Dear All, I have a problem with my Matlab source code written to solve the generalized nonlinear Schrödinger's equation using a split-step solver. The code includes the convolution integral with the Raman response function using the built-in function "conv", but the results do not match the experimental ones. The spectra seems to be almost mirrored to the central frequency as compared to the correct results. The same problem occurs when I solve the NLSE including the higher order nonlinear terms. I can not find my mistake. Is there any proper source code to solve the GNLSE?