How to use reference tracks for mixing, step by step

Pick a released song in your style, turn it down to the same loudness as your mix, then compare one thing at a time: frequency balance, dynamics, stereo image. Matched loudness plus one question per listen is what turns "theirs sounds better" into a fix you can make.

Most producers already "use references". They play their mix, play a song they love, feel their heart sink, and go back to twisting knobs. That's not referencing. That's comparing. Here's the difference, in 7 steps.

Step 1: Choose the right reference

A good reference does the same job as your track. Look for:

• The same style and tempo range. A sparse acoustic song can't tell you how loud your 808 should be.
• A similar arrangement. If your chorus has drums, bass, a vocal and two synths, find a chorus with roughly the same parts.
• A sound you'd be proud of. You're choosing a target, so pick the one you'd love your track to sit next to.

Pick two or three, not ten. More references means more opinions, and you'll start chasing all of them.

Use the best quality file you can get: a lossless download or a purchased WAV beats a ripped video. Low-quality files lose top end and stereo detail, so you'd be matching a damaged picture.

Step 2: Put it in your project the right way

Drop the reference on its own track in your session, and make sure it skips your master bus processing. If your mix bus has a compressor, a limiter or an EQ, the reference must not go through them, or you're comparing your mix with a squashed copy of the song.

Most DAWs let you route a track straight to the output, or place your metering plugins after a point the reference bypasses. Do it once, save it in your template, and never think about it again.

Step 3: Match the loudness. This is the step everyone skips

Released songs are mastered. Your mix isn't. So the reference is almost always louder, and louder always sounds better to the human ear: fuller, brighter, punchier.

If you compare at different levels, you'll "fix" problems that don't exist. You'll boost the low end and the highs, and end up with a harsh, boomy mix that still sounds worse.

So turn the reference down until both sound equally loud. A loudness meter that reads LUFS (a measure of how loud a song feels, not just its peaks) makes this precise: match the short-term readings of the loudest section in both. If you don't have one, use your ears and flip back and forth until neither one jumps out.

Want the numbers behind loudness? Read what LUFS to master to for Spotify.

Step 4: Compare the same moment

Chorus against chorus. Verse against verse. Drop against drop.

A quiet intro against your loudest section tells you nothing. Loop four to eight bars of each, at the same point in the song, and switch between them.

Step 5: Ask one question per listen

Here's the core of it. An untrained ear hears a mix as one feeling: "it sounds bigger". A pro hears three separate things, and checks them one at a time:

1. Frequency balance. Is my low end heavier or lighter? My low mids thicker? My highs brighter or duller? Tip: put a low-pass filter on your monitoring and listen to only the lows of both, then a high-pass for only the highs.
2. Dynamics. Does their kick hit harder against the rest? Does their chorus feel denser? Do their quiet parts stay quiet?
3. Stereo image. Press mono. Does their mix hold together better? Are their wide sounds wider, and is their center more solid?

Write one line for each answer, like "my low mids are thicker, my highs are duller, my kick is quieter against the bass". Now "theirs sounds better" has become three things you can fix.

Step 6: Switch fast, and rest your ears

Your ears adjust to whatever they hear for more than a few seconds. Listen to each version for about 5 to 10 seconds, then switch. Longer, and they start to sound the same.

Take a break every 30 to 45 minutes. Tired ears hear dull mixes as fine and push the highs too far.

Step 7: Make one change, then check again

Fix the biggest gap first, usually frequency balance. Then go back to the reference at matched loudness. Did the gap shrink? Did something else get worse?

One change, one check. Ten changes and one check is how mixes go in circles.

If your biggest gap is a thick, cloudy middle, our guide on why a mix sounds muddy walks through it.

Common mistakes with reference tracks

• Copying the reference instead of learning from it. You want your song to sit next to theirs, not to clone it. Different songs need different choices.
• Matching a mastered reference with your unmastered mix in level. Match the balance between parts, not the final loudness. Loudness comes at mastering.
• Switching references halfway through. Pick them at the start and keep them for the whole mix.
• Referencing only at the end. Check early, while it's cheap to change things.

Free for 7 days

Get an expert's ears on your track

Try our community, Music Production Mastery, free for 7 days. You get a 30-minute call with one of our experts, every premium plugin and app, and our weekly live masterclass.

Start my free week

Join free now, and stay if you love it.

Why this works

Comparing without a method leaves you guessing. Referencing with matched loudness and one question at a time gives you a fixed point to measure against, every time you sit down.

And it compounds. Every track you reference this way teaches your ears something new about frequency balance, dynamics and stereo image, so the next mix starts closer to the sound you hear in your head.