This work addresses the issue of retrieving efficiently sound samples in large databases, in the context of digital music composition. We propose a sequence generation mechanism called musical mosaicing, which enables to generate automatically sequences of sound samples by specifying only high-level properties of the sequence to generate. The properties of the sequence specified by the user are translated automatically into constraints holding on descriptors of the samples. The system we propose is able to scale up on databases containing more than 100.000 samples, using a local search method based on constraint solving. In this paper, we describe the method for retrieving and sequencing audio samples, and illustrate it with rhythmic and melodic musical sequences.