Note: I started writing this article FOUR YEARS ago. And it’s still going on. Might as well publish it, eh…
It’s a familiar pattern.
Somewhere along the way (thanks internet!), something that is relevant in certain specific circumstances became standard for everything all the time. A practice that’s applicable and required in some situations has become How It Must Be Done. This is true of a lot of practices in audio, and anyone who’s read anything on this blog will be familiar with my stance on these sorts of things.
The particular focus of this
moan discussion, is the practice of placing a low end mono-isation plugin on the master output, either from the start of the mix process, or after the mix is complete. The particular method/plugin used isn’t as important to this article as the practice itself, though I’ll touch on one or two (very important) considerations in that regard.
It’s also worth reiterating that what bothers me most is that people apparently stopped asking themselves why they’re doing it, and are unquestioningly doing what the internet tells them to. But with that being an ever-present theme on this blog, I’ll keep the pontification on that particular aspect to a minimum here. I’ll try anyway.
There are one or two terminological things to clear up first as well. I’ll be talking a lot about stereo information here, but for the purposes of this article I’ll be referring almost exclusively to side/antiphase/decorrelated/out of phase/difference (you get the idea) content as opposed to traditional left/right stereo. The very subject itself obliges one to think in mid-side (or, mono/stereo, or, sum/difference) terms rather than traditional left/right, so I’ll be using ‘stereo’, ‘side’, ‘difference’ and ‘out of phase’ interchangeably. If I mean something else, I’ll try to make that clear.
If the above paragraph means nothing or makes no sense, then you’re probably not doing what I’m moaning about and can set about making better use of your time than reading my drivel.
So with that done, let’s start with a quick summary of the problem, before we move onto solutions, and then, inexorably, onto solutions that become problems in their own right.
The Decorrelated Low End ‘Problem’ aka ‘You gotta mono da lowz bruv’
If you play the same thing in the left channel and the right channel of a stereo signal, you effectively have mono sound – there’s no difference between the two, and therefore no perceived stereo width. When you have something different in the right channel to what you have in the left, you have stereo width. When you have two signals that are of the same/very similar frequency or frequencies that are out of phase, you have decorrelation (aka antiphase). You might have a hi hat on the left channel and a bass on the right; that doesn’t necessarily mean you have decorrelation, or at least, any significant decorrelation. You can think of decorrelation/antiphase as (well, it is) the left channel and the right channel going in different directions at the same speed, at the same time.
Now, we all love a bit of stereo width; it gives a sense of envelopment, depth, and interest. So what’s the problem?
The primary drawback to having stereo width is that *if* you play the signal back monophonically (e.g. out of a radio speaker in your kitchen, on a low bandwidth Internet stream, out of a mobile phone) then you’ll lose anything that’s decorrelated when the two channels are summed. 1 – 1 = 0.
Mostly this kind of thing doesn’t bother people that much these days – it just isn’t as much of a problem as it used to be. Even phones play in stereo these days. When making, say, music intended for a club sound system, it’s not a problem because those systems play back in stereo. Except a lot of clubs choose to run their subs in mono, meaning that any stereo bass content will be lost.
Similarly, this springs to mind for many people when thinking of vinyl. ‘You can’t have stereo bass on vinyl because it will make the needle jump off the record’ and all that. So, standard practice when cutting vinyl is to mono the low end below a certain frequency using elliptical eq, ie a high pass filter in the side signal.
So, you have two purely safety issues – not causing problems with vinyl cutting, and not causing problems on single/mono sub systems.
(As an aside, in conversations I’ve had with working cutting engineers, they rank scratchy/sibilant content up between say, 5-11kHz, as at least as much of a problem as out of phase low end, especially when it too is out of phase. Yet nobody seems to be monoing their high end…)
The Solutions to the ‘Problem’
There are a number of ways to approach this, some of which I’ll mention here. It should become clear where I’m heading fairly swiftly.
– Don’t have out of phase low end in your sounds in the first place
– High pass sounds that have unnecessary low end content individually with normal stereo eq
– High pass the sides of sounds that need mono/centre, but not side, information
– High pass groups/busses carrying sounds that don’t need low end
– High pass the sides of the master buss/mono the entire low end
It’s the last one that everyone is doing now, by rote, without thought, and without considering that there are better ways. Namely, all those above it.
High passing sounds (with normal stereo eq) that have excess low stereo/side information individually in the mix is a totally different thing to removing all that stereo information from everything in one go later on.
Firstly, you’re only doing it to one sound. So you can do it individually, in a tailored way, for each sound. This can provide better depth and separation to your mix, keeping low mids (mmmmmm, low mids) in sounds that need them and not in others, to varying degrees.
The Problems With The Solution
Doing this stuff at the last stage, with a default plugin on your master buss, there are a number of problems.
You’re not only collapsing all that content to mono, running the risk of increased perceived congestion in the middle, but losing content that actually contributes meaningfully to the perceived width of the track. Below say, 80-150Hz (depending), ok, you have to be careful, but aggressively monoing everything below 300Hz is a bit ‘breeze block in a gnat’s face’.
Let’s put it this way: if you sent me your mix to master and I told you I was going to collapse about a third of the spectrum to mono, you’d probably take your work elsewhere. But people willingly do it on their own mixes all the time. I’m not even sure why people do it really; they remove loads of useful stereo information and then wonder why their mixes don’t sound big.
But there’s more to it than that. If you’re using a minimum-phase eq to do this task (most normal eqs, the ubiquitous bx monomaker, most stereo tools type plugins, etc) the inherent frequency-dependent phase shift in the side channel means you are progressively flipping the stereo image by differing amounts around the cutoff frequency.
In short, this sounds you’re twisting your stereo image – badly – by different amounts over a comparatively small frequency range. At the 180 degree phase point, stuff that was on the right is now on the left. And as discussed elsewhere on this blog, the sharper the filter you use to do it, the more twisted your image gets.
Try it. Get an eq plugin that lets you choose linear or minimum phase, and apply an HPF in the sides. AB between the minimum and linear phase options. Hear that? It sounds bollocksed, eh. But people do it all the time coz u gota mono da lowz bruv.
Yes, that’s right: people are messing up huge portions of their mix by applying a process that arguably isn’t even necessary, by default. Ladies and gentlemen, the internet.
I have many customers with plugins like this on their main mix buss. They love the big enveloping sound when I put some low-low mids back into the side channel. ‘It sounds so deep/warm/full’ etc. Yes of course it does – I made use of a load of space that had been cleared unnecessarily. But there’s no way to undo the weird-sounding phase-related image shift associated with that stuff.
I want to really hammer this home: because of a concern specific to vinyl, and occasionally to sound systems, people are doing this much too much, far too aggressively, all the time.
That’s not to say that it never needs to be done. I’m only asking – as ever on this blog – that every case, every sound every channel, is given due attention and treated appropriately. Because you might not need to do this stuff at all.
Blindly doing it all the time to everything is not only completely unnecessary in many cases, but may also be limiting the sonic potential of your mix and undermining all your efforts to make things sound great.