A/B Testing Your Mix: Getting Objective Feedback on Two Versions

The Problem With "Which One Do You Prefer?"

When you've made two versions of a mix — maybe one with more reverb on the vocals, or a different level of compression on the drums — asking someone "which do you prefer?" is rarely useful. You'll get an answer, but you won't understand why. And the preference might be arbitrary, based on whichever version they heard first or simply which was louder.

Effective A/B testing asks more specific questions and uses a systematic approach to evaluate meaningful differences.

What to Compare

Not all A/B comparisons are equally useful. The most productive comparisons isolate a single variable:

• Version A has more low end. Version B has less. Which sits better in a car stereo?
• Version A has heavier vocal compression. Version B has lighter. Which sounds more natural?
• Version A uses reverb on the snare. Version B is dry. Which has better energy?

When you know exactly what changed between versions, feedback becomes actionable. Instead of "I like version B better," you get "the less-compressed vocals in version B feel more emotionally present."

The AI A/B Test Approach

SongTools.ai's A/B Comparison tool lets you upload two audio files and specify the question you want answered. The AI analyses both versions against your specific criteria and returns a comparative assessment covering the dimensions you care about.

This is particularly useful when you're too close to your own work to hear the difference objectively. After hours in the studio, your ears become fatigued and your judgment becomes unreliable. An AI comparison gives you a fresh, criteria-driven perspective.

Building a Testing Protocol

The most systematic approach to A/B testing mixes involves:

• Loudness matching — Before comparing, ensure both versions are at the same perceived loudness. Even a 1dB difference will make the louder version seem "better."
• Blind evaluation — Have someone else play the versions to you in random order so you don't know which is which.
• Specific criteria — Define what you're evaluating before you listen: low end impact, vocal clarity, stereo width, overall energy.
• Multiple playback systems — Test on speakers, headphones, and in a car. A mix that sounds great on studio monitors might fall apart on earbuds.

When Two Versions Both Sound Good

Sometimes both mixes are genuinely good but different. In this case, the decision isn't about which is objectively better — it's about which version best serves the song's purpose.

A vocal-forward, intimate version might be right for a singer-songwriter ballad. A denser, more aggressive version might be right for the same song if it's being placed in a film trailer. Context matters.

Use A/B testing to clarify what each version does well, then choose based on where the music is going.