Equalization Part 2: The Art of Equalization

Equalization is nothing less than attempting to obtain a balanced frequency spectrum.  You don't want any frequency to be exaggeratedly loud or retreated behind other frequency ranges.  If you run a spectrum analyzer on your music, this will show you the frequency spectrum of your signal.  In an ideal world, many people say you'll want it to be flat.

When you analyze commercial CD's, you'll see that a lot of them have a gentle rolloff that can start from 500 to 1500 Hz. Older records will usually have a rolloff of 6 dB/oct, newer records usually sound brighter, and use about 3dB/oct. You don't have to follow this rule, but a good practice is to listen to records you like (avoid mp3's please, that's a very bad reference for listening to frequencies and stereo imaging), then run them through an analyzer, and see what they did. You'll be amazed how much tunes use the same rolloff. But, as in most guides, don't blindly follow what is said. Different tunes will need different applications. The explanation to that rolloff, is that most people find the sound of a live symphonic orchestra pleasing. If you'd analyze this, you'll probably find out that there's some kind of rolloff of the high frequencies going on. As people like to borrow from pleasing stuff, the link is made quite easily.

Again, I point to the fact that you need to do what is best for the music, not because everyone else does it. If your music sounds better without the rolloff, for god's sake, leave it as it is. I have to make one thing clear though. For a quick comparison, it's ok to take commercial tunes as an example. But don't try to make it sound completely like them, because you won't be able to. Don't forget those tunes already have been mastered, so with mixing only you won't always be able to achieve the same results. Always keep that in mind. Of course, if you can come close to such results without mastering, so much the better, there will be less work at the mastering stage. But don't be blinded to try reaching that sound for hours and hours.

Corrections

There are mainly three uses of equalization. The first is a corrective measure for flaws in the sound (bad mic placement, room acoustics, badly tuned instruments, ...). It's not that common in completely electronic made music (if the sound isn't good, you tweak the Synth, that's all), as it applies mostly to recording with microphones. In the case you will record stuff with microphones, pay close attention. In this situation, the less EQ needed is the better.

You must know that "normal" EQ's have a major flaw. They rotate phase. Simply put, phase means where the pressures and depressions are in the sound waves. A sound wave that plays alone doesn't care about phase. Phase only poses problems when two waves play together, and more importantly, at the same frequency. Phase is a relative occurrence. Let's take an example. You have two speakers, that play a beat. But one of the speakers is wired oppositely. In practice, it means that on the beat, one speaker will move outward (create a pression), while the other speaker will move inward (create a depression). When you stand in the middle (both sound waves reach you in the same time), you'll have phase cancellation. The opposite waves counteract each other. In a perfect world, if both waves are in phase opposition (180°), and both have the exact same level, they'll cancel each other out completely (no sound anymore).

If they are both completely in phase (compressions and depressions are at the same place, 0°) they'll amplify each other by 6 dB. If the phase difference is 90°, you'll have a 3 dB gain. The best option is always in phase, you can reduce the amplified level. If you have phase cancellation, good luck compensating. That's why most mixers have a phase reversal switch. Take a snare drum recording for example. Usually it's recorded with two microphones: one on the top and one on the bottom facing the other direction. When the drummer hits the snare drum, they'll both receive the pression, but reversed in phase.

Why is this important in EQ'ing? Well, normal EQ's have a flaw, they rotate phase. The more you boost/cut, the more they rotate (around the cutoff frequency). This can result in unwanted phase cancellations. As your main aim here is correction of flaws, the last thing you want to do is add another one. So when you use the EQ as a corrective measure, the less you need to correct is the better.
You'll try to reach the sound you want by finding the correct room for the recording, use the right microphone for the job, and place it well. EQ would be your last resort if anything else fails. In the pro world, there's a rule that says, if you need to correct more than 3 dB somewhere, you did a bad job at the recording. Ideally, you'll do the recording again, better this time. If it's not an option, well then you have no other option and you'll have to be happy with what you can do. Also, if you apply massive EQ'ing, the filter can induce distortion (they "ring" around the cutoff frequency), that can be audible. The transients (attack of the sound) can get smeared, and in the worst occasions, pre echoes can even be heard.

