Equalization, please tell me there’s no math involved…
Take a deep breath and cast the horrid flashbacks of math class aside, there is no mathematical knowledge necessary — or at least not much. My online mentor Graham Cochrane of The Recording Revolution explains EQ as a volume fader for individual frequencies.
You see, the range of human hearing goes from roughly 20 Hz to 20,000 Hz (commonly referred to as simply “20K”). Audio engineers break this frequency spectrum into 5 major bands: the lows, low-mids, mids, high-mids, and highs.
Most sounds that we make and hear cover a wide range of frequencies. The frequency spectrum analyzer below shows the frequencies of a male speaking voice. You’ll notice that the loudest frequencies in this voice are a chesty low-end rumble at 100 Hz, some throaty low-mids, and a peak at 5K from the high pitched consonant sounds like “s” or “c” (known as “sibilance” — which is a good example of itself). What we can do with EQ is raise and lower the volume of certain frequency ranges in order to sculpt a more pleasant sounding balance.
Great, show me a picture so I can copy your settings!
Well, not so fast. Any given voice recorded through any given microphone in any given environment will result in a different balance of frequencies. There is no EQ template that will work on every recording since each one is unique. There are, however, a few tips and tricks that will put you on the fast track to improving your audio quality drastically.
High-Pass Filter? What is this, pole vaulting?
A high-pass filter, also called a “low cut”, does just that — it cuts out the sound below a specified frequency. For the purpose of recording your voice, there is almost no reason to keep any of the low end frequencies below 100 Hz. Any frequencies in this range are likely to be the subtle rumbling vibrations that the microphone may pick up from passing cars or someone stomping around in another room.
As a rule of thumb, you can play back your audio in your DAW and slowly drag your high-pass filter up from 0 Hz until you start to notice that the sound is becoming too thin and bright. Once you hit that point, slowly roll back the filter a bit until you hear the roundness of the low-end come back. Much like the story of Goldilocks, EQing is often a process of tweaking back and forth until the sound is just right.
Keep in mind that you should not be fooled by what the graphical interface is showing. Ultimately what you hear is more important than what the frequency graph might appear to show. Try closing your eyes while you drag filters around in order to focus your attention on how they are changing the sound rather than how they are changing the graph.
Wow, my recording doesn’t sound so heavy and muffled anymore!
Exactly! And that’s just the first EQ move we’ve made. The next will go even further in cleaning up your audio.
As we saw before, the human voice is heavy in the low-mids. Very often this frequency band can be scooped out a bit to add clarity. This is called “subtractive EQing.” Boosting frequencies can lead to harsh sounding results that may crank your audio signal too loud if you’re not careful. It is generally favorable to remove some of the frequencies you don’t want in order to reveal more of the ones that you do want.
In the example below, I adjusted the width (referred to as the “Q”) of the yellow peak filter so that it covered the low-mid range, and reduced the volume by 3 decibels (dB). It is a good idea to make EQ changes in 3 dB intervals. If you find that cutting 3 dB is not enough, go to 6 dB, and so on until the problem frequencies are well balanced.
EQ changes can be subtle, and it takes practice to train your ear to hear what needs to be fixed. As a beginner, focus on wider, more drastic filters to set the general balance of your EQ. If you are unsure whether an EQ move is helping or not, try turning the filter on and off to hear the change it’s making.
It’s like a layer of mud was cleaned off my recording!
I know — it’s great isn’t it? Cleaning up the low end and the low-mids is most of the battle. From here you may even be able to call it a day. If you’re still not happy with the audio quality, you can experiment with more filters until you find the right balance.
A very helpful trick is to raise the volume of a peak filter by a large amount, say, 18db. Then, you can drag this filter from left to right until a frequency jumps out at you for sounding unpleasant. Of course, any frequency boosted by 18 dB won’t exactly sound normal, but believe me, you’ll know when you hit a frequency that needs to be tamed. You will develop a much better sense for this the more familiar you get with the frequency spectrum.
Here are some other tips and tricks to help get you in the EQ mindset:
I know I said boosting is less favorable than cutting, but a slight boost to the midrange around 1-2K can add some nice clarity and presence for vocals, especially if the voice is talking over background music.
Harsh Frequency Notch
If there is a particularly harsh frequency, especially in the high-mids, that is sticking out, like the sibilance we saw before around 5k, you can take a peak filter with a narrow Q and do a dramatic cut — or “notch” — to that specific frequency.
This may be more relevant in the world of mixing music, but a high shelf boost can add some high end excitement that gives the recording more energy. Be careful though, as these high frequencies can easily become harsh and unpleasant.
What if I’m not really sure what EQ changes to make?
Explore! EQ is all about trial and error. Feel free to drag filters around the frequency spectrum to find the problem areas. If it’s too muddy, clean up the low end. If it’s too harsh, tame the high end. If it’s in between — you get the idea. Think like Goldilocks and your YouTube videos will have professional quality audio that sounds just right.
Albano, J. (2014, August 20). [EQ Bands]. Retrieved October 26, 2016, from https://ask.audio/articles/5-essential-mixing-tips-achieving-good-eq-tones