Musician performance automation
What an individual player does, captured so it can be reproduced rather than re-negotiated.
The take was fine. You just weren't all the way in it.
Arrangement Based Automation — the production method. One of two doors into ABA Media Platform, a one-person media studio in Greeley, Colorado. The other door is the tool it produced.
Some part of you was still on the routing, or the level, or whether that edit was going to hold.
Nobody plays badly because of a patch bay. What happens is quieter than that: the performance comes out competent and a little guarded, and nothing in the session log says why.
Arrangement Based Automation is a way of arranging the work so that does not happen.
What it is
Settle the technical decisions before anyone plays.
Get the technical decisions settled and automated before the creative work starts, so that when the musician plays, nothing technical is competing for their attention.
That is the whole method in one sentence. The rest is the discipline of actually doing it — moving analytical work forward in the schedule rather than letting it arrive, as it usually does, in the middle of a take.
Why it works, and why not for the reason you have heard
It is about attention, not hemispheres.
The familiar version of this argument says the analytical left brain interferes with the creative right brain. We do not make that argument, because it is not supported. The hemispheres are not separate personalities and they do not take turns; they work together, and the pop version of the story has been contradicted for years.
The defensible reason is simpler and harder to argue with: analytical work and creative judgment draw on the same limited pool of attention. Every open technical question is a withdrawal from that pool, and it is being made at the exact moment you need the balance.
This distinction matters more than it looks. A method resting on a claim that does not hold collapses the day someone checks, and takes the parts that were true down with it.
The instrument
A question to ask before the next session.
This one costs nothing and you can run it tonight, on a session that has nothing to do with us.
Which technical decisions are still open?
List them before the performer arrives. Not the ones you expect to go wrong — the ones that are simply undecided. Routing, levels, headphone balance, what happens to the take afterwards.
Every item on that list will be answered during the take, by somebody, using attention that was supposed to be going somewhere else. Answer them first and the list is empty when it matters.
That is the entire principle, and it works whether or not you ever automate anything. The automation is what makes it repeatable rather than a good intention.
What automation is holding
Three layers, so the performance can be performed again.
Section automation
The same at the level of a section, so a group moves together without a person riding it live.
Solo performance automation
The line that has to breathe differently from everything around it, held separately for exactly that reason.
With those three held, the last pass stops being a mixdown and becomes something closer to conducting: live control over sections and soloists, decided by ear, on a performance that can be performed again tomorrow the same way.
What the milestones are for
Every job ends in the same conversation.
Not the mix. The invoice. A number agreed months ago meets a number arrived at since, and the gap between them gets argued by two people who no longer remember which decision caused it. That conversation is not a billing problem. It is what happens when a job is settled once, at the end, instead of nine times along the way.
The way the job runs now
One estimate. One invoice. One argument.
01Carried
02Carried
03Carried
04Carried
05Carried
06Carried
07Carried
08Carried
09Carried
Then you send the invoice.
Everything that went wrong across the whole job arrives in one number, months after the decisions that caused it, and it is argued with a client who has already spent the money.
The same job, run in milestones
Nine settlements. Nothing left to argue.
01Signed & settled
02Signed & settled
03Signed & settled
04Signed & settled
05Signed & settled
06Signed & settled
07Signed & settled
08Signed & settled
09Signed & settled
The invoice is a formality.
Every hour that missed the estimate, over or under, was documented in red and settled at the milestone that caused it — while both people still remembered why.
The bars are a shape, not a measurement. There are no figures on this page because there is no average job, and a made-up number on a drawing about estimate integrity would be the exact thing the method exists to prevent.
Not one of those nine is an argument, because none of them is a surprise. Each milestone closes on a verification task carrying three signatures — client, producer, ABA engineer — and every hour that missed the estimate, over or under, is documented in red on the signed milestone. Those overruns are reconciled before the next milestone opens. The estimate is checked against the actual nine times over the job, instead of defended once at the end of it.
Not published yet
The stage list is not on this page, and that is deliberate.
The Milestone Production Process — the actual stage-by-stage sequence, with what enters each stage and what stops it — exists as working documents and has not been written up for publication. It is not on this page because it is not finished, and a summary written to fill the space would be the kind of confident approximation this company keeps a linter to catch.
When it is published it will be published whole, the way the rulebook is. Until then this page says what is settled and names what is not.
Where to next
The instrument, the rulebook, and the conversation.
The tool the method produced
AudioVibes measures a finished master against the target it was made for, and names the difference. It runs in the browser, on the machine in front of you.
The rulebook
Every claim on this page and the approved sentence it came from — including the one still open, and the one that was withdrawn and left on the page where it used to be true.
Who signs it
One owner and six AI workers, working to that rulebook. The signatures on a milestone are only worth what the name behind them is worth, so the name is on the page.
An hour, your own job, your own numbers. Milestone 0 is that conversation, and it is the one that decides whether the rest is worth doing.