mike_mccue
However, It just doesn't seem like a "musical" solution.
It's not, it's a
mixing solution. You still hear the notes, they're still there, if someone plays an A chord you'll still hear an A chord. However, you have more options with the spectral balance among instruments during the mix. It may only take pulling back the notches by 2dB to make a beneficial difference.
Writing parts with chord voicing that compliments a song isn't a bad idea either.
But you really don't know how the spectral distribution of that voicing will play out in the mix. If someone said "well maybe the fundamental of the root key will clash with the other fundamentals so I won't write a guitar part using any fundamentals," that would not be musical at all. I also don't think most arrangers do a spectral analysis of an acoustic guitar to find where the body resonances boost certain frequencies. I recorded Linda Cohen's Contreras nylon string guitar several times, which had a HUGE body peak at exactly 220Hz. I ended up having to notch it by about 9dB (!) just so I could set decent levels. But it sounded great live.
What I'm describing doesn't eliminate anything, it just gives precise control over additional level and frequency-related parameters not incorporated in typical EQs.
I don't care if anyone uses this technique or not any more than I care whether someone records instruments that don't oversample at 96kHz. I'm not selling anything, I'm simply presenting a technique that some people will hopefully find as useful as I have when facing certain mix issues. It's not necessary to overthink this, it's basically just a specialized variation on parametric EQ.