It is well accepted that the conversion of an evanescent wave into a propagating wave is critical to far-field subwavelength imaging. However, subwavelength resolution can also be achieved using the multiple signal classification (MUSIC) algorithm for the situation of low conversion. In order to explore the difference of imaging performance between these two approaches, an acoustic superlens of length about one wavelength is designed to convert the evanescent wave into a propagating wave, which can be harnessed by the MUSIC algorithm. It is confirmed that the conversion of the evanescent wave into a propagating wave plays a role in improving the imaging resolution against noise, and the imaging resolution is improved by both the MUSIC algorithm and an acoustic superlens.