Explain this Parametric EQ setting to me, please!

[rasputin1963]

Hey guys,

 

Here is a screencap of my favorite EQ at work:  Audio Elemental EQIUM.

 

Notice that I've deployed the filter called HI SHELF II

 

The regular high shelf simply boosts (or attenuates) all frequencies above a given range;   but HI SHELF II  not only boosts,   but places an attenuating  scoop/depression  right before the boost.    The higher the high-frequency gain,  the deeper is the scoop occurring right before it.   What is that scoop all about?   Why,  artistically or engeneering-wise,    would one want to do that,  exactly?

 

Thanks,  ras

[Anderton]

The answer, of course, is...why not?

But seriously, back in the days of analog filters, there were often interactions among stages and anomalies. Quite a few filters add "alternative" responses that attempt to re-create this kinds of responses; this may be one of them.

[MikeRivers]

A filter with anything but the simplest response (6 dB per octave cut or boost, or single frequency resonance) is a combination of filters each with its own frequency response. Multiple filters are "stacked" to make steeper slopes, flat response after a boost or cut, and such. The dip in the high frequency shelf is likely a result of another filter to make the rise steeper (or less steep) than a simple high frequency boost.

As to why it might be more (or less) useful is that it tends to emphasize the action of the filter that you're adjusting since the gain drops a little at a frequency just before it's raised.

Thinking in the other direction, synthesizer filters often have a "resonance" parameter that causes a little peak in the frequency response to emphasize the frequency at which the filter starts working. The frequency response curve of a high frequency boost with "resonance" looks a bit like a step waveform with overshoot.

 

[WRGKMC]

All presets are simply a jumping off points.  One may or may not work depending on your material and the recorded tracks you use it on. 

That particular screen shot has a dip that will attenuate guitars, snare, and some vocals and boost the drum cymbals. If it works, great. If not, you have to customize its response to what you need, not let a preset built by some software author dictate what your music needs. 

A 12db boost of everything above 4K is painting with a broad brush. Its more like spraying lacquer over a finished portrait to make it glossy. That's about it. Its likely a preset designed for a mains or drum bus to compensate for a lack of high end in the mix. The problem with setting is it isn't customized to boost individual instruments and will boost not only the instruments that have high end frequency but it will also boosts any noise that exists up there and add noise and hiss to the recording. If you like hiss fine, but its not a good surgical tool to fix real issues.  

The better way to fix a mix if that setting is being chosen is to go and fix the individual tracks so they additively add that kind of brightness to the mix.

Instruments produce notes at set frequencies. They also produce harmonics above the root note in mathematical multiples of the root note. How much of this harmonic content is captured recording depends on the quality of the tracking, mics and sound source.

If the high frequency content is lacking its because there wasn't enough of it tracking. If you try and boost ghost frequencies that don't exist, you only succeed in boosting unwanted noise between the notes and wash the soundscape out to a blur. 

Remember, music consists of notes and rests. The rests of silence between notes carry just as much if not more weight than the notes themselves. People often try to boost the energy of notes beyond their capacity and in the process destroy the dynamic contrasts of the silence at the bottom between transients.  The canvas of the portrait is a silent background. Notes are the transparent color you place on the canvas. The white canvas, the silence,  is what gives the transparent paint (notes) different shades. If you try and cover up all the silence, the silence between notes disappears and you end up with noise that is unpleasing to the ears.

A tuned EQ should target the specific notes or overtones of those notes you want to boost. If say a vocalist doesn't have allot of high frequency in his voice and the mic he uses tracking rolls off at 12K, you only add noise by trying to boost frequencies above that mics rolloff limit.  

If the tracks lack high end harmonics, and collectively produce a muted recording, then you should start over and re-evaluate the recording techniques and front end gear making those tracks. Its the only real fix that exists. Cutting mids and lows to let more highs remain is no different than boosting high only.

From there, creating high frequencies that don't exist in a mix can only be done through synthesis. There are some harmonic generator plugins that will take the root notes and create harmonics of those root notes. The process isn't much different than what an octaver circuit does in cutting the sine waves into halves, quarters, etc and reshaping them to look like harmonics of the root note, the same way an instrument does this naturally. Over use will sound harsh and plastic. Its not a do all fix all method of fixing a bad mix. Mild use can add some shine to a mix poorly tracked. It will also avoid boosting all the noise that can exist in the upper frequency bands caused by analog preamps, line noise, ghost reflections etc. 

When you get to the point tracking where you don't need EQ to make the tracks sound good, then EQ becomes a simple artistic tool used to surgically highlight a part so it stands out above the others (or gets pushed back into the sound scape letting the other parts shine) vs.. Being an overworked medical tool resuscitating tracks that were never in focus to begin with. 

[Lee Knight]

Cause it sounds bitchen. It does. That little dip just prior to the steep incline of boost, while I suppose an artifact of some analog designs, brings a certain emphasis to your boost frequency. You see the same sort of behavior in the Waves Ren EQs. I love the way it sounds. That behavior being particularly useful at smaller shelf boosts. 

 

Think of it like a resonant low pass filter. Say you set the roll off at 3k and turn the resonance up. Now you have a little peak right where you want to begin rolling off. Why??!?!! Cause it sounds bitchen.

[veracohr]

Maybe it incorporates one of the filter types that have lobes which rise back up after a zero, like the Chebyshev Type 2 or Elliptical shown below. That dip could correspond to the zero?

 

 

[Zooey]

McDSP filters have that same response (and you can choose how much, if any, of it you want).