LeugenAcademie provides professional lie detection tests based on Layered Voice Analysis (LVA™) — the world's most advanced voice analysis method, developed by Nemesysco after 25+ years of scientific research.
Layered Voice Analysis (LVA™) is a breakthrough technology developed by Israeli company Nemesysco, founded in 1997 by Amir Liberman. Unlike the traditional polygraph test — which measures skin resistance, blood pressure and breathing — LVA analyses the voice alone.
The human speech process is one of the most complex actions the body performs: lungs, vocal cords, tongue, teeth and lips work together under orchestration by the brain. When someone lies or experiences emotional tension, that orchestration changes in a subliminal, uncontrollable way.
LVA detects more than 150 voice parameters per speech segment — including frequency patterns entirely new to phonetics — and classifies them into 14 emotional dimensions. The system is language-independent: it analyses how someone speaks, not what they say.
The Emotional Diamond™ visualises 8 of the 14 emotional dimensions that LVA detects in real time. The position and size of the profile shifts with each answer.
When speaking, the brain simultaneously orchestrates dozens of muscle groups. Cognitive processes such as concealing information increase mental load measurably — and this leaks unconsciously into the voice.
LVA7 CORE extracts over 150 parameters per speech segment via 3 base formulas: frequency modulations, energy levels, temporal patterns and statistically anomalous deviations from the individual baseline.
Every individual has a unique emotional baseline. LVA does not detect "lies" as such, but sharp deviations from that baseline — regardless of nationality, language or gender.
Analysis takes place in real-time during the interview and retrospectively on recordings. Both modes yield independent datasets combined by the operator in the final report.
LVA analyses pure vocal patterns, not semantic content. The method works identically in Dutch, English, Arabic, Chinese or any other language — validated in 40+ countries.
The core algorithms are protected by two international patents held by Nemesysco Ltd. Remaining internal processes are proprietary trade secrets.
A professional test is a carefully structured forensic investigation. Each step is designed to maximise the quality of the analysis and give the test subject a fair process.
Fraud investigations, workplace theft, integrity issues, screening for sensitive positions and internal investigations following incidents.
Supplementary evidence in civil proceedings, divorce cases, maintenance disputes and criminal defence.
Relationship problems, suspected infidelity, family disputes over inheritance or property where mutual trust needs to be restored.
Security screenings, integrity tests for civil servants, supplementary tool in interviews and criminal investigations.
In fact, no machine can detect lies — a machine does not understand what a lie is. To determine what needs to be measured, it is important to first understand exactly what a lie is. At LeugenAcademie we distinguish 10 different types of lies. Knowing the type of lie you are looking for is decisive in formulating the right test questions.
A deliberate, complete untruth where the speaker knows the statement is factually incorrect. The most recognisable category, but not the most common.
Part of the truth is told, while essential information is deliberately omitted. Technically not untrue, but deliberately misleading.
Answers are phrased so generally or ambiguously that they allow multiple interpretations. LVA detects the deliberate imprecision in the wording.
The speaker answers a different question than asked, or steers the conversation in another direction. A classic tactic in interrogations and media appearances.
Relevant information is not voluntarily shared, without being directly asked. LVA explicitly measures whether someone is consciously withholding information.
The speaker has convinced themselves that the untruth is justified ("it's for their own good"). The emotional charge is therefore different.
"I don't remember" as a deliberate strategy. LVA distinguishes authentic memory loss from a deliberate avoidance strategy.
Facts are deliberately presented somewhat differently or more extremely than they are — exaggeration or understatement with misleading intent.
Lies primarily intended to spare the other person or salvage a social situation. Lower stress response due to reduced emotional loading.
Unpremeditated, impulsive lying without forethought. Recognisable by specific timing and energy patterns in the voice that LVA detects.
The question "does LVA measure lies?" deserves a scientifically honest answer: no — no machine measures lies. LVA measures something more fundamental: deviations from individual emotional equilibrium (homeostasis) at the moment of speaking.
