Kevin's approach to EQ was philosophically simple: more is more. If the vocals sounded thin, boost the mids. If the bass felt weak, boost the lows. If the highs lacked sparkle, boost the highs. By the end of soundcheck, every band on the 31-band graphic EQ was above the center line. Most of them by 6 dB. Some by 10.

The net effect was approximately 8 dB of broadband gain added before the amplifiers. Kevin had, through the medium of equalization, simply turned everything up. The PA was now running 8 dB hotter into the same room with the same microphones at the same gain-before-feedback threshold.

The feedback started during the first verse. A beautiful, crystalline 2.5 kHz howl that built from a shimmer to a scream in about two seconds. Kevin's response was to pull down the 2.5 kHz band on the GEQ by 6 dB. This helped. But now 1.6 kHz started ringing, because every other frequency was still boosted into the danger zone.

By the end of the show, the GEQ looked like a mountain range — peaks at every frequency Kevin had boosted, valleys where feedback had forced his hand. The frequency response was worse than when he started.

The Moral: EQ should primarily cut, not boost. If everything needs boosting, the problem is gain structure, not tone. SonaVyx's AI diagnostic analyzes your transfer function and recommends subtractive corrections first — because the best EQ move is almost always taking something away.