When we talk about live sound, it is very common to talk about equalization: EQing an artist at soundcheck, EQing the sound system… However, it is very important to use equalizers as little as possible live. In fact, more and more often we see that the graphic EQ at FOH (FRONT OF HOUSE) positions is flat, something that was unthinkable not so long ago.
The equalizer is the device that lets us modify the frequency response curve of an audio system. This is done using filters: by adjusting their controls, we alter the incoming signal and thus modify the response of the sound system until we achieve an ideal response for the venue and the type of music in question.
What we are after when using equalizers is a response curve that is as neutral as possible, that is, one in which energy levels are distributed equally across each octave.
It is used to correct acoustic defects or, for artistic reasons, to modify the response of a particular instrument. In the latter case, it should not be overused.
Frequency spectrum. Frequency response
The frequency response curve of an audio installation is the graphic representation of the sound intensity at a point in the room relative to the frequency being emitted. It is obtained by placing a sound level meter in the listening area along with a spectrum analyzer that can display this curve graphically.
If the resulting plot is a horizontal line, it means there is no loss or gain in sound intensity at any frequency. Usually, some frequencies will show a loss of signal and others a gain. The equalizer must correct these variations by applying the inverse of the initial frequency response. In this way we will try to achieve a flat final response.
The frequency response of a listening room varies greatly depending on the furniture it contains and how it is arranged, causing absorption or reflection depending on the materials.
If the energy absorbed by the walls is small, sound intensity builds up quickly, reaching pressure levels above those that would be created in free space under the same conditions, while also distorting the sound message.
If, on the other hand, absorption is high, the level reached in the room is on the order of that obtained in free space, giving the room a flat frequency response.
The frequency response is also affected by the fact that some of the sounds reaching the listener do not come directly from the sound source. These are reflected or indirect sounds.
A very important consequence follows from this: we will not have the same ratio of direct to indirect sound at every point in the room. This is because the waves travel different paths and take different amounts of time. Therefore, it is only possible to equalize one area correctly. The changes in sound caused by the presence of furniture and surfaces in a venue are:
- Alteration of the temporal structure of the sound by reflections arriving at different times.
- The existence of room modes causes some frequencies to be boosted more than others, while others cancel out.
- An increase in level due to the arrival of the reflected wave, which subjectively adds to the direct wave, producing a higher sound level.
- The equalizer's response must be the inverse of the room's. That way, it will boost the frequencies that are lost and attenuate those that are excessive.
In practice, to equalize a room or acoustic space, you don't need to make many corrections to get a fairly flat acoustic response. We call a response flat when any areas that deviate do so by a maximum of 2 decibels from the ideal line. These deviations will be barely perceptible to the human ear.
General equalizer parameters
The center frequency (Fc) is the frequency value on which each filter acts. It corresponds to the value at which its action is greatest.
Bandwidth (BW) determines the width of the working area. It indicates the extent on either side of the center frequency, F1 and F2. This value is taken at the point where the level reaches -3dB. A large value indicates action over a wide range of frequencies.
BW=(f2-f1)
The Q factor, also called selectivity, is the ratio between the center frequency and the bandwidth; it indicates the slope of the filter's response curve, also called its bell. The lower this value, the more uniform the filter's action within its bandwidth.
Q=Fc/BW
Example: If Fc= 1400Hz, F1=1000Hz and F2=2000Hz
What would the BW be?
BW=f2-f1=1000Hz
What would the Q be?
Q=Fc/BW= 1.4
From these formulas we can see that the higher the Q factor, the narrower the bandwidth. In our example the bandwidth is one octave; if the Q were 0.7 the bandwidth would be two octaves, while it would be only half an octave with a Q factor of 2.8. Gain is the amount of boost or cut the filter applies to the signal. It is expressed in decibels for each filter and generally ranges within ± 12dB.
The human ear does not respond linearly to frequency but logarithmically: that is, the step from 100Hz to 150Hz (a 50Hz difference) sounds identical to the step from 1000Hz to 1500Hz (a 500Hz difference). That is why the piano keyboard is divided into octaves, in which the frequency of the first note of one octave is double that of the first note of the previous octave.
Types of equalizers
There are two main types of equalizers: the graphic equalizer and the parametric equalizer:
The graphic equalizer
It gets its name from the layout of its faders, which are arranged in a way that lets you visualize the compensation applied.
There are graphic equalizers with anywhere from five to 31 controls. The most widely used in professional audio is the third-octave equalizer, which has 30 control points. Graphic equalizers let you boost or cut the signal at a given frequency by about 6 to 15dB.
It consists of an input stage feeding a bank of parallel filters, whose number varies depending on whether it is an octave, half-octave or third-octave equalizer, all of which feed into an output summing stage. These are selective band-pass filters, each assigned a different working band, with their center frequencies spaced logarithmically so that together they cover the entire audible spectrum.
Each filter has a gain control to cut or boost the frequency band it acts on. In the center position nothing happens; that is, it stays at 0dB. Since the faders sit side by side in sequence, the imaginary curve joining their positions corresponds to the response the equalizer applies to the signal.
ISO. This standard is used to unify equalizer design criteria, since if we look at the table, the chosen frequencies are not the result of dividing each octave into three or, what amounts to the same thing, multiplying a given frequency by 2^(1/3); rather, they are frequencies chosen for that ISO standard.