When someone deliberately deceives, the brain activates a complex set of cognitive processes that run simultaneously: retrieving the false information, monitoring its consistency, tracking the conversation partner's reaction, and suppressing the truth. This measurably increases cognitive load — and that leaks, unconsciously and uncontrollably, into the voice.
In addition to possible deception, LVA also measures:
Why are professionals increasingly choosing LVA voice analysis over the classic polygraph method?
| Feature | LVA™ Voice Analysis | Polygraph test | Cognitive interview |
|---|---|---|---|
| Invasiveness | ✓ Non-invasive (voice only) | ✗ Electrodes, cuff, sensors | ✓ Non-invasive |
| Remote analysis (phone/recording) | ✓ Fully possible | ✗ Physical presence required | ◑ Limited |
| Language dependency | ✓ Fully language-independent | ✓ Language-free | ✗ Highly language-dependent |
| Validated accuracy | ✓ 94–98% (peer-reviewed) | ◑ 80–90% (context-dependent) | ◑ Methodological |
| Real-time operator feedback | ✓ Live Emotional Diamond | ◑ Visual graph readout | ✗ Not applicable |
| Scientifically peer-reviewed | ✓ 15+ international publications | ✓ Extensive | ✓ Extensive |
| Countermeasures by test subject | ✓ Very difficult to manipulate | ✗ Known countermeasures exist | ◑ Partial |
| Emotional Diamond profile | ✓ 14 dimensions, per question | ✗ Not available | ✗ Not available |
LVA technology has been the subject of extensive scientific research by independent universities and institutions worldwide. Below is a selection of published research supporting the reliability and validity of the technology.
LVA voice analysis scores 94–98% accuracy in peer-reviewed research (CRIMSA, 2001; AFRL, 2002). The technology does not detect "lies" as a binary category, but measures deviations from the individual emotional baseline (homeostasis) — a neurophysiological fact, not interpretation.
A polygraph test measures physiological reactions via electrodes: skin conductance, blood pressure and breathing — requiring physical presence. LVA voice analysis only requires a voice recording, by phone, recording or in person. LVA is non-invasive, language-independent and analyses 150+ voice parameters that are far harder to consciously manipulate.
LVA analyses parameters that are partly unconscious and uncontrollable — particularly the neuro-motor coordination of speech production. Known polygraph countermeasures have no demonstrable effect on LVA parameters. That said: no method is 100% foolproof.
A full test has three phases: (1) Intake and preparation: approx. 30–45 minutes. (2) Calibration and test: 45–90 minutes depending on number of questions. (3) Data processing and report: 1–3 working days depending on chosen package.
Yes, always. Explicit, informed consent is required under GDPR. The test subject signs a detailed consent form before the test begins. Hidden recordings or analyses without consent are legally prohibited and ethically indefensible.
In the Netherlands, LVA results are not automatically admissible as evidence in criminal proceedings. They are regularly used in civil proceedings, employment law, family law and as supplementary forensic evidence. We always advise consulting a lawyer regarding the specific legal application.
Yes — this is one of LVA's greatest advantages. Analysis can be performed on phone recordings, video conferences or existing audio recordings. The technology is validated for use with standard microphones and over telephone lines.
All processed audio data, test results and personal data are handled in accordance with GDPR. Data is used exclusively to fulfil the assignment and destroyed after completion in accordance with legal retention periods.
All test packages include the complete LVA™ voice analysis, intake, test administration by a certified operator and a verbal result. The only difference is the number of questions.
Would you also like a written report? Then choose a report package as an addition — see further down this page.
By default you will receive a verbal result from the certified operator after the test. If you also want a professional written report — for example for a legal file, HR procedure or personal archive — choose one of the report packages below as an addition to your test.
Prices exclude any travel costs for on-site testing outside the office (Nieuwerkerk a/d IJssel). Custom pricing available for organisations and multiple test subjects. Contact us for a no-obligation quote.
Our team of certified LVA operators is available for a no-obligation consultation. We will help you determine which test configuration best suits your specific situation.