaahhh)!
Following the R128, loudness is measured in LU (loudness units) on a Full Scale, so we get LUFS. Where 0LUFS stands for the loudest possible (pink noise at 0dBFS). To give a comparison, modern day electronic music is rated in between -5 and -9dBFS, pop music between -8 and -12dBFS and classical from -16 down till somewhere at the middle of the earth… Measuring LUFS by yourself is simple, just make sure you’ve got an EBU R128 approved meter.
Is my track going to sell more if I have a higher LUFS count?
Interesting question. This first of all depends on how your customer is buying your track, if they buy it because they heard it on the radio, then loudness isn’t influencing the sales at all. Same goes for shops that already include loudness normalization in their previews, because the customer won’t hear a difference! In all other cases, there may be a chance that the customer would want to buy your track. On the other hand, studies show that there is NO connection between sales and the loudness of a track.
However, think about this. Louder is perceived as better. Record labels listen to hundreds of demo’s every day, and I doubt if they normalizing it. If your track can stand out 1LU more then somebody else, I really think that it will have a bigger chance in getting signed!
So, how loud should my production be?
Well, that is is hard to say, at this point and time, loudness still counts. But, with more and more platforms performing loudness normalization (Spotify, YouTube, iTunes), it is just a matter of time. On the other hand, over-compressing has became a sort of “standard” sound in the electronic music business. For me, everything is about making a commercially accepted sound, its not about what I want from an engineering point of view, its about what people are used to hear, and I sometimes find it very challenging to squeeze the last few LU out of a track without making weird artifacts.
This is all for now, I’m already planning to write a follow-up article about this subject. Loudness is hot, and there are a lot of developments going on in this area. I’m pretty interested where we stand with loudness in a few years from now!
If you are also interested in getting your track to commercial standard loudness levels, think about ordering our mastering service!