I discovered a (very) old analog console in storage at my workplace, waiting to be verified as operational before being donated to a local audio program. In the course of the testing, I was playing with the channel EQ and noting the response on the analyzer. It occurred to me that unless someone bothered to hook their console up to a test rig, they might go the entire life of the console without ever seeing the curves created by the EQ. This is perfectly fine – our ears, not our eyes, should be the guiding factor when working with EQ.
But what if the EQ isn’t doing what we think it is? Maybe a 2 kHz boost is actually centered at 1 kHz, for example. Wouldn’t it be in our best interest to know that, to avoid throwing off our carefully calibrated ears?
The most critical skill in the EQ toolset is developing the mental associations between tonal ranges and the frequencies in question, so it follows that those associations could be skewed by extensive use of an EQ that’s doing something unexpected under the hood. (Classic example: the Pulteq EQP-1A, which has both boost and cut knobs for the same frequency. In reality, the boost and cut filters are slightly offset in frequency, and so can be used simultaneously, but you wouldn’t know that from the front panel.)
Table 1 shows the parameters for the three channel strip settings in Figure 1 (located at the beginning of this article) as represented by the channel strip silkscreen labels. The first thing to do is to look at the curves that are actually produced by these settings, with the help of an analyzer. Note that a rough result could be obtained using pink noise and a real-time analyzer (RTA), but a much better view of the goings-on will be provided using a full-blown fast Fourier transform (FFT) analyzer thanks to the higher frequency resolution.
The analyzer traces in this piece were produced with Rational Acoustics Smaart, but for the purposes of this experiment any analyzer will do. (ARTALABS offers a free demo of its ARTA analyzer software that will work nicely. Thanks to Pat Brown for that tip.)
Figure 2 provides the curves produced by the EQ settings I showed at the beginning of the article. Compare these to your sketches – how close did you come?
The general shapes are probably as expected, but the details might be surprising. For example, the 12 kHz high-frequency (HF) band, when set to maximum attenuation, actually reaches -3 dB at 1.2 kHz. Likewise, the 80 Hz low-frequency (LF) band, at max cut, reaches -3 dB at 580 Hz.
These gentle slopes are common with analog consoles, but it certainly pays to remember that HF and LF cuts and boosts may have significant effects through so much of the spectrum. Figure 3 shows how the 80 Hz LF EQ band extends its influence well into the midrange. By adjusting the 80 Hz control on this particular console, I’m also emphasizing a lot of content in the next higher octave as well, and knowledge of that might lead me to change my approach.
For instance, if I’m dealing with a 200 Hz resonance I don’t like, boosting the 80 Hz band might exacerbate the issue. On the other hand, after seeing the curves, I know that the same LF band would be extremely effective at dealing with proximity effect on a vocal mic.
Viewing these curves on a screen as the controls are adjusted can make all the difference when it comes to an intuitive understanding of what’s happening when EQ is applied. Note that our ears should always been the final arbiter when making EQ decisions, but a good grasp of the filter shapes we can create using consoles goes a long way towards the effective use of equalization. More knowledge about what these filters are actually doing allows me to be more focused with my application of them. As mentioned, digital consoles come with this convenience built in.