Voice Authenticity FAQ
Common questions about ORAVYS voice authenticity analysis, bio-acoustic technology, deepfake detection, privacy, and pricing.
General Questions
Q What is voice authenticity analysis?
Voice authenticity analysis examines the bio-acoustic properties of speech to determine whether audio is genuine human voice or synthetically generated. It also detects incongruences between verbal content and vocal biomarkers, such as stress patterns, cognitive load indicators, and emotional micro-expressions in the voice. ORAVYS uses 110+ specialized engines analyzing 137 acoustic features per audio frame to provide comprehensive voice authenticity assessments.
Q How is this different from a voice stress analyzer?
Legacy voice stress analysis tools typically rely on a single acoustic feature (often micro-tremor frequency) and have been widely discredited in peer-reviewed research. ORAVYS is fundamentally different: it uses 110+ independent engines, each examining different acoustic dimensions, powered by a proprietary deep learning ensemble architecture trained on 1.51M+ validated samples. The multi-engine approach provides scientific rigor and resilience that single-feature tools cannot match.
Accuracy & Technology
Q How accurate is ORAVYS deepfake detection?
ORAVYS achieves industry-leading accuracy on deepfake detection, validated using strict speaker-disjoint cross-validation that prevents speaker leakage between training and test sets. The production ensemble was trained on 1.51M+ samples from multiple corpora, applying standard audio augmentation techniques to ensure robustness across recording conditions.
Q What audio formats and quality are required?
ORAVYS accepts common audio formats including WAV, MP3, M4A, OGG, and WebM. For optimal results, recordings should be at least 5 seconds long with minimal background noise. The platform includes built-in voice isolation to handle noisy environments. Audio is internally resampled to 22,050 Hz for analysis. You can also record directly in the browser using the recording interface.
Q What are the 178 engines?
The engines are specialized analysis modules organized into categories: core forensic (21), forensic deep-dive (28), vocal biomarkers (15), personality traits (16), professional metrics (5), voice quality (11), temporal dynamics (3), relationships (4), and meta-analysis aggregators (4). Each engine independently analyzes specific acoustic dimensions and contributes a confidence-weighted score to the final assessment.
Privacy & Compliance
Q Is my voice data kept private?
ORAVYS takes privacy seriously. Audio samples are processed in secure cloud infrastructure hosted on Google Cloud Platform and are not shared with third parties. Users can request deletion of their data at any time. All processing complies with GDPR requirements. The platform does not store audio longer than necessary for analysis completion, and voice data is never used to train models without explicit consent.
Q Is ORAVYS compliant with the EU AI Act?
Yes. ORAVYS is designed from the ground up to comply with EU AI Act Article 50. The platform provides probabilistic assessments with full methodology transparency. It never makes deterministic claims about individual behavior or labels anyone as deceptive. All outputs include confidence intervals, per-engine breakdowns, and clear disclosure of the analytical methods used. ORAVYS is a decision-support tool for qualified professionals, not an automated judgment system.
Pricing & Access
Q How much does ORAVYS cost?
ORAVYS offers flexible pricing: $14.90 for a single analysis, $49/month for the Pro subscription (unlimited analyses), and $299/month for Business teams. Enterprise and academic pricing with API access and custom SLAs is available on request. Visit the pricing page for full details.
Q Is there an API for developers?
Yes. The ORAVYS Enterprise API provides RESTful access to the full 110-engine analysis pipeline. It supports Bearer token authentication, tier-based rate limiting, and returns structured JSON with per-engine scores, anomaly flags, and metadata. A JavaScript SDK is also available for browser integration. See the contact page to request API access.
Q Can I use ORAVYS for academic research?
Absolutely. ORAVYS offers an Academic API program with discounted rates for university researchers. The platform's standardized 110-engine output provides rich feature sets for studies in psychology, linguistics, communication sciences, and speech pathology. Contact us for academic pricing and batch processing capabilities.
Still Have Questions?
Our team is here to help. Reach out for technical questions, enterprise inquiries, or partnership opportunities.
Contact Us Try a Free Analysis