In our proposed framework we present a novel watermarking procedure to embed for copyright protection and authentication into digital audio by directly changing the audio samples then after modifying the audio signals. The modified audio signals are divided into no. of samples each sample is decomposed adaptively by the method of novel Empirical Mode Decomposition (EMD).those decomposed samples after decomposition called as a Intrinsic Mode Functions (IMFs),In this Intrinsic Mode Function  the low frequency  mode  table under different attacks is presented and then after audio perceptual quality of the original audio signal is preserved. Finally in our proposed algorithm we show the robustness efficiency of the hidden watermark for additive noise re-quantization, MP3 compression,, filtering, cropping and re-sampling. A comparison analysis shows that our proposed framework has high end performance than the other watermarking schemes proposed recently in the literature.