Ok, slightly geeky audio post alert 😉
Whilst I was putting this site together, I fell into the trap of the now decades old “Loudness War” (go Google it if you want the whole gory history). See, this is what happened…. I put together a start on my Mix Examples page and thought “I know, I’ll go look at a bunch of other sites and see what kind of stuff they have up”. That’s when the trouble started. The thing was, clicking back and forth between other sites and this one, my mixes were sounding weak. Bugger. Maybe I can’t do this after all. I quickly realised that the mixes I put up were actually just a bit quieter. In my slightly stressed condition, I rushed back and remastered the tracks to a comparable volume, but then they lost something unless I played them really loud (and then I couldn’t listen for long). I was about to give up the whole thing until my sensible brain kicked in (I have one, somewhere). Take a breath, use your volume control, then decide. So, with some careful matching of volume I did a better comparison and, what do you know, I preferred what I’d done…..
It all comes down to how our immensely clever ears and brain work. For whatever (probably evolutionary) reason, presented with two similar sounds at different volumes, we’ll favour the louder one. Our ears also have a different frequency response at different volume levels* – with louder sounds we’ll hear more bass and high frequencies, compared to the mid frequencies (things will sound more powerful and bright = more exciting = better). So, starting way back (yeah radio stations, I’m looking at you), people started to try and make their stuff sound relatively louder than other people’s stuff (if the listener leaves their volume control alone – which most do once they start listening). However, there’s only so much volume you can add to a recording without altering it in some way. Here’s a fairly typical percussive sound (it’s a snare drum)…
…so, within this file (being as we’re in the digital age) you could turn it up by 4dB (0dB is the absolute max that can be stored) without changing the sound at all – so the perceived loudness would be -24dB. That peak can’t go above 0dB. But our ears don’t respond fast enough to perceive peaks as “loudness”. We perceive loudness based on the average energy of the waves hitting our ears (ok, that’s slightly simplified, but more or less correct), so in this case the perceived volume is 24dB less than the peak loudness. So, to get your stuff to sound (relatively) louder than other people’s stuff, welcome compressors into the world (thanks again, radio stations ;-)).
Compressors (and their extreme siblings, Limiters) allow you to reduce the dynamic range of the sound so you have more headroom to turn it up. They do this by selectively decreasing the level of the peaks (hence “compressing”) whilst leaving the rest alone. Here’s a compressed version of that sound…
…so now there’s room to make it 11dB louder – making the perceived loudness -19dB (5dB louder than was possible before compressing). There’s a price though – we’ve changed the sound. With careful compression, you can increase the perceived volume quite a bit without it sounding too different to most ears, but once people started the race to the top of the loudness scale, like Berserkers in battle, everyone took it too far. The result? Distortion. Too much compression sucks all the life out of the sound. Even though our ears don’t respond to the peaks as loudness, those peaks still get interpreted by our brains. Those first 10-20ms of sound are critical to us recognising what the sound is, and where it is, and how far away it is. All sorts of qualities are held in that transient (the start of a sound) and compression changes that, if only with the relationship between the transient and the rest of the sound. In mixing, we use that positively to alter individual sounds on purpose – to make them more punchy, or gritty, or sustained, or to bring out some interesting character of the instrument/voice. You can do the same when mastering the whole track if you do it well, and your goal is to improve the sound rather than just to make it louder.
So, what’s the good news? These days, more and more media players and online streaming services (because people playing CDs and vinyl know how to use the volume knob, right?) are using some form of replay volume adjustment. There’s an informative article by Ian Shepherd detailing some of the loudness adjustments which are made by music streaming services. What it boils down to is that, if you “loudness war” compress your music too much (to try and make it sound comparatively louder) YouTube, or Spotify, or whoever will just TURN YOU DOWN! As a result, your track will not only be the same volume as everyone else, it’ll sound like crap too. It’s a lose lose situation. Most forms of music still benefit from the skillful use of compression and limiting at the mastering stage to add life and detail, but the days of over-compression are definitely fading away. If we want it to sound better, we just have to make better music – whatever volume someone plays it at – and that’s a good thing.
End result? I’ll be producing masters that use compression to bring out the best in the music, and not just to make it loud.
* Interesting factoid – our ears are most sensitive to frequencies in the 2-5kHz range, which happens to correspond with the strongest overtones of the sound of a crying baby. Yeah, thanks for that Mother Nature.