How to A/B pedals without lying to yourself.
Almost every comparison video on YouTube is wrong, and almost every comparison you do in a guitar store is worse.
Two sides. One page.
Jason argues the piece. Johnny argues against it. Read both, then call the winner below.
The take is shown below.
Most pedal A/B tests are not actually tests. They're rituals. You plug in pedal A, play a riff, plug in pedal B, play the same riff, and announce that you can hear the difference. You probably can. The question is whether the difference you're hearing is the difference between the pedals or the difference between everything else that changed when you swapped them.
I've done a lot of pedal A/B tests over the years and I've been wrong about my conclusions enough times to develop some methodology. None of it is original. Most of it is borrowed from how engineers do double-blind audio tests. But almost no guitarist actually does any of it, including, until embarrassingly recently, me.
If you're going to compare two pedals, here's how to do it so the comparison means something.
First: level-match. This is the one nobody does and it invalidates almost every comparison on YouTube. Pedals have different output levels. Two overdrives that sound identical at the same gain setting can sound completely different if one is 2dB louder, because human ears interpret louder as 'better' and 'fuller' and 'more present.' You will pick the louder pedal every single time, even if it's objectively worse. The fix: use a loop switcher or an A/B box that lets you toggle between the two pedals with the volume of the loud one trimmed down. If you don't have one, set both pedals up, play through one, adjust its output until it matches the other when measured at the amp speaker with an SPL meter on your phone. The free app is fine. Within 0.5dB is the goal.
Second: blind. Have someone else switch the pedals while you're not looking. This is annoying. Your bandmate hates doing it. Do it anyway. The number of times I have confidently picked pedal A in a sighted test and then picked pedal B in a blind test of the same two pedals is humiliating. Sighted preference is mostly preference for the pedal you expected to like.
Third: same context. Play the same musical phrase, in the same key, with the same picking dynamics, through the same amp at the same volume, in the same room. If you change two of those variables at once you have no idea what you just learned. The guitar-store A/B test, where you sit on a stool with a different guitar than you'd actually use and an amp you don't own at a volume you'd never play at, is generating data about pedals in a context that has nothing to do with how you use them.
Fourth: longer than you think. A/B tests where you switch every five seconds favor pedals with immediate impact — louder, brighter, more compressed. A/B tests where you play a full song through one pedal, then a full song through the other, favor pedals that feel good to play through over time. Both are valid. Most differences that seem important in a five-second test disappear in a five-minute test, which tells you something about how much of pedal marketing is built around the first impression.
Fifth: record both. Phone audio is fine. Listen back the next day with fresh ears and no idea which is which. You will learn things about your preferences that will surprise you.
I did this exercise last summer with three Klon-style overdrives and discovered that the cheapest one — a kit-built clone I'd put together for about $80 — was indistinguishable from the boutique pedal I'd been gigging for two years. Not 'close to.' Indistinguishable in a blind test, level-matched, recorded. I sold the boutique one.
That's the cost of doing this honestly. Sometimes the answer is that you spent money you didn't need to spend. That's also the value.
— Jason
The double-blind method finds the pedal that survives scrutiny. Records get made on the pedal that starts a fire. Those are different pedals.
The methodology is correct and the epistemology is upside down. Level-match, blind the switching, control the variables: do all of it and you will reliably discover which of two overdrives is objectively distinguishable, which is a fact about the pedals. But the purchase decision was never about the pedals. It is about which box makes you play longer, reach further, book the extra rehearsal. That variable lives inside the guitarist, is invisible to the SPL meter, and is the only one that ends up on tape.
The louder-is-better bias the essay wants to correct is not a bug in the listener. It is the actual experience the audience will have. If pedal A at its natural output sounds bigger than pedal B at its natural output, then A is bigger in every room that matters, and trimming it down to lab-match B is constructing a scenario that will never occur on a stage to make the pedals tie. Fair tests between unequal contestants are a laboratory courtesy. Gigs are unfair on purpose.
There is also the small matter of who will actually do any of this. The loop switcher, the phone SPL meter, the half-decibel tolerance, the friend swapping cables behind your back. The essay admits almost nobody complies, including its author until embarrassingly recently, and the reason is not laziness. It is that the ritual it mocks, plug in, play, feel something, announce a winner, is the correct protocol for measuring the thing being bought, which is enthusiasm. Placebo you can afford is tone you play better through, and the essay said that sentence itself without hearing it.
Keep the method for magazine shootouts and forum arguments, where facts are the product. At the shop, trust the fire. The fire is the data.
Discussion
Loading comments…