We introduce a representation for rhythmic patterns that is insensitive to minor tempo deviations and that has well- defined behavior for larger changes in tempo. We have com- bined the representation with an Euclidean distance measure and compared it to other systems in a classification task of ballroom music. Compared to the other systems, the pro- posed representation shows much better generalization be- havior when we limit the training data to songs with different tempi than the query. When both test and training data con- tain songs with similar tempo, the proposed representation has comparable performance to other systems.