OPEN a week in, and what comes next

OPEN went out a week ago today, and some of you wrote back.

daniel wrote that the cloud algo “sounds really damn nice… sound design heaven, clean as well.” eckart called it the “most interesting one in a long time.” thank you, both of you, and everyone who picked it up.

the email i learned the most from was the one that stung. a customer who has owned WIDE since april wrote that the OPEN manual lost him completely. “it could have been in french.” then he rewrote it himself, line by line, and sent me the rewrite.

he was right. so all six manuals now start differently: what the plugin does in plain words, a quick start, then what you hear when you turn each knob (with audio you can play). the numbers are still there, further down, for anyone who wants them.

what comes next

OPEN had to pass a blind test before it shipped. my plugin against the ones people already own, levels matched, order shuffled, and i vote without knowing which is which. there’s a danish saying, man skal ikke skue hunden på hårene (don’t judge a dog by its fur). a nice interface and a good first listen are the fur.

now the other four go through the same thing. i want to take all five to the next level in one big update, so most of my days have gotten pretty lab-heavy: blind tests, measurements, listening against references, and a newer version of the framework underneath. nothing ships until it holds up in that test more than once.

no date yet. i’d rather ship it late than ship it untested.

this week’s film

the paper everyone quotes about music and dopamine is from 2011. it’s careful work, and fifteen years later nobody has independently run that exact result again.

the one experiment that tested cause is smaller and stranger. raise people’s dopamine with a drug and they enjoyed the music more and would pay more for it. lower it and both went down. how they said they felt didn’t move.

a result counts once someone runs it again. same reason my blind test runs more than once.

what to take into your next session

ask three questions of any mixing rule that comes with a number. who measured it, on how many people, and has anyone run it again. if two of the three have no answer, it’s a habit someone wrote down, and you can treat it like one.

stop using your own chills as the test of a mix. around one in ten people report getting them rarely or never. among people who said they got chills, visible goosebumps showed up in only 40 to 57 percent of them, depending on the study. check the track against a favourite reference, the next day.

when you play a draft for someone, ask two separate questions. “would you play it again?” and “how did it make you feel?” in the one causal experiment, dopamine moved how much people enjoyed the music and what they’d pay for it, and left how they said they felt alone. treating “play it again” as the first kind of answer is my step, the paper doesn’t test it. but a friend calling your track emotional and a friend putting it on repeat are two different results.

and if one of my manuals ever reads like french again, tell me.

kernaudio.io