Here is a session I have lived in more times than I want to admit.
Two rhythm guitars, panned wide. A piano or a pad underneath. Everything in tune, everything played well, all of it recorded on a decent day with a decent mic. And the whole middle of the mix is a gray smear. You can hear that something is happening in there. You cannot hear what.
So you do what everyone does. You open an EQ on guitar one and pull three dB out around 300. Better. You do it to guitar two. Better again, sort of. You notch the piano at 400 because that is where the internet says mud lives. Now everything is thinner and it is still a smear, just a quieter one. Two hours later you have four EQs doing surgery and the section sounds worse than the rough.
I lost a lot of nights to that before I understood what was actually happening.
EQ moves energy, not information
An EQ changes how loud a frequency range is. That is the entire job. It does not change what note is there.
If your two guitar parts are both playing an open C chord in first position, they are playing the same notes, in the same octave, with the same top voice. The information is identical. You can pull 300 out of one of them all day and you will never make it a different part, because it is not a different part. You are just making one copy of the same thing quieter in the middle, which is the definition of thin.
Separation in a mix is mostly not an EQ outcome. It is a note-choice outcome. Two things sound distinct when they are actually distinct.
The test that takes thirty seconds
Solo the two parts that are fighting. Bypass every plugin on both. Play eight bars.
Now name the notes. Not the chord symbol, the actual notes, top to bottom. On guitar one you are playing C, E, G, C, E. On guitar two, capo off, same shape, you are playing C, E, G, C, E.
If the top two or three notes of each part are the same pitches in the same octave, stop reaching for the EQ. You do not have a frequency problem. You have one part recorded twice.
I do this before I open a single plugin now. Most of the time the answer is embarrassing and it takes five minutes to fix.
Fix it on the fretboard
This is where guitar is a gift, because the same chord lives in six places on the neck and every one of them sounds like a different instrument.
Capo the second part. Capo on 5, play a G shape, you get the same C chord with a completely different top voice and a different set of open strings ringing. The parts stop overlapping and you did not touch a fader. This is the single highest-value move in the list and it takes about ninety seconds.
Take the low strings off one of them. Play the triad on the top three strings only, up the neck. Let the other guitar own everything below the third string. Now one part is the body and one part is the shimmer, and they are not competing for the same 200 to 500 Hz that you were about to carve out of both.
Take notes out entirely. Two-note shapes are underrated. The third and the seventh tell you what the chord is. Everything else is reinforcement, and reinforcement is exactly what makes a section gray. A part with two notes in it has room around it by construction.
Move one part in time instead of in pitch. If both guitars are strumming eighths, one of them can hold whole notes, or hit the downbeat and get out. Separation is a rhythmic axis too, and it is the one people forget.
The pad or the piano gets the same treatment. A pad playing a full five-note voicing in the middle of the keyboard is the most common mud source in independent records, and the fix is not a high-pass at 300. The fix is playing three notes instead of five, or moving the left hand down an octave and leaving the middle alone.
Why this actually matters, and it is not the guitars
The lead vocal lives in a range. For most singers that is roughly 200 Hz up through the low thousands, and it is not negotiable, because that is where the human voice is.
Every part you stack in that same window is asking the vocal to work harder. When you EQ your way out of this, you are shaving the guitars down to make a hole for the voice, and the guitars end up sounding like a phone speaker so that the vocal can sit in a gap you manufactured.
When you fix it with voicings instead, the hole is already there. Nobody is playing in it. The vocal does not need a carve, it needs an empty room, and the arrangement is what builds the room.
What EQ is actually good for
I am not anti-EQ. I use it on almost everything. But it is a tool for problems that live in one source.
The boxy resonance from tracking an acoustic two feet from a wall. The honk a particular mic has at 1.2k. A cymbal wash that needs a shelf. A low rumble from the amp fan. Those are real, and no amount of clever voicing fixes them, because they are not about notes.
The rule I use: if a single track sounds wrong soloed, that is an EQ problem. If two tracks sound fine soloed and bad together, that is an arrangement problem wearing an EQ costume.
The part that is hard
The reason we reach for EQ is that it is available at 11pm and re-tracking is not. The performance is already in the box. Opening a plugin feels like progress and picking the guitar back up feels like going backwards.
But retracking a rhythm guitar with a capo on the fifth fret takes five minutes. I have spent three sessions on EQ moves that a capo solved. That math never once came out in favor of the plugin.
Fix it while the mic is still up. Your mix engineer, who is also you, three weeks from now, will not have that option.