So, continuing on the EQ theme I thought we’d think about terms and types of filters you might run across.

1) Shelving EQ’s: A shelving filter takes all the frequencies above or below a specified starting frequency and attenuates or amplifes them (i.e. cuts or boosts them).  Many pro-sumer consoles, such as a lot of Mackie boards, have shelving filters as the top and bottom EQ’s at something like 85 Hz on the bottom and 12kHz or 15kHz.  These filters are useful, but they are even more useful when you can specify the frequency at which the shelving begins, which isn’t the case for most all boards except fairly top-of-the-line consoles.  With a sweepable-shelving fitler you can boost that string definition on the electric bass and get that really nice “click” from the bass drum.

2) Sweepable-mids: A sweepable-mid filter (or semi-parametric) allows the sound operator to chose a center frequency from which the attenuation or amplification of frequencies extends both upward and downward in the frequency range by a fixed width.  It is common for mid-line sound consoles to have one or two of these types of filters.  Normally these filters are also limited in range but will overlaps.  So one filter will be able to sweep from, say, 45 Hz to 1 kHz and the other from 500 Hz to 8 kHz.  These filters are really where usable EQ starts.  With these you can help eliminate feedback at that certain frequency or remove the ugly mids from that cheap youth-group-worship-leader’s guitar.

3) Parametric: A parametric filter is able to do what the sweepable-mid filter can, but can also adjust the width of frequency range being affected.  This allows much more precision in the frequencies affected.  These filters also normally have overlapping ranges.  A “fully” parametric filter can affect any frequency in the audible range.  Fully-parametric filters are normally found on system processors, digital boards, or outboard gear.  The nuances these filters allow speak for themselves.  Smoothing out that Oboe sound, taking large bites out of the trumpet’s bark, boosting that perfect sub-woofer moving frequency from the kick drum, or optomizing the entire sound system are all where the parametric EQ shines.

So these are the basic tools.  There are, of course, more specialized filters.  Sweepable-notch filters are sweepable filters with a very small, fixed width and are good for feedback problems.  A graphic equalizer is a series of fixed-notch filters (normally every 1/3 octave) and is used for system optimization and possibly on a particualrly difficult lapel or podium microphone.

The next post is going to go over frequency ranges and terms like “box-y” or “tubby” or “harsh.”

So, I’m going to be geeky and write a little about “EQ”.  Hopefully this will be the first in a series of posts, but no promises or timetables.  ”EQ” is the part of the sound-mixing console or other outboard gear where we get to use filters to affect different frequency ranges to change the way a particular input sounds (or is colored, shaped, etc.)  Let’s start at the top, which in this case is at the bottom:

The high-pass filter explained: Although this tool is normally grouped physically with the pre-amps on most mixers, this filter is an EQ function.  Most high-pass filters on pro-sumer sound boards are simply in or out and are fixed to start affecting frequencies at either 85 or 100 Hz.  More professional boards allow you to choose a frequency up to around 400 Hz and on some digital boards you could pick a frequency going much higher than that.  This filter allows frequencies above the stated or chosen frequency to pass unaffected while attenuating frequencies below the stated or chosen frequency (normally by -12 or -18dB per octave, sometime more steeply).  The standard is that the stated frequency is the point at which -3db is reached and then the filter shots down exponentially at the ration from there.

The high-pass filter used: The high-pass filter is used 1) to remove unwanted air, handling, or clothing noise so the roar of the air conditioner or the bump of a mic-stand don’t get amplified like crazy. 2) A high-pass filter is used remove low-end all-together from signals that would only be duplicating it.  For example if you close-mic a drum kit you can high-pass the overheads to elimate excess bass that is already being picked up by the close mics on the tom-drums and kick, etc. 3) High-pass filter are kind of like a broad sword when it comes to extremely low-end feedback removal.  Using a good high-pass filter at just the right point can really help lapel and pulpit mics that often need to be almost in front of the speaker clusters because of their use on stage.  An example of both #2 and #3 would be the use of the filter on a reverb effect that has too much low end reverberation.  You already have the low-end from the instrument channel and high passing the reverb might allow you to use more without worrying about feedback in that cello microphone.

On a related note.  Low-pass filters do the opposite of high-pass filters (imagine that!). And although you won’t find them on sound consoles, they are fairly important for the audio-post world.   It is my understanding that because of a phenomenon deemed the “Nyquist Effect” signals that are being digitized need to be extremely low-passed at a frequency that is below half of the digital sampling rate.  If this isn’t done weird sounding ‘artifacts’ can be recorded in the digital signal.  So if you are saving space on that small mp3 podcast files by using a sample rate of say 22,050 Hz you should be low-passing the audio around 10,000 Hz or lower.