A raised cosine filter based interpolation for the empirical mode decomposition (EMD) technique is presented in this paper as an alternative to the cubic spline interpolation used in the original implementation. By adjusting a parameter in the proposed technique one can control the quality of decomposition of the algorithm. This results in superior performance compared to the original approach which could either mean faster convergence or lower final error under different conditions. An efficient fast Fourier transform (FFT) based implementation of the proposed technique is presented. Further, an adaptive version of this algorithm is introduced that blindly adjusts the parameter value to ensure efficient signal decomposition for any combination of component frequencies of a two-tone signal.