AI Music Detector by SiliconSense

SiliconSense provides an advanced music analysis tool to determine the origin of audio tracks. Key features include: • Probability score for synthetic origin • Detailed report on spectral artifacts • Timeline indication of specific generated segments • Server-side analysis with immediate file deletion This free online detector is designed to identify traces from popular music generation platforms like Suno and Udio. It functions by analyzing the unique spectral "fingerprints" left behind by neural vocoders during the creation process. The system employs two models: a robust CQT (Constant Q Transform) model for resilience against pitch or tempo manipulation, and a spectral artifact analysis specifically targeting the 1–8 kHz frequency band. Users receive an estimate of the likelihood that a track was synthetically produced, along with specific details on where and how these artifacts manifest. The tool is ideal for content creators, researchers, and anyone needing to verify the authenticity or origin of audio content. It supports various audio formats including MP3, WAV, M4A, FLAC, and OGG, with analysis focused on the first five minutes of tracks up to 80 MB. The service ensures privacy by deleting all uploaded files immediately after analysis, providing a secure and anonymous way to gauge audio authenticity quickly.