Loudness Normalization: Why It Matters for Mix Comparisons
Learn how Mix By Ear's loudness normalization removes volume bias from your A/B comparisons, helping you make better mixing decisions based on actual quality—not just who turned it up louder.
You upload two mixes to compare. One sounds obviously better—punchier, fuller, more exciting. You pick it. But here’s the problem: that mix was just 0.5 dB louder.
This isn’t a hypothetical. It’s a documented phenomenon that has fooled professional engineers for decades. And it’s exactly why Mix By Ear includes loudness normalization.
The Problem: Your Ears Lie About Loudness
Human hearing has a built-in bias: louder sounds better. Even tiny volume differences—as small as 0.1 dB—can sway your preference toward the louder option. This happens unconsciously and affects everyone, regardless of experience level.
Here’s what louder does to your perception:
- Bass feels deeper (Fletcher-Munson curves make low frequencies more apparent at higher volumes)
- Highs feel crisper (same effect on the treble end)
- Everything feels more exciting (louder triggers a stronger emotional response)
The result? A mediocre mix that’s been turned up will beat a better mix that’s quieter. Every time.
This is why A/B comparisons without level matching are essentially useless. You’re not comparing mix quality—you’re comparing who pushed the fader higher.
The Solution: LUFS-Based Normalization
Mix By Ear uses LUFS (Loudness Units relative to Full Scale) to measure and match the perceived loudness of your tracks. Unlike simple peak metering, LUFS models how humans actually hear sound.
Why LUFS, Not Peak or RMS?
| Measurement | What It Measures | The Problem |
|---|---|---|
| Peak | Loudest single sample | A snare hit can peak high while the mix sounds quiet |
| RMS | Average signal level | Doesn’t account for how we perceive frequency content |
| LUFS | Perceived loudness | Models human hearing with frequency weighting and gating |
LUFS applies a frequency curve that matches human hearing sensitivity and ignores silent passages that don’t contribute to perceived loudness. The result is a number that accurately represents how loud something feels.
How Mix By Ear Uses Loudness Normalization
When you enable normalization in your project settings, Mix By Ear analyzes each track’s loudness and adjusts playback gain so all tracks sound equally loud. No audio is permanently altered—it’s just a playback adjustment.
You get two modes, each designed for specific comparison scenarios:
Streaming Mode (Integrated LUFS)
What it does: Analyzes each track’s overall loudness and matches them to a target level (typically -14 LUFS, matching Spotify and YouTube).
Use this when:
- Comparing full mixes or masters from start to finish
- Testing how your master will sound on streaming platforms
- Getting feedback on final masters before release
- Comparing your mix against commercial references
How it works: If Track A measures -10 LUFS and Track B measures -14 LUFS, Track A gets turned down 4 dB during playback. Both now play at -14 LUFS, removing loudness as a variable.
Example scenario: You have three master candidates—one from your own DAW, one from a mastering engineer, and one from an automated service. Enable Streaming Mode to hear them at matched loudness and focus on tonal balance, dynamics, and overall vibe instead of who mastered it loudest.
Comparison Mode (Short-Term LUFS)
What it does: Continuously adjusts gain based on short-term loudness (1-3 second windows), keeping tracks matched moment-to-moment.
Use this when:
- Comparing specific sections (just the chorus, just the drop)
- Tracks have very different dynamics (one compressed, one dynamic)
- A/B testing different processing on the same section
- Comparing how two mixes handle a specific transition
How it works: Instead of one fixed gain adjustment, the system rides the level in real-time. If Mix A has a quiet verse and loud chorus while Mix B is uniformly loud, Comparison Mode keeps them matched in both sections.
Example scenario: You’re comparing two versions of a mix—one with heavy bus compression, one more dynamic. With Integrated normalization, the quiet verses of the dynamic mix would sound much quieter than the compressed version’s verses, making section-by-section comparison unfair. Comparison Mode keeps each section matched, so you can accurately judge how each approach handles the verse and the chorus.
Quick Reference: Which Mode to Use
| Scenario | Recommended Mode |
|---|---|
| Comparing final masters | Streaming Mode |
| Testing streaming platform loudness | Streaming Mode |
| Comparing specific song sections | Comparison Mode |
| One mix is compressed, one is dynamic | Comparison Mode |
| Getting feedback from non-engineers | Streaming Mode |
| Detailed A/B of processing choices | Comparison Mode |
When to Turn Normalization Off
Normalization removes loudness as a variable. But sometimes loudness is exactly what you’re evaluating.
Turn off normalization when:
- Comparing compression/limiting approaches: You want to hear the actual impact of different loudness strategies
- Evaluating master loudness for a specific platform: You need to hear what it will actually sound like at its intended level
- Demonstrating the loudness difference to a client: Sometimes you need to show why one approach “hits harder”
In Mix By Ear, normalization is a toggle in your project settings. Enable it by default for honest comparisons, disable it when loudness itself is what you’re testing.
Streaming Platform Loudness Targets
When using Streaming Mode, it helps to know what loudness levels the major platforms target:
| Platform | Target | Notes |
|---|---|---|
| Spotify | -14 LUFS | Default setting; users can choose -11 or -19 |
| Apple Music | -16 LUFS | Slightly more conservative than Spotify |
| YouTube | -14 LUFS | Always on, cannot be disabled by viewers |
| Amazon Music | -14 LUFS | Similar behavior to Spotify |
| Tidal | -14 LUFS | Default on, user can disable |
| SoundCloud | None | Plays at uploaded loudness—no normalization |
The industry has largely converged on -14 LUFS as the standard. This is why Mix By Ear defaults to -14 LUFS in Streaming Mode—it matches what most of your listeners will actually hear.
What This Means for Your Masters
If you master to -8 LUFS (very loud), streaming platforms will turn it down 6 dB. If you master to -14 LUFS (moderate), it plays as-is. The ultra-loud master gains no advantage and loses dynamic range for nothing.
This is why the “loudness war” is essentially over. Platforms normalize everything, so the hyper-compressed master just sounds flat compared to a more dynamic one played at the same level.
Best Practices for Mix Comparisons
Start with Normalization On
Make it your default. The vast majority of comparisons benefit from level matching. You can always turn it off for specific tests.
Use Consistent Modes
Don’t compare Track A with Streaming Mode and Track B with Comparison Mode. Keep it consistent so your comparisons are valid.
Listen at Moderate Volumes
Even with normalization, your monitoring level affects perception. Around 80 dB SPL is the sweet spot for critical listening—loud enough to hear detail, quiet enough to avoid fatigue and Fletcher-Munson distortion.
Trust Your First Impression Less
With normalization enabled, your initial gut reaction is more reliable—but still take time. Quick A/B switches (under 1 second) help you catch differences before your brain adapts.
Check Peak Levels After Normalization
If a very quiet track gets boosted significantly, watch for clipping. Mix By Ear respects true peak limits, but it’s worth understanding: a track mastered at -20 LUFS boosted to -14 LUFS means +6 dB of gain. If peaks were already near 0 dBFS, that’s a problem.
The Bottom Line
Loudness normalization isn’t about making everything sound the same—it’s about removing volume as a confounding variable so you can actually hear the differences that matter: tonal balance, stereo width, dynamic feel, and overall vibe.
Without it, you’re not comparing mixes. You’re comparing fader positions.
Mix By Ear’s normalization modes give you the tools to make fair comparisons, whether you’re choosing between master candidates, testing processing decisions, or getting feedback from clients who don’t know a compressor from a composter.
Enable normalization. Trust your ears. Make better mixing decisions.
Ready to compare your mixes fairly? Create a project and try loudness normalization in your next session.