Why use a reference
The ear adapts to the room, the speakers and your own mix within minutes. A commercial reference track in the same genre recalibrates your listening: it shows how loud the vocal should be relative to the backing track, how much low end there is, how wide the sides are, how bright the top end is. The comparison, however, is only valid at equal volume: a track that is 1 dB louder is perceived as “better” (more bass, more treble, more presence).
Choosing references
- 2–3 tracks in the same genre and with a similar arrangement (same instrumentation, same tempo, same density).
- One reference for technical aspects (low end, brightness) and one for aesthetics (vocal, effects).
- Recent references for loudness, classic references for balance.
Loudness matching
- Import the reference into a mixer track that goes directly to the output, without passing through the master plugins (in FL Studio: a dedicated track with routing straight to the interface output, or a comparison plugin on the master that switches before the chain).
- Measure the integrated loudness (or short-term in the chorus) of the reference and of your mix.
- Turn the reference down (almost always louder than the unmastered mix) until the loudness matches. Example: reference −8 LUFS, mix −16 LUFS: lower the reference by 8 dB.
- Compare equivalent sections (chorus with chorus) for 10–20 seconds, then go back to the mix.
Useful plugins: ADPTR Metric AB, Mastering The Mix Reference, iZotope Ozone (Reference and Tonal Balance Control), Plugin Alliance bx_refcomp; in FL Studio, a reference track and a Fruity Balance for the level.
What to compare
| Aspect | How | Question |
|---|---|---|
| Vocal/backing track | Listening, short-term LUFS of vocal and instrumental | Does the vocal sit up front like in the reference? |
| Low end | Sub and Bass bands, listening at low volume | More or less weight? Longer or shorter notes? |
| Low mids | 250–500 Hz band | Is the mix muddier? |
| Brightness | Presence and Air bands | Darker or harsher? |
| Width | Correlation, Side/Mid | Narrower or wider? |
| Dynamics | Crest factor, LRA | Does the chorus open up like in the reference? |
| Depth | Listening to the reverbs | Drier or wetter? |
Tonal balance curves
Tools such as Tonal Balance Control show the typical spectral range of a genre; the app's analysis shows the relative energy per band. Compare your mix with the reference band by band: differences of 1–2 dB are normal, while anything over 3 dB indicates a choice to reconsider by ear.
Common mistakes
- Comparing a −16 LUFS mix with a −8 LUFS master without matching levels: the reference always wins.
- Copying the spectral curve without listening: two tracks with the same curve can sound different (arrangement, key, transients).
- References of a different genre or tempo: an 80 BPM track with sparse instrumentation is not a reference for a 128 BPM track.
- Using a low-quality MP3 reference to judge the highs.
- Running the reference through the master plugins: it gets compressed and limited a second time.
Starting settings
| Parameter | Value | Why |
|---|---|---|
| References | 2–3 tracks in the same genre | Reliable comparison. |
| Format | WAV or lossless, normalization off | Real loudness and highs. |
| Reference level | Lowered until loudness is matched | Fair comparison. |
| Comparison length | 10–20 s per equivalent section | The ear does not adapt. |
| Routing | Reference outside the master chain | No double processing. |
| Spectral tolerance | ±1–2 dB per band | Beyond 3 dB, check by ear. |
Frequencies to watch
| Range | What to do |
|---|---|
| 20–120 Hz | Weight of the low end: compare the Sub and Bass bands at equal volume. |
| 250–500 Hz | Mud: the unmastered mix often has 1–3 dB more than the reference. |
| 2–5 kHz | Vocal presence and harshness. |
| 8–16 kHz | Air: commercial masters are often brighter than raw mixes. |
Example chain
| Step | Tool | Settings | Why |
|---|---|---|---|
| Reference | Dedicated mixer track / Metric AB / Reference | Level matched | Fair comparison. |
| Loudness measurement | Youlean Loudness Meter / Ozone / Wave Candy | Integrated and short-term | Level matching. |
| Spectral analysis | Tonal Balance Control / app analyzer | Bands for comparison | Balance differences. |
| Corrections | EQ and balance in the mix | Adjustments on the tracks | The mix is fixed in the mix. |
How to check the result
- Integrated loudness difference between mix and reference matched within 0.5 LU before every comparison.
- Relative energy per band of the mix within ±2 dB of the reference in the Bass, Low-mid, High-mid and Presence bands.
- Mix correlation within ±0.15 of the reference.
- Crest factor of the unmastered mix 2–6 dB above that of the mastered reference (normal: mastering reduces it).