Referencing and A/B at Matched Loudness

Human hearing is unreliable in one very specific way: when two versions differ in level, we almost always prefer the louder one. A louder signal sounds fuller in the bass, brighter in the treble and more detailed overall — that's how our ears' sensitivity changes with volume. In mastering, where every move changes level, this bias will mislead you constantly unless you remove it.

Matched-loudness comparison

The rule is simple: only compare two things at the same perceived loudness. Concretely:

• When you bypass a processor, its output level should match its input. Most compressors, saturators and EQs have an output gain control for this; some have an automatic gain-compensation option.
• When you compare your master against the unmastered mix, turn the master down (not the mix up) until both measure the same short-term or integrated LUFS.
• When you compare against a commercial reference, turn the louder one down to match.

Use a loudness meter to match, and then fine-tune by ear over a few seconds of the same section. Within about 0.5 LU is close enough for most decisions.

If the master still sounds better than the mix when they're level-matched, your processing is genuinely improving it. If the mix sounds as good or better, you're only making it louder — and the streaming service would have done that for free.

Choosing references

A good reference is a commercially released song:

• In the same genre and era, with a similar arrangement density. Referencing a sparse acoustic song against a full EDM master teaches you nothing.
• That you know well on your monitoring. You want something whose sound you trust in your room or headphones.
• From a lossless source when possible. A low-bitrate stream adds its own artifacts to the high end.

Two or three references are better than one; each is a different but valid answer.

What to listen for

Compare one thing at a time, on the same kind of section (chorus against chorus):

1. Low end. Is your kick and bass heavier, lighter, or similar? Is the sub as controlled?
2. Mid-range. Is the vocal or lead as forward? Does your mix sound boxier or more hollow?
3. High end. Is it as bright? Harsher? Duller?
4. Width. Does your mix feel narrower or wider? Does the reference keep the low end centred?
5. Dynamics. At matched loudness, do your drums hit as hard? Does the chorus lift as much over the verse?

Only after that compare raw loudness — and remember that the reference's loudness tells you about its genre's conventions, not a target you must match on a streaming service.

A/B in the Studio

SoundGen lets you attach a mastering reference to a project and flip between A · Mix and B · Reference in the Master view while the song plays. Its Match loudness option turns the louder side down so you hear tone, not volume. Use it the same way you would any A/B: short switches, same section, one question at a time.

Avoiding reference fatigue

• Switch quickly. Long listening at one level re-calibrates your ears; quick switches catch differences.
• Don't copy, interpret. A reference shows you a direction. You're not trying to make a clone; your song has its own arrangement.
• Take breaks. After about 20 minutes of focused comparison, your judgement drifts. Step away, come back, and check whether your last few moves still sound right.

Final check

Before you commit, listen to the complete master once from start to finish without stopping or adjusting anything — at a moderate level, ideally on a system you haven't used during the session. Listen for the whole-song experience: does the intro sound as good as the chorus? Is the ending clean? Is anything distracting? Fix only what still bothers you.