Remove Problems in your Audio with EQ - EQ Tutorial Series - Part 3

(This post is part of the EQ Tutorial series)

At a glance, EQ seems simple: just draw a curve, and you’re done. Your voice sounds “magical.”

Well, maybe it's not "magical" exactly, but you get the point. 

EQ is, in essence, about adjusting a line.

Anyone can drag that line, but knowing where to adjust it makes all the difference.

Simply put, you can only bend the EQ curve in two ways: up or down.

• Dragging it down is a cut. It reduces unwanted frequencies.

• Dragging it up is a boost, enhancing certain frequencies.

Let’s start with cutting frequencies to remove problems. If you've watched my tutorials on Audacity or Adobe Audition, you've seen me demonstrate a basic EQ.

This EQ uses a high-pass filter around 80 Hz, sometimes called a low roll-off for speech.

The goal? Cut frequencies between 20 Hz and 80 Hz.

If you’ve seen the EQ chart from Producer Hive, you know that 20-80 Hz is marked as "low-end rumble." 

In the human voice, the fundamental frequency generally starts above 80 Hz. Fundamental frequency means the lowest frequency of speech.

So, if you see frequencies below 80 Hz in your recording, it’s likely noise. Our voices don’t naturally produce frequencies in that range.

A quick test with a spectrum analyzer will reveal this low-end rumble. 

That’s why I recommend a high-pass filter at around 80 Hz in my tutorials.

Now, if cutting low-end frequencies reduces noise, can we cut other frequencies to fix other issues?

Yes, but it’s more complex.

For instance, sibilance (a sharp “S” sound) is a common issue. We usually handle this with a De-Esser plugin, which essentially cuts specific frequencies.

According to the EQ chart, the sibilance zone is between 5-8 kHz. Cutting this range can reduce sibilance. 

However, this area also includes other sounds. If you cut it too aggressively, your recording can sound unnatural.

Each voice has sibilance at slightly different frequencies, usually in the 5-8 kHz range. Many De-Esser plugins allow you to target specific frequencies within this range.

In summary, because voice frequencies overlap with sibilance frequencies, you can’t just cut them out broadly.

The solution is a narrow cut: targeting only a small frequency range. You can identify the problem frequency by doing an "EQ Sweep" and then applying a narrow cut.

This technique works for other audio issues, too. If your recording sounds boxy, nasal, whistly, or harsh in specific frequencies, a narrow cut can help.

Cutting frequencies is one way to improve sound quality. The opposite is boosting.

If you find a frequency range that enhances the audio, you can apply a boost. But don’t boost randomly—there’s a method to it, which I’ll discuss in the next post.

If you’re serious about EQ and want a custom setup, I offer a Custom EQ with Macro/Preset for Audacity and Adobe Audition. 

You can get a free demo by sending me your raw recording. 

For details, visit the custom EQ page, where you’ll also find my email for any questions.