This paper describes estimating signals that are sparse superpositions of sinusoids. Overall it is clear and with strong empirical results. However, given how much it builds upon the past work in [23], I don't think it meets the standard for an oral presentation.