Creation

The second use is a creative tool on itself. Here, to hell with the phase rotation, as you'll tweak the sound so much it will MAKE the sound you want. Here there's no right or wrong, you do what you want, to achieve the sound you want. But again, keep an eye out when boosting. You might to do some crazy stuff, but remember that boosting doesn't only add the wanted stuff, but also brings the noise and other nasty things you don't want up. But of course, in some occasions that's exactly the point.

Fitting it in the mix

Getting each aspect of your track to "fit in the mix" is the most important use of the EQ. The aim here is to make your mix sound consistent. Whether you started from an acoustic recording, or from electronic instruments, you'll have to create the feeling that the different instruments are part of one same tune. A lot of amateur productions really have problems in this department, you can really hear the different parts are thrown together, but they don't "glue." Of course, EQ'ing isn't the only tool that can help here, but unless an effect (reverb, delay, compressor, etc...) is an integral part of the sound, try to fix problems in a "dry mix" first. With dry, I mean no effects. Fitting instruments in the mix doesn't only mean that, but it's also here you'll define the identity of an instrument. Depending the role the instrument will have in the mix, you'll EQ it differently. If a lead Synth needs to play solo, you'll leave it fuller. If the Synth only serves as a rhythmic reinforcement, you can cut more frequencies, to leave more room for the more "important" sounds. You can also emphasize a certain part of the instruments sound, so it cuts easier through the mix.

First of all, start with a rough level balancing. Start with all faders at the same level, and then make quick adjustments (you know, if the hihats are too loud, reduce them) until you get a decent mix to start with. If you have lowcut filters (like most hardware mixers have) engage them on the high frequency instruments (for example, hihats really have nothing to do in the low frequency range). Listen one or two times to this mix, and note any apparent problems. Try engaging the lowcut filter on other instruments/sounds (like vocals, and lead synths, etc...). If it doesn't make an apparent difference, or the attenuation of bass doesn't pose you a problem, leave them engaged. Of course, you want to keep the bass on the low frequency instruments. This will already leave some more room for the bass instruments. In any music, and in contemporary music above all, the kick and bassline carry all the rest. So getting them right is of big importance. Solo them together. A little word of care about soloing though. The human ear has the ability to adapt to sounds very quickly, so don't spend hours fiddling on one sound until it sounds good.

The aim of EQ'ing is to make it sound good IN THE MIX, not on it's own. And the adaptation of the ear makes it even more difficult. If, for example, you find the kick to lack bass, you add bass to it. After some time your hearing will have adapted to it, and maybe you'll have the impression it still lacks bass. Again you add. But after you take a rest, letting your ear get fresh again, and you hear that sound again, you might jump up surprised at how much bass you actually added. So caution when you EQ in solo. Get out of the solo once and a while, and compare how it sounds in the mix. If you have an EQ bypass button, switch it on and off sometimes to compare the results with what you started with. This is true for most instruments, but even more for bass and kick. They are playing in the same playground.

A mistake a lot of people make is they want punchy and bass heavy sound. So they add bass to the kick. Suddenly the kick muffles the bass. So they add bass to the bassline. Then they don't hear the kick anymore, and start all over again. In the end, it's one big mess. The kick and bass are fighting with each other. The solution to this is planning. You need to define which instrument will take care of the very low frequencies. He and only he will take care of it, the rest needs to stay out. For example, a common practice is to add bass to the kick around 80-100 Hz. The bassline should stay out of this range now. If you want to hear the bassline more clearly, add some EQ on higher harmonics of the bass. Maybe 200 Hz could do the trick. If the bass has some kind of pluck, you might want to emphasize that, or not, depending on how you want it to take part in the tune. These numbers are only examples. They will differ on each tune. And those rules apply for every other instrument also. Of course, it's a bad idea (and frankly impossible) to have no overlap at all (like instrument A sounds in one frequency range, instrument B will start sounding in the frequency range just next to it, without overlapping). But you should try to give each sound it's own space in the spectrum. If you have two lead synths, give one a little push in one frequency range, and give the other Synth a push in another range. Not the same one, as they'll fight each other then.