The parametric equalizer
These control the three fundamental parameters: bandwidth, center frequency and signal amplitude. They are more complex to operate, since all the parameters have to be adjusted.
Ideally, a parametric equalizer should have four frequency bands to work with: low frequencies (20 to 250Hz), low mids (250 to 2,000Hz), high mids (2000 to 4,000Hz) and highs (4,000 to 16,000).
If we modify any frequency, we also modify its phase response. So imagine heavy EQ on several frequencies when doing sound for an event: the system's phase response would be severely altered.
This is due to the effect filters have on phase; remember that a graphic equalizer is nothing more than a set of band-pass filters.
That is why, if the sound system is properly tuned, there will be no need to touch the graphic EQ, and for this reason we see that most graphic equalizers at live concerts are completely flat.
Practical use of equalizers
Classic use of a graphic equalizer to correct a sound system
To properly equalize our sound system according to the acoustics of the venue we are in, using the classic method, ideally we will need the following:
-1 CD with pink noise (Test)
-1 spectrum analyzer
-1 measurement microphone or, failing that, an omnidirectional condenser mic with as flat a frequency response as possible across the whole spectrum.
-A stereo 2X31-band equalizer
Once our graphic equalizer is properly connected in series between the console and the power amplifiers, either directly or through the "inserts," we play a pink noise CD (if our console doesn't have its own generator) and, with all the console's EQs bypassed, we look at the result on our spectrum analyzer.
We simply correct the frequencies that our analyzer shows as excessively attenuated or boosted; we must keep in mind that when we lower the problem frequency on the equalizer, the adjacent bands above and below will also drop, so one solution is to raise these adjacent frequencies slightly. Although measuring equipment is very reliable, it is better to listen to the result, since in most cases a flat curve is not synonymous with a perfect response, so the best thing is to play a CD we know well and listen to often, to compare the acoustic correction we have made with the analysis our ears make, since obviously the information reaching our ears is much richer than that provided to the spectrum analyzer.
NOTE: The correct and most complete way to tune a sound system in a venue would be with reference to the phase response of the components and how they combine.
This would require software or hardware capable of performing transfer functions.
Practical use of a parametric equalizer to correct musical instruments
It should also be noted that these EQ settings depend a great deal on the artistic taste of the sound engineer and/or the musician, who for certain musical styles look for exaggerated levels at certain frequencies.
In the case of a kick drum, the frequencies produced go down to 40Hz, so to better control the attack we should boost or cut around 63Hz; in fact, many microphones designed for kick drums have a slight peak at this frequency. Some of the most annoying frequencies are usually in the 300 to 400Hz area.
Electric guitars cover a wide frequency range starting at around 80Hz. If they sound too dull we can cut around 200Hz, while if they are too harsh we will cut around 1000Hz; for acoustic guitars we should watch the area between 200 and 500Hz.
The singing voice contains a wide range of frequencies, with the frequencies between 2 and 4KHz being where we get the most intelligibility. Remember that, according to the Fletcher-Munson curves, this part of the spectrum is the one our ears are most sensitive to. We can see how many vocal microphones have a peak from 5KHz up.
Frequency range of each instrument within the sound spectrum
40-50Hz
This is where those lows so fashionable in dance music live. Here we should mainly put the kick drum to work and avoid placing vocals in these frequencies.
70-100Hz
Now, besides the kick, we have the bass; the important thing is to differentiate these two instruments.
200-400Hz
In this frequency range we can shape the sound of the snare; we can also add warmth to vocals or fill out the guitars a bit.
400-800Hz
We can clarify the bass in these frequencies; this range also contains the snare's first harmonics, which are very important for defining its attack.
In this frequency range we can also reduce the ringing of the toms.
800-1000Hz
This range is very important for vocals, as it gives them more power; it will also give the bass clarity and the kick more punch.
1-3KHz
Here it is worth boosting between 1 and 2 kHz to define guitars and basses; obviously, when we reach this very sensitive part of the ear's range, we should pull the vocals down around 2.5KHz.
3-6KHz
This range is quite critical for vocals. Basically, as mentioned above, we can lower the level around 3KHz, but we should also do so around 6KHz, since that is where the sibilance in speech lies; it follows that the area between 3 and 5KHz will give us more clarity.
6-10KHz
From these frequencies up we can open up our mix, since the sound has more air. We will also notice how we give acoustic guitars more freshness.
10-16KHz
Here we should forget about the lowest instruments, simply cutting these frequencies for them (kick, bass,...) with a low-pass filter, for example. In these frequencies we can give more air to vocals and cymbals too.
All these tips are not strict rules and can vary slightly with each instrument. Moreover, it can often happen that we EQ a particular instrument on its own and, when we mix it with the other instruments, it doesn't sound the same and we have to touch up the EQ. We should test at various levels (remember the Fletcher-Munson curves) and never spend too long on a single sound, since our ears tire very quickly of the same sound repeated many times.
On the other hand, remember that for live sound EQ, in practice, there is no need for a lot of work on frequencies above 16KHz, since, as we have studied before, high frequencies are attenuated more with distance than low frequencies. We will also use this theoretical knowledge to improve our mixes, so that we can push a sound source back or bring it forward in our final mix. And let's not forget panning to improve the mix by placing each instrument in its correct sonic plane.
