Recent studies show that multiwavelet methods give consistently better performance over traditional methods, such as denoising. It is because multiwavelet gives better representation by using more than one scaling and wavelet functions which have shorter support while keeping desired properties such as orthogonality and symmetry. However, there are still many burden in applying discrete multiwavelet transforms to scalar signals, especially the prefilter design for higher multiplicity (>2). In this paper, we first suggest a simple and direct prefilter design method for any multiplicity. The resulted prefilter is orthogonal and possess second order approximation power. Then, we investigate the performance of higher multiplicity wavelet with the proposed filter on the application of denoising. Simulation results show that the higher multiplicity wavelets usually give better denoising results.