Two Common Confusions about Console EQ
I've seen a couple things lately related to console EQ that I wanted to shed some light on and hopefully get a good dialogue going. But we're gonna keep it short and sweet.
"Air Band" EQ
A couple consoles running at at 96 kHz will allow tuning EQ bands out past 20 kHz. (Some call this "air," which is a misnomer as air attenuates HF the most so it would have the opposite effect, but hey.) These filters produce audible effects. You are not hearing the effects of frequencies over 20 kHz. Even if it's leaving the console (which it's usually not, but you can check with an analyzer), virtually all competent system DSP presets / active loudspeakers / system techs will remove this stuff so it doesn't fry the PA. These filters have effects far down into the audible range.
Here is the HF band of a Midas Pro series console tuned to the max of 25 kHz, showing all four shape settings (PEQ and three different high shelf):
These filters have a clearly audible effect down to 4 kHz. That's what you're hearing.
"That console's EQs are crap."
Sometimes we forget that equalization is a mathematically determinate concept. Under the hood, we are defining responses with mathematical expressions. They look like slightly more complicated fractions. The basic structure determines if we have a HPF, LPF, BPF, APF, etc. Then tweak your variables for gain and Q and off you go. In the context of digital mixing consoles, if the curves match, they will sound the same, and can be made to null to silence. "But wait," you say, "I clearly hear differences in EQs between consoles."
Yes, you do! Here are three popular digital consoles, with their EQs all set to the same settings.
The "same" settings are obviously different enough to be audible. What's really happening here is largely an issue of how manufacturers choose to define the "Q" parameter. This same experiment was done with multiple system DSPs over at SynAudCon with similar results.
So what you really probably mean by this is that you're not familiar with, or not a fan of, the relationship between the channel strip Q control and the under-the-hood values.
If we work backwards and use an analyzer to see what the EQ is actually doing, we end up needing different settings for each console but we can match the curves completely, at which point the effect on the audio signal is obviously also the same.
Why do we need to use an analyzer? Because some consoles don't accurately portray the combined filter response.
Using Photoshop to overlay the console EQ display with the actual analyzer trace tells the story:
Because I know someone's going to ask, the X32 was dead on the money.
Interestingly enough, Midas sidesteps this issue entirely by showing each individual filter, not the combined response, forcing you to "use your ears" (what a concept!):
