The Mel Spectogram is a 3D representation of the sound. It has the time in abscissa axis, the frequency in Hz on the orderly axis and the third dimension the volume in dB, represented with a color spectrum from black (-∞ dB) to yellow/white (+0 dB). This representation is widely used for [sound classification](Sound%20classification.md) by applying image recognition patterns with #machine-learning and #deep-learning algorithms. Going further : - [Getting to know the Mel Spectogram - Towards data science](https://towardsdatascience.com/getting-to-know-the-mel-spectrogram-31bca3e2d9d0)