Vanta – Engine Notes

Why loudness targets don't equal perceived loudness

Ref: PHY-01

Hitting -14 LUFS is a mathematical target, not a musical one. Here is why the meter often lies about how "big" your track feels.

Technical Answer

Integrated Loudness (LUFS) measures average energy over time. However, human ears are more sensitive to mid-range frequencies and transient peaks. A track can be mathematically "loud" but feel "small" if its crest factor is low and its transient information is smeared by over-limiting.

The Vanta Stance

We prioritise the "Snap." Vanta looks at the Peak-to-RMS ratio (Crest Factor) before it looks at the LUFS target. If reaching your requested loudness requires destroying the rhythmic energy of the mix, our engine will automatically stop short. We believe a dynamic track turned up by the DJ always sounds better than a crushed track fighting the limiter.

Expand Note

The problem with automated mastering chasing LUFS

Ref: INT-04

Most "AI" mastering services are programmed to impress you with volume. But volume at the cost of phase and dynamics is a bad trade.

Technical Answer

Algorithms that use linear gain to hit a brick-wall target often ignore the cumulative distortion being added. When an engine "smashes" into a limiter to hit -7 LUFS, it creates inter-sample peaks (ISPs) that cause consumer DACs to distort during playback.

The Vanta Stance

Transparency over volume. Vanta is built to be "Local First" and "Deterministic." We don't use generative AI to guess what your song should sound like; we use a 32-stage expert system to ensure your mix translates perfectly to every system. We never "smash"—we prepare the signal through dual-stage clipping so the limiter barely has to work.

Expand Note

Why mono compatibility still matters in streaming

Ref: STR-02

Your fans listen on single Bluetooth speakers, phones, and in clubs. If your stereo width kills your mono correlation, your mix disappears.

Technical Answer

Phase cancellation occurs when L and R channels have out-of-phase information. In a mono summation (like a phone speaker), these frequencies literally cancel each other out, leading to "ghost" vocals or disappearing bass.

The Vanta Stance

Controlled Width. Vanta uses Mid/Side (M/S) intelligence to monitor stereo correlation in real-time. If it detects that widening the side channel will drop the correlation below the "Danger Zone" (0.5), it pre-emptively clamps the enhancement. We refuse to let your mix break for the sake of artificial width.

Expand Note

What happens to your master after AAC/MP3 encoding

Ref: QC-09

Lossy encoding isn't just "lower quality"—it actually changes the peak levels of your audio. If you don't account for it, you get digital clipping.

Technical Answer

The psychoacoustic models used in AAC and MP3 encoding reconstruct the waveform in a way that can cause new peaks to exceed 0dBFS, even if the original WAV was limited to -0.1dB. This is why "True Peak" metering is critical.

The Vanta Stance

The -1.0dBTP Safety Margin is just the beginning. While industry-standard automated services skip this entire step to save on processing costs, Vanta is the only engine that performs a live **"Encode Survival"** check. We run a hidden background transcode of your actual master against weighted stress-models of the target AAC and MP3 encoders before you save. If we detect that the encoder's psychoacoustic reconstruction will push your signal into clipping, our **Correction Loop** automatically calculates the precise trim required to clear the safety gate—typically resolving even worst-case overshoots in a single pass. We've validated this across thousands of laboratory encodes to guarantee that your master arrives at the listener exactly as you heard it: clean, punchy, and distortion-free.

Expand Note

Why crest factor matters more than loudness

Ref: PHY-05

The difference between "loudness" and "impact" is the space between the peaks. Learn why the air in your mix is your best friend.

Technical Answer

Crest Factor is the difference between the peak and the RMS (average) level. A high crest factor means your drums have room to "hit." A low crest factor means everything is pushed into a flat wall of sound, which is fatiguing to the listener.

The Vanta Stance

Transients are sacred. Vanta's "Transient Integrity Guard" (TIG) uses a look-ahead architecture to identify incoming rhythmic pulses. It creates a micro-headroom pocket for the transient before the limiter hits, ensuring that your kick and snare retain their "snap" even in a commercial-level master.

Expand Note

Why most spectrum matchers damage mixes

Ref: INT-09

Reference-matching seems like a shortcut, but forcing your track into another track's "shape" often ignores the unique physics of your recording.

Technical Answer

Spectrum matching tools often apply thousands of tiny EQ bands to force a frequency curve. This creates massive phase rotation and group delay issues, making the mix feel "blurry" or "phased," especially in the low end.

The Vanta Stance

Deterministic Matching. Vanta doesn't force a blind EQ curve. We use 9 primary spectral dimensions to understand the *character* of your mix, then apply broad, musically-defensible corrections. We'd rather be 90% "close to typical" and 100% phase-accurate than perfectly "matched" and sonically broken.

Expand Note