EQ

EQs have no flavours#

In my opinion, EQ and compression are the two tools to learn before anything else. Nobody needs a course on how to use a flanger. And with EQ, the good news is that you only need one.

The industry says you need a whole bunch of EQs for different flavours: this one for guitars, that one for kick drums. So I null tested 16 of them, analogue-modelled ones, expensive ones and free ones, against a regular digital EQ. Properly matched, almost everything cancelled to the point of inaudibility, even on white noise, the hardest thing to null. When I quantified it on real music, almost every result sat below the just noticeable difference, even with the delta exaggerated by 20 dB. One emulation did something measurably nonlinear. I have no opinion on whether that is good or bad.

The exceptions are odd shelf shapes and odd phase behaviour, which are not desirable flavours, just weird curves a normal digital EQ can recreate. Linear phase is the one genuine alternative, mostly useful in the highs, and a good all-rounder digital EQ probably might have it already. So in almost all the cases that actually matter, one high-quality, full-featured EQ gives you every tone you need (null testing).

What you hear is the interface and the Q#

Two things often fool people. The first is the interface. If it looks like it came out of an old console, you can think it sounds warm, because what you see influences what you hear (perceptual bias). The second is Q. Frequency and gain are calibrated the same everywhere, but a Q of one on one plugin can be a Q of two on another. If you type in the same knob values, of course they sound different.

The vintage sound is in the transformers#

The filter network of a classic passive EQ is resistors, capacitors and inductors. They are linear components. A capacitor does not saturate. Whatever character those units have mostly comes from their transformers and amplifiers, not from the EQ curve.

I love the sound of transformers. I have built analogue EQs myself. But if I want transformer saturation, I add it with a tool made for that, rather than accept unspecified mystery distortion baked into an EQ emulation. See distortion and analogue modelling. I don't find workflow a convincing argument either: if you cannot dial in a high shelf and a bell cut on a digital EQ in two seconds, it's better to learn EQ than to buy more EQs.

EQ fixes a problem you can name#

I use EQ to fix a specific problem I have heard in the source. Putting a vintage EQ's signature curve on everything is random: if it happens to fix the mix, that is a coincidence. A broken clock is right twice a day. For the same reason, "cut 1 dB at 160 Hz on every mix" is not how professionals usually work.

For me, mastering EQ means precise moves made for a reason: a narrow dip on a resonance, or tilt, one of the most foundational moves there is. In mastering I work with a strict negative approach, cutting and hardly ever boosting, because a boost can trick your ear into preferring it simply for being louder (loudness matching).

Many EQ problems are not EQ problems#

• Dull vocal. Vocal highs are very dynamic, so a naive shelf can make it thin, harsh and sibilant. I de-ess first, then add air.
• Weak sub. You don't always start by boosting the bass. One option is to generate harmonics just above it. The ear infers the fundamental, even on speakers that cannot go so low.
• Buried part. Often masking, not level. The fix can be to thin or rewrite the competing parts, pan, or duck only the frequencies that clash, like the guitar mids under a vocal.
• Boomy kick One fix is to cut the bass out of the kick and turn the kick up.

A dynamic EQ needs attack and release#

I use a static band for constant problems, and a dynamic band for problems that come and go, like sibilance or a one-off scream that jumps out at one frequency. A dynamic EQ without attack and release controls is as useful as a chocolate teapot.