Audio and image watermarking is one of the important and popular research areas for ownership protection of the media(image, audio, video etc) to prevent illegal distribution of the media and can also be used for media authentication, broadcast monitoring. A lot of work has been done on images, video watermarking, Audio watermarking as image and video use visual characteristics for watermark embedding, on the other hand, audio use human auditory system (HAS) properties for watermark embedding and watermark detection. A prior knowledge may not be available about various quantization steps that generate the maximum robustness and imperceptibility, thus DE is adopted in our watermarking scheme. In this paper, we propose a new method of DWT-DCT-SVD method are proposed using the DE algorithm to optimize the quantization. The time or frequency –scale modification leads to desynchronization and cause the false detection error, which should be avoided. For the synchronization problem, one way to solve it is to bind the watermark with the invariant audio features. With this method the watermark can be detected without synchronization error. Our scheme, the second way is used to address the desynchronization problem. Synchronization code is adopted in front of the watermark to orient the location of inserted watermark. DWT-DCT coefficients, are split into different segments, the segment length is obtained by dividing the length of the approximate coefficients by the watermark length. © 2015 Elsevier B.V., All rights reserved.