Of course, this implies that you should have thought about this aspect in the creation of the sounds too. It's generally a bad idea to let several similar sounding instruments playing at the same time. But when we say give each sound its own space, we don't mean that the ranges can't overlap. There are people actually trying to fit individual instruments in very defined frequency ranges, with those ranges having as small an overlap as possible. There are different schools of thought about this, but if you don't know what you're doing, don't venture there. Personally, I think leaving the instruments "complete" will sound better. For example, the bassline doesn't only have pure bass. The harmonics of the bass sound are important too (like I said, it could have a pluck to it, or some kind of other clarity). If you remove those, the bass can quickly become lifeless. The same applies for all sounds. When we mean: "fit it in its own space," we mean that the main focus of the sounds shouldn't be in the same range. Like explained above with the kick and bass.

You should really see your mix as a wall. Each sound will be a brick that will help construct the wall. You can't place two bricks in exactly the same spot, as that will result in a mess. When you use compression, experiment also with placing the EQ before or after the compressor, as this will yield different results. Of course, after you EQ, the levels in your mix can change, so compensate accordingly. For example, if you cut some low frequencies in a lead Synth, that Synth can probably be mixed in louder, you might have less low oomph, but the increase in volume might be exactly what you need to have the power you wanted. More bass doesn't necessarily mean more power! The following is a chart that describes some of the frequency ranges, and what they do to several instrument classes. I don't know who made it but props to him/her ( and if this is yours, please get in contact with us, so we can either give you credit, or remove it depending on your wishes).

Of course, this chart isn't the bible. It doesn't mean that it WILL have the desired effect. But it's a good start if you don't have a clue. In general, you'll have to find the frequency you want to affect. The best option is a full parametric EQ, (but a semi is usually good enough). The best way to find the frequency is to set maximum EQ gain (yes boost all the way up), and slowly (well, not TOO slow either, remember ear adjustment) sweep the center frequency until you find the frequency you want to adjust. Once you found it, you can apply the boost/cut you wanted. This method is referred to as "focusing the EQ."
Be sure to check it in the mix, with EQ on or off. The aim is to make it sound better, not worse!

Some final thoughts

There's not much more to it really. The rest is up to you. You can read all the guides you want, in the end it all comes down to practice, and how you really want it to sound. If you already have a bad idea in your head, you'll never make it sound right.

So at last some things you should consider:

• Don't use the EQ for fixing problems all the time. I see a lot of people inserting an EQ on the master to fix problems. They hear there's too much bass in the mix, so they remove bass on the master EQ. Not the right way to work. If your mix sounds too bassy, it means the low frequency instruments are mixed in wrong. Try to fix problems at the source.
   
• Keep in mind EQ'ing is balancing. A balance has two sides. If something sounds too dull, you have two options. Most people will increase high frequencies. Have you ever thought that decreasing some low frequency content might do the trick too? We said cutting is usually preferred over boosting.  If you want to change something in the balance, always try cutting in the "opposing" range to solve the problem first.
• If you're really serious about producing, it can help if you know what instruments take up each frequency range. I made a training CD for myself where I can practice frequency recognition. Take a track you know well that contains enough frequency content. Take a 10 second sample or so. With a band graphic EQ, make a sample each time one EQ range is boosted at maximum (so you'll end up with 32 10 second samples [31 bands + the original]). Listen carefully to each sample, and listen on which instruments it acts on. Eventually, you'll be able to pinpoint certain frequencies pretty accurately. It can be very useful in mixing and mastering to know this, as this can considerably speed up your work.
   
• Low frequency sounds require musch more energy to have the same apparent loudness than high frequency sounds (just look at the balance between the kick and the hihats). That's the reason why engaging the lowcut filter on high frequency sounds can be important. Maybe you don't hear a difference, but there can be some low frequency content from another instrument that is taking up a considerate amount of energy. Energy that could be spent to squeeze just that tad more into the mix to make it more powerful.

You do all the of this to make your music sound better. If you look back at everything that is said in this guide, you'll see everything is pretty technical. But an EQ is a tool that you need to understand to the point where you can use it without thinking. In the end, you don't want all that technical babbling to get in your head, you just want your music to sound great. Don't forget that the tools are there to achieve that goal. If something didn't improve your sound, don't be afraid to start over with a different approach, or even take it off completely. If it doesn't help the mix, it only makes it worse, that's one of the things you'll find out pretty soon.  Remember, you want a coherent production. The aim of this all is to make everything glue together. How you do it is up to you. Just keep in mind that audio is an illusion. If you can fool the human ear and tell it your track sounds great, it will believe you.

Related Sites According to Google