With ever greater engineer awareness around the virtues of keeping things quieter, many could be forgiven for thinking that the race to the top is no longer a problem. Here UK mastering engineer, and founder of Production Advice Ian Shepherd highlights why the industry might not quite be out of the red just yet. Over to you Ian…

A Nemesis Returns

I wrote a post back in 2009 called “How Spotify will end the Loudness Wars”. And another in 2015 called YouTube just put the final nail in the Loudness War's coffin.

But with hindsight, I have to admit that I was being more than a little over-optimistic! Fifteen years later, super-loud masters are still business as usual, especially for mainstream releases, which routinely get as loud as -8 LUFS or higher, and every day I see people online asking “how do I get my masters as loud as…” followed by the name of their favourite artist.

Meanwhile, most people today are listening via the major music streaming services, which have loudness matching (normalisation) enabled by default to avoid complaints from users being ‘blasted’ by sudden changes in loudness. When I asked in a popular audio group on Facebook, over 75% of artists, engineers and producers said they disable these features in order to hear the music ‘as it really is’. But in stark contrast, only 17% of normal users - the fans - do this. On YouTube, which has over 80% of the users, you can’t disable normalization.

So the cruel irony is that artists and engineers - the people who care most about the way the music sounds - aren’t listening to it the same way their fans do, and spend a huge amount of time and effort trying to achieve high-LUFS results that most people will never hear.

So why is this ? Why are we still pushing for extreme LUFS levels when normalization means there’s no need to ‘compete’ any more, and in fact masters with balanced dynamics sound just as big and punchy at matched loudness, often more ?

Here are the most common reasons I see people giving:

• High LUFS are necessary for commercial success.

• High LUFS are part of The Sound for many genres.

• High LUFS masters still sound louder even when they’re played with matched LUFS.

The thing is, even though all these ideas are accepted as facts by many people, it’s not quite that simple. Research has failed to find any connection between LUFS and commercial success - music fans simply don’t care. And simply pushing the LUFS up into a limiter alone rarely achieves ‘the sound’ of a genre. There are so many other factors that are more important - tonal balance, contrast, arrangement, stereo width, density, intensity, saturation, distortion… there’s much more to sounding loud than just measuring loud.

The final point is more complicated. I’ve had the experience myself of comparing two different masters, where one measures much louder than the other and it still seems that way, even when the LUFS are matched. Listening more closely almost always reveals there’s more going on, though. EQ balance plays a massive part in perceived loudness, for example - and the internal dynamic structure of the song plays a crucial role as well - the contrast in level between the verse and the chorus, or between the intro and the drop, for example. In my experience the raw LUFS values very rarely tell the whole story.

Dynamic Range Day

It’s much more effective to demonstrate this in practice than simply describe it though, which is while I’ll be providing some concrete examples for you to listen to as part of a livestream I’ll be doing on Friday, 22nd March - as part of Dynamic Range Day. I’ve been running the event since 2010 to raise awareness of the issue of the Loudness War - and in a year when the new Beatles’ single measured -6 LUFS, it feels more important than ever! If this is a topic that interests you, please come along and join in.

So: if I’m right and super-high LUFS don’t give any practical advantage online, why are the Loudness Wars still an issue ?

I think perhaps the most likely answer is simply that there are still plenty of places where loudness matching isn’t in use yet. Some users disable normalization; most browser players don’t normalize yet; some, like SoundCloud and Bandcamp, don’t use it at all at this point; and of course CDs always play at the original loudness level - until the listeners adjusts the volume control, at least.

Which means that even if more than 80% of people are hearing their music with the loudest songs reduced in level, that’s still about 1 in 5 listeners who aren’t. And if a louder result can be achieved without compromising the quality of the results, why not master loud?

Balancing Dynamics And Loudness

For me, that’s a big “if”. Beyond a certain point it’s really tough to get things to sound great, in my experience - the effect of more saturation, more limiting, more parallel compression and more clipping inevitably takes its toll - especially when heard with matched loudness. And the fact that plenty of songs have much lower LUFS but it doesn’t affect their popularity on services without normalization like Soundcloud, for example, supports the research that says music fans really don’t care either way.

But the pressure from major labels and artists who firmly believe loud masters are crucial to their sound and their success is still intense. So ultimately it’s a balancing act we all need to decide for ourselves, which is why understanding how all this works, and testing it for yourself, is so important.

My own opinion is that normalization is simply going to become more widespread (and more effective) as time goes on, so I encourage my clients to future-proof their masters by optimising for the majority of listeners. The reaction I get to this approach is unanimously positive - I encourage you to give it a try, too !