The worst note I ever got on a mix was "it's good, man." The second worst was "the chorus needs more energy."
The first one told me nothing. The second one told me something was wrong and then pointed me in a direction that turned out to be the opposite of the fix. I spent an evening pushing the chorus harder: louder drums, a doubled guitar, more compression on the bus. The next version came back with "hmm, now it feels kind of flat."
It did. I'd made the whole chorus louder and it lost the lift it had coming out of the verse. The problem was never the chorus. The verse was too big, so the chorus had nowhere to go.
That evening taught me more about collecting notes than about mixing.
A note has two halves
Every note a listener gives has two parts welded together. There's what they heard, and there's what they think should be done about it.
The first half is almost always right. If someone says the chorus doesn't hit, something real happened in their ears at that moment. They're reporting an experience, and nobody knows that experience better than the person who had it.
The second half is a guess. "Needs more energy" is a prescription from someone who isn't sitting at the session, doesn't know what's on the tracks, and is describing a symptom in the only words they have. A drummer will tell me the kick is buried when the real problem is the bass sitting on top of it. A singer will ask for more vocal when the vocal is loud enough and the guitars are just crowding the same frequencies.
So I split every note in half. I keep the symptom. I hold the prescription loosely, as a clue to where the listener's attention was, not as an instruction.
Ask where, not what
The single biggest change in the notes I get came from changing the question.
I used to send a mix and ask "what do you think?" That question invites a verdict. People feel pressure to say something smart, so they reach for mix vocabulary they half know, and I get "more punch" and "warmer" and "the low end is muddy" on a song with almost no low end.
Now I ask for locations. Tell me the moment you stopped paying attention. Tell me the moment you leaned in. If something bugged you, give me the timestamp or the section, and you don't have to explain why.
That's it. No adjectives required.
A timestamp is the most useful thing a listener can give me. "1:42, I got bored" is gold. I can go to 1:42 and find out why. Usually it's obvious the second I look: the second verse repeats the first verse's arrangement exactly, or the guitar part that was interesting for eight bars is still going at bar thirty two. The listener couldn't have named that. They didn't need to. They just had to tell me where it happened.
One round, then wait
The other thing I changed is how often I ask.
I used to send every bounce to everyone. A version a night. Notes trickled in from different people about different versions, and I'd be fixing a comment about Tuesday's mix in Thursday's mix, where the thing they heard had already moved.
Now I work in rounds. I get a version to a place where I'd be willing to call it done if I had to. That version goes out to a small group, the same version to everyone. Then I don't touch the song until the notes come back.
Two good things fall out of this. First, everyone is commenting on the same audio, so I can compare notes. When three people independently flag the same bridge, that bridge has a problem, even if all three describe it differently. When one person flags something nobody else heard, I listen for it, but I don't rebuild the song around it.
Second, the waiting period is its own listening tool. By the time the notes come back, I've been away from the song for a few days. I hear the mix closer to how they heard it. Half the time, I already know what they're going to say.
Write the notes down where the song lives
I keep every note with the version it was about. The song, the version number, the timestamp, who said it, and the exact words they used. Not my summary of what they meant. Their words.
That record has saved me more than once. A note from three versions ago that I dismissed turns out to be the same thing a different listener is saying now in different language. Or I notice that one collaborator's notes almost always land on vocals and another's almost always land on the drums, which tells me as much about their ears as about my mix. I start weighting notes by who's giving them and what they reliably hear well.
It also keeps me honest when I decide not to take a note. If I'm overruling someone, I want to see the note in front of me and know I made that call on purpose, not that I forgot it.
The note I give myself last
After a round is collected, I make a short list before I open the session. Not a list of fixes. A list of moments: 0:38 lift is weak, 1:42 attention drops, 2:55 the ending arrives too fast. Each one is a place, and each one gets a fix I decide on at the desk with the faders in front of me.
The listeners told me where. My job is to work out why.
The best collaborators I've had understood this without being told. They never tried to mix the song through a text message. They told me what happened to them while it played, and they trusted me to do the rest. Now I ask everyone for exactly that, and the notes are better for it.