I wanted to see how precise a Q of 20 to 24 is.
 
I've always assumed that the Sonnitus Q factor could be regarded at face value. In other words, emulation concepts like "hybrid", "E-type", and "G-type" don't apply.
 
If it's correct that the Q factor in Sonnitus EQ is what you see is what you get then:
 
Q = 20 = 0.072127 octave bandwidth at -3dB = 0.865524 of a semitone bandwidth
Q = 24 = 0.060108 octave bandwidth at -3dB = 0.721296 of a semitone bandwdith
 
That's pretty precise.
 
 
 
 
I'm hoping someone will comment further about the comparison of phase response in an analog vs a digital filter such as Sonitus EQ. I always get confused.
 
 
 
edit spelling

 

mike_mccue
The examples you cite, E, A, D, G open chord (first position?) don't repeat the root, rather they include an octave as a harmony.

 
Okay, "same note letter" as the root. I'm sure everyone knew what I meant but if not, they do now. I rushed through that post so I wouldn't miss the beginning of "Edge of Tomorrow" (which FYI I found thoroughly enjoyable).
 

Using 6 dB cuts to illustrate that it is easy to hear a difference is an easy way to make it easy to hear a difference.

 
When posting an audio example, it would be pointless to adjust it so subtly no one could hear a difference, or so heavy-handedly it sounds like crap. The point of posting an audio example is to compare before and after so people can decide for themselves whether the results are beneficial. If they think the results are beneficial, then they will be aware of a new tool for their toolbox and perhaps use it. If they don't feel it's beneficial, they won't. I adjusted the depth to produce what I perceived as a beneficial result with the chosen signal source. Jeff preferred the processed version so to at least one set of ears, I apparently adjusted it correctly
 

I think you are stretching facts here to justify your novel idea, Mr. Anderton.

 
The justification or lack thereof doesn't rest with me, but with the audible results. All I care about is whether clicking on this effect's "enable" button produces something that sounds better. In some cases, it does. In other cases, it doesn't...which means it's no different from any other signal processing technique.
 
Don't worry, no facts were injured in the making of this technique. Guitars exhibit prominent frequencies that relate to particular pitches and notch filters can selectively adjust the levels of frequencies...match up those two sets of frequencies, and it makes an audible difference. Fact-wise there's not much that can go wrong.

mike_mccue
If it's correct that the Q factor in Sonnitus EQ is what you see is what you get then:
 
Q = 20 = 0.072127 octave bandwidth at -3dB = 0.865524 of a semitone bandwidth
Q = 24 = 0.060108 octave bandwidth at -3dB = 0.721296 of a semitone bandwidth

FWIW I added a second control to vary the bandwidth so I could experiment with that. The notches can stand to be a little bit wider with some material but if they get too wide, then it does sound more like a Pultec EQ, as Jeff more or less predicted. That reduces the level of a much wider range of frequencies, which of course can still have merit; but for that application, it makes a lot more sense to use a Pultec than to mess around with a bunch of notch filters.

mike_mccue
 
I'm hoping someone will comment further about the comparison of phase response in an analog vs a digital filter such as Sonnitus EQ. I always get confused.

 
Here's an interesting paper on the issues involved with phase and stability regarding digital filters, particularly FIR types. I have no idea what architecture the Sonitus or ProChannel EQ filters use, but my next test is to try this with the LP 64 so I can hear if it makes any difference.
 
This is a great text from Analog Devices that's loaded with analog filter theory and includes nifty-looking phase response graphs for analog notch filters. Fantastic for geeking out.
 
 

Craig you are probably right in saying although I hear it as a broad improvement it is obviously something else such as the more complex comb you are setting up here. Comb filters do sometimes sound different to how they look! It is an interesting idea and there is no reason why it should not work. By choosing those centre frequencies rather carefully it results in something that may sound a little different to broader EQ.
 
I have had some experience with this in terms of synthesiser filters and making synth sounds for example. Using comb filters over a sound but not necessarily modulating the centre frequencies of the notches. ie just using the sound of a comb filter only.
 
I see the same sort of approach here.
 

 

Jeff Evans
Comb filters do sometimes sound different to how they look! It is an interesting idea and there is no reason why it should not work. By choosing those centre frequencies rather carefully it results in something that may sound a little different to broader EQ.
 
I have had some experience with this in terms of synthesiser filters and making synth sounds for example. Using comb filters over a sound but not necessarily modulating the centre frequencies of the notches. ie just using the sound of a comb filter only.
 
I see the same sort of approach here.

Yes, very much so. People often thing of comb filtering as a "moving target" but as you point out, it has static applications as well.
 
The reason why I come up with these things isn't because I sit around wondering what to do with notch filters, but because I have some problem and start looking for a solution. For all I know this simply emulates what happens with mic placement...maybe when we set up stereo mics, what sounds "good" to us is when the comb filtering caused by using two mics attenuates some of the "rogue" signal peaks...anything's possible!

Anderton
 Yes, very much so. People often thing of comb filtering as a "moving target" but as you point out, it has static applications as well.
 
The reason why I come up with these things isn't because I sit around wondering what to do with notch filters, but because I have some problem and start looking for a solution. For all I know this simply emulates what happens with mic placement...maybe when we set up stereo mics, what sounds "good" to us is when the comb filtering caused by using two mics attenuates some of the "rogue" signal peaks...anything's possible!

Hi Craig,
Just playing around with that this morning. Cool technique, and it works really well with acoustic guitar, particularly my big Martin D35. It seems to buy you some "space" for other stuff to fit into. I've only tried it with a single mic this morning, but I gotta set up a pair of mics and see what happens with stereo or spaced pair.
 
FWIW, the Melda MAutoDynamic EQ works really well for this. You tell it how many additional harmonic notches to create and the depth et voila, all done with a single band.
 
Thanks for posting it up here.
 
Dean
 
PS - a bit OT, but I know we've chatted in other threads about console emulation. Along similar lines, I just got a pair of Peterson Goodwyns Colour modules and have been A/B'ing against various software. I may post some comparisons in another thread (BTW, if you have a bit of time to spend with solder gun the Colour modules are super cool). Anyway, carry on folks.