Skip to content

Reference tracks are for calibration, not imitation

A reference track resets ears that have adapted to your room and your mix. Match its loudness first, then compare the balance, not the tone.

Updated · 6 min read Lesson 3 of 44 in Mixing & Mastering

In short

  1. After hours on one mix your ears adapt to its balance, and a reference resets that baseline.
  2. Match the reference's loudness to your mix before comparing, or the louder master wins every time.
  3. Use the reference to check balance and translation, then fix problems on the tracks instead of copying its tone.
In this lesson (8)

The ear fatigue loop

You have worked on a mix for four hours straight. You have tweaked the kick, adjusted the vocal compression and brightened the acoustic guitars, and in the studio it sounds great. The next morning you play it in the car and it falls apart: the bass is muddy, the vocal is buried and the upper mids are harsh.

Your ears adapted. Hearing adjusts to whatever it listens to for a long time, so after hours on one mix in one room, that mix's balance starts to sound normal, faults included. A you know well is how you get back to a fixed point.

Why it matters

You need a fixed point

A colourist checks their eyes against a grey card. A reference track does the same job for your ears. Play a professionally mixed song that you know translates well on many systems, and you remind yourself what a balanced low end, a clear vocal and a smooth top end sound like in your room, on your speakers, today.

The comparison only works if it is fair. A released master is usually limited and much louder than a mix in progress. Play it at its own level and it will win every comparison, because louder sounds better.

Figure 1. An example of level matching. Your unmastered chorus reads -18 LUFS and sets the level. The released reference reads -8 LUFS, so it goes down 10 dB to meet your mix before you compare anything.

Hear how strong that pull is, even with only 1 dB between the two sides:

Listen

Blind test: which one sounds better?

A synth loop or a real mix, played twice with one side 1 dB louder. Pick the one you prefer, then find out which was louder.

Loading demo…
100%

Start with your speakers or headphones low.

The science

Matching loudness, not peaks

Loudness meters follow ITU-R BS.1770, which weights the signal toward how the ear hears and reports the result in , loudness units relative to full scale. A difference of one loudness unit is one decibel, so the gain change you need is simply the difference between the two readings:

ΔG=Lmix−Lreference\Delta G = L_{\text{mix}} - L_{\text{reference}}

If your chorus reads -18 LUFS and the reference chorus reads -8 LUFS, ΔG=−10\Delta G = -10 dB: turn the reference down by 10 dB. Compare like with like. An integrated reading over a whole song includes quiet intros and breaks, so match the of the same kind of section, chorus against chorus.

Matching peaks does not work. A limited master and an open mix can have the same peak level and very different loudness, because the master's average level sits much closer to its peaks.

Matching also needs repeating. Each check resets your baseline, and the drift starts again as you keep working.

Figure 2. A sketch, not a measurement. Without a reference, your sense of a neutral balance drifts the longer you work. A matched reference check pulls it back each time.

Try it

Set up a reference channel

Do this once per project, with two or three references in the same style.

0 of 7 steps done

You will hear balance differences that had become invisible, usually the vocal level and the amount of low end, and you will hear them as specific changes to make.

Common mistake

Copying instead of comparing

To match the reference's frequency curve, you put an on your master bus. A released master is the result of the arrangement, the recording, the mix and the mastering together. A big high shelf on a dull mix will not turn it into that record; it will make the dull parts harsh and the bright parts piercing. Find the parts that cause the difference and fix them where they live.

The second mistake is copying the reference's tone. Your song has its own arrangement and its own character. If the reference has a bright, thin acoustic guitar and your song needs a warm, full one, keep yours. Compare how loud the guitar sits, not its exact sound.

Takeaway

The reference is a mirror

A reference track shows you problems; it does not tell you what your song should be. Use it to check whether the low end translates, whether the vocal is clear and whether the still snap.

Finish by playing both your mix and the reference on a phone speaker at low volume. If the balance between the vocal and the snare holds up in the same way on both, you can print the mix with confidence.

3 questions

Check yourself

  1. Your chorus reads -18 LUFS and the reference chorus -8 LUFS. What do you do before comparing?
  2. Why can a mix that sounded right at midnight sound wrong in the car the next morning?
  3. The matched reference has a brighter top end than your mix. What is the better fix?

Sources

Show all 2
  • ITU-R BS.1770-5 (2023). Algorithms to measure audio programme loudness and true-peak audio level. International Telecommunication Union.
  • Senior, M. (2011). Mixing Secrets for the Small Studio. Focal Press.

Next lesson · 4 of 44

Mud is often masking, not dirtMud usually comes from several parts piling up in the low mids and hiding each other. Find the pile-up and thin the parts that do not need it.

Previous: Monitoring level changes the balance you hear

Portrait of Virzy Guns

Written by Virzy Guns

Songwriter and producer, founder of Virzy Guns Production. Now building HealingWave.