Okay, this is my first try to make some kind of classification of digital equalizers. It’s mostly based on some defects or features they have in their responses. These defects or features give digital equalizers their unique sound. Sometimes they sound “digital” in bad meaning of this word (i.e. “harsh”) but the other side of digital sound is clean, pristine and maybe too cold sometimes. I don’t have much time to make this classification really glossy so it’s some kind of a draft.
In my opinion there’re 7 main properties of digital equalizers, which affect their sound:
- Frequency response behavior near Nyquist frequency
- Phase response behavior near Nyquist frequency
- Frequency response ringing
- Types of Curves
- Time domain response
Now I’m going to try to illustrate possible cases for each property by some images mostly created by VST Plugin Analyzer. All pictures were created at 44.1 kHz sample rate. Also I do not mention real plugin names used.
1. Frequency Response Behavior Near Nyquist Frequency
Check this picture. There’s a typical 1 kHz peaking curve. Also I shifted this curve using copy and paste in an image editor to show how it’d look for 10 kHz.
Now if someone has implemented cool EQ using equations from “Cookbook formulae for audio EQ biquad filter coefficients” by Robert Bristow-Johnson (RBJ) the curve at 10 kHz could have this kind of a look:
So you can see the right side of the bell shape is distorted. Well it sounds like high Q harsh boost. There’re ways to mathematically match Q at 10 kHz with Q at 1 kHz but the shape distortion remains and such distortion has its own “digital” sound.
Read more of this post