Grasping 3 band eq

Just remember that raising the Q factor will narrow the bandwidth over which the EQ is applied.

Here is a rough guide to give you an idea. A Q of 0.1 will be several octaves wide. A Q of 1.0 will be just under 1.5 octaves.
Remember octaves get larger as the frequency moves higher. 20 to 40Hz is one octave. 3000 to 6000Hz is one octave.


Q vs Bandwidth.jpg
 
TalismanRich said:
Just remember that raising the Q factor will narrow the bandwidth over which the EQ is applied.

Here is a rough guide to give you an idea. A Q of 0.1 will be several octaves wide. A Q of 1.0 will be just under 1.5 octaves.
Click to expand...
OK. That answers the previous question. My guess was the other way around 🙃 Thanks, Rich.
 
Thanks talisman. Yes I sensed that was what you were driving at in the earlier message . Makes good sense. Basically if someone leaves a cake on the kitchen floor and you step on it it's Q height is only 1mm but it's now twice the width -as a rough visual metaphor
 
Last edited:
Record some pink noise and hear what the controls do. You’ve got the theory, so now time to experiment. Having had to calculate resonance with component values 40 years ago, the Q of a filter in my head translated to ‘quality’ it was harder to make narrow filters than wide ones. The bell curve, rather than straight line, is mainly due to the graph horizontal axis not being linear, so the display always exhibits the same basic shape. In a DAW it’s easy to create high Q, so really narrow notches or boosts can be done, but in analogue circuits, lots of gain at a very narrow bandwidth, is beyond what the circuits can produce. I still think in a DAW that most times I’m widening the curve, which sounds more musical, and problem solving with a graphic or separate parametric or graphic. I’m guilty, like others, of constantly adding or subtracting multiple points based on what I am seeing on a screen, not hearing with my ears. The most useful are often the gentle roll offs or boosts, top and bottom, not spikes and dips in the middle. Bass and treble, really.
 
Yes it's going to be a lot of trial and error with this . Instincts tell me not to narrow the default at this time which is 0.5 . Otherwise i'm possibly more inclined to cut out something narrow and anomylous sounding this way . I'm hearing a ton of Junk in the piano in this track . I figure alot is mid and lower mid wool . I'm trying to rationalise that this track is offering a pulse more of the top area and broadly the chord information. Sacrificing the tone somewhat to allow other instruments to come through . There's also a notion I picked up on somewhere that a broad decision for which instruments will occupy which space in the spectrum is a fair philosophy
 
I think now that you're beginning to grasp the idea, we'll move on to the actual formulas the rest of us here are already quite familiar with...

f.jpg
 
Higher number = narrower filter. 0.5 is pretty wide, 1 is - well, it's sort of typical I guess. 3-5 is pretty sharp, 10 is more like a notch filter.
 
You must log in or register to reply here.
Back
Top