[Note: I’ve been meaning to write this article for almost two years. The question just came up so I’ve included my response here, to kill two birds with one stone 😉 ]
“…if you have any advice to help me catch problematic eq problems before i send off to you to master let me know. I would probably now tackle this by doing a peak eq sweep over my premaster of about +6db over the full range and listen for any areas that have horrendous peaky distortion and resonance then make the appropriate changes to the instruments as we have done in this instance… does this seem a sensible approach?”
Personally I would very strongly recommend AGAINST sweeping eq boosts. There are exceptions of course, and on occasion a narrow boost can be helpful to be sure you’re in exactly the right place. But in 99% of cases, if you’re looking for stuff you intend to cut, then sweep a cut. Many reasons for this, but primarily;
– You don’t kill your ears.
– Sweeping a cut lets you hear how it affects everything else, rather than hiding everything else behind something horrible you want to get rid of anyway.
– You don’t pick up on resonances that might not actually be a problem, turning them into problems unnecessarily.
– It allows you to use a much bigger eq change to find what you need; sweeping an 18dB cut is much nicer than sweeping an 18dB boost.
– You don’t skew things by distorting/pushing things downstream (including your speakers and room).
The general approach I use (for most processing, including compression), I refer to as ‘overprocessing’. I should note that this is nothing new here, I’m not laying claim to some revolutionary thing. It’s just a handy term that encapsulates the following general approach;
1) Do way too much
2) Find the right spot
3) Bypass/set to zero
4) Do the right amount
The whole concept underlying this is as follows; the audio already knows what’s wrong with it, it contains everything you need to know. Overprocessing is a means for it to tell you.
Let’s put it this way; if you know that an 18dB cut (or 30dB compression, or whatever) actually sounds pretty bloody good, then the chances of it sounding bad when it’s only doing 1dB are massively reduced.
So, expanding slightly, with eq corrections it means;
1) Make an estimate of where the problem is before you do anything at all.
2) Make a deep, fairly narrow cut and sweep it.
3) All of a sudden, at a certain point, everything will sound amazing – open, clear, huge, massive, all that stuff. The eq’ed-sounding ‘phasiness’ from the deep eq cut will disappear, and it will feel like the eq is a part of the sound. This is the place.
4) Adjust the bandwidth so you aren’t clearing out masses, but it doesn’t sound eq’ed. [N.B. this is the part that takes most practice]
5) Reset to unity gain, clear your mind and listen again. Make sure you hear the stuff you are targetting come back. Make a mental evaluation of how bad that problem is – awful? Try 3dB. Not too bad? Try 1dB.
6) Make the cut.
When I do this it takes 10 seconds, absolute tops. Often it’s even faster with analogue, as it will be either one switch position or the other. Less farting around.
In time, and with practice, you’ll become extremely fast, extremely accurate, and you’ll find you need much, much less eq because you’re getting things in the right place from the start. Audio Dim Mak. And in line with all the other stuff I harp on about, less processing is always a good thing. Of course if something needs a little low/high end lift or a more general tonal change then you might not need to do this dance, but it works anyway. Of course there exceptions and caveats and so on but I’d rather keep this on point.
This has worked for years for myself and many other engineers, and is a superb habit to get into. I teach it to kids when I give workshops and sell it as the best technical tip I can give them. Because, well, it is.