Using blind A/B to check your work

Something I know about myself is that I’m extremely susceptible to cognitive biases. I mean, we all are, right? None of us is immune from the human condition. We want to believe things.

Two of the things that I most want to believe are that I know things and that I have good ideas, specifically in regards to choosing processing for sounds in my mixes. And, I mean, I do! I think I do. But how can I be sure?

Blind A/B testing!

Blind A/B testing is where you set up a situation where you can, with your eyes closed, click back and forth on a single button, and that button will A/B a process that you’re considering.

Perhaps the button is simply the bypass on a plugin. Or maybe it’s literally the A/B button in a plugin, where you can store two different settings and switch back and forth between them.

Here’s how I do it:

• Make sure the A and B processes are gain-matched as much as possible, like literally down to the tenth of a dB if necessary. This is critical — if one is slightly louder than the other, our brain will pick the louder one every time, because our stupid brains think that louder things are better. This is how the loudness wars happened.

• Hit play on the relevant section (looping it if necessary).

• Position your cursor over the relevant button.

• Close your eyes!

• While the song is still playing, rapidly click the button a few times with your eyes closed, until you’re no longer sure which state the button is in.

• With your eyes still closed, continue to listen in a more leisurely way, clicking the button back and forth between the two states.

• One of the two will appeal to you more; when you’ve arrived at which one it is, open your eyes and note it.

Repeat a couple of times to be sure. And then go with the one your ears are choosing! The point of the exercise is to remove your various biases — only your ears can tell you the true answer.

Ears don’t lie — jamie