If you are evaluating Voice Stress Analysis software for your agency, this guide walks through what modern VSA platforms should offer — and why a growing number of agencies are choosing DecepTech™ VSA over legacy systems like CVSA® and LVA.
The Voice Stress Analysis category has been around for five decades. Legacy systems like CVSA® (developed in the 1980s) and LVA (Layered Voice Analysis) established the market, but the underlying technology has advanced considerably since then. Modern Voice Stress Analysis is capable of far more than what was possible even ten years ago: offline AI speech recognition, automatic speaker identification, multi-language support, and high-resolution spectral analysis are now achievable on a standard laptop. This page explains what to look for when evaluating VSA platforms, and where DecepTech™ VSA differs from older systems.
Before comparing any two products, we recommend agencies use this checklist to score candidate systems objectively. Not every criterion will weigh the same for every agency, but a thorough evaluation covers all of them.
DecepTech™ VSA version 1.4.3 (our current release) has been designed specifically around the nine criteria above. Here is how the product delivers on each.
DecepTech's algorithm detects the Lippold micro-tremor in the 8-14 Hz band, based on the foundational 1971 Scientific American research by Olaf Lippold. See our Research page for the full bibliography.
Stress indicators appear on the examiner's screen as the subject speaks. No post-processing wait, no separate rendering pass. Examiners can steer the interview based on what they observe live.
Our Vosk-based offline speech recognition engine automatically places word-level annotations centered over each spoken word. No manual transcription required.
DecepTech 1.4.3 uses 128-dimension voice vectors with cosine similarity to automatically label each annotation as [E] Examiner or [S] Subject — eliminating manual speaker tagging.
English (GigaSpeech model) and Spanish editions available, with automatic language-aware model selection. Critical for agencies serving bilingual communities or operating internationally.
All speech recognition and stress analysis runs locally on the examiner's computer. No cloud services. No internet required. No data leaves the machine — critical for sensitive investigations.
An 80-hour Basic Certification covers protocols, practice exams, and chart reading. An optional one-week Advanced Course focuses on case studies and narrative analysis. All certified examiners receive lifetime mentoring at no additional cost.
Works with any recording format. 44 kHz, 48 kHz, 16 kHz, stereo, mono — the software detects the native sample rate and adapts automatically. No file conversion required.
DecepTech checks for updates silently at startup and displays a notification when a new version is available. Version 1.4.3 is the current release; 1.5.x with machine-learning scoring is in active development.
In conversations with agencies moving from CVSA®, LVA, or other legacy systems, these five themes come up repeatedly.
Legacy systems often require examiners to mark each spoken word by hand — an hour of tape can take hours to annotate. DecepTech's AI Auto Annotation does this automatically during analysis, freeing examiners to focus on interpreting results.
U.S. agencies with significant Spanish-speaking populations, and agencies in Latin America, need native Spanish language support. DecepTech ships with both English (GigaSpeech) and Spanish models, automatically selected by the software.
A two-week Basic course is long enough to produce truly competent examiners rather than graduating students who still feel uncertain. Lifetime mentoring means when a complex case arises years later, the examiner has someone to consult.
Agencies want to know their investment won't become stranded. DecepTech's version 1.4.x delivered Vosk integration, automatic speaker diarization, and multi-language support in successive releases, with the 1.5.x ML-based scoring engine in active development.
Agencies that prefer to finance their purchase can do so over 12, 24, or 36 months with a 5% financing charge. Each training seat is valid for two years, giving scheduling flexibility when personnel changes occur.
No single VSA system is right for every agency. Here's a straightforward look at where DecepTech is a strong fit, and where another solution might be a better match for you.
We would rather recommend another product than sell you the wrong tool. If DecepTech isn't the right fit, we'll say so.
Request a live demonstration or download our full product overview. No sales pressure — we'll walk through the software and answer your agency's specific questions.