As we mentioned in the previous tip (QGM 18), vocal sibilance, that annoying “sss” or “sh” sound, can play merry havoc with our ears. However, if we cut EQ in the high frequency range where sibilance occurs, which is around 6kHz and upward, we run the risk of removing the bright, open sound of the vocals, rendering them dull and lifeless. We need a means of selectively removing sibilance. Back in the good old analog days of stone cutting tools and therapeutic bloodletting, we used a somewhat complicated combination of EQ and compressor tied together via the compressor’s sidechain.
Since the net abounds with tips on this archaic technique, we’ll just skip ahead to the DAW world, where all we have to do is find the sibilance in the vocal track and automate our EQ plug-in to reduce the offending frequency when it occurs. This is similar to what the EQ/compressor sidechain setup does, but instead, we can be surgically accurate. As you find the sections where sibilance occurs, boost the gain of the high frequency range and sweep through the frequencies until you find the one where the effect is most pronounced. The loop playback function of the DAW really comes in handy here for repeatedly listening to a passage while searching for the right frequency.
Once you’ve isolated the fiendish frequency, set a very narrow “Q” or octave range/bandwidth, and cut until the sibilance becomes civilized. Set your EQ plug-in to read automation, and have it reduce gain by that amount, but only during the sibilant portion of the word (which coincidently, will be a word starting with the letter “S”) and then return gain to normal.
With DAW software, you have the luxury of dedicating as many EQs to your vocals as you want, so that one can do tone-shaping while the other merely reduces sibilance. Or, if you have a 7- or 10-band EQ plug-in, you can dedicate one band to de-essing.