AudioStellar

AudioStellar is an experimental sampler designed for the Linux platform that utilizes machine learning techniques to organize and manipulate audio samples. The application functions by analyzing a collection of audio files and mapping them into a two-dimensional space based on their sonic characteristics. This spatial arrangement allows users to navigate through their sample library by proximity, where sounds with similar timbral or rhythmic qualities are positioned near each other.

Users can interact with this map to trigger samples or create new soundscapes by selecting regions within the coordinate system. The software focuses on the exploration of large sample sets, providing a visual interface for sound design and experimental music production. As an experimental tool, it prioritizes non-linear access to audio data over traditional file-based browsing methods, requiring users to pre-process their audio directories to generate the underlying data model.