AlienSeas Demo: Medusa Pelagia Noctiluca

This video demonstrates the relation between an image and its representation in the audible space. All frames have been obtained from a single picture representing a Medusa Pelagia Noctiluca in all its details. To obtain these frames I used the singular-value decomposition (SVD) technique to factorize a matrix M in the product of three matrices U Σ V :

• U is an m × m real unitary matrix;
• Σ is an m × n rectangular diagonal matrix with non-negative real numbers on the diagonal;
• V is an n × n real unitary matrix;

The diagonal entries of Σ are known as the singular values. Hence, the low-resolution frames are obtained taking at each iteration a different number of singular values, multiplying the matrices together to reconstruct the image.

These singular values have been transformed into frequencies by a sort of affine transformation. As you can hear and see, the first image is equivalent to a single sin oscillator. Therefore, adding more details to the image also means adding more harmonics to the audio signal, which also means more noise!

This demo is part of a wider project called AlienSeas currently under development.