Some tools don't take kindly to negative values and break in unknown ways (nu_correct and inormalize come to mind). The use of sinc/bspline interpolation during resampling is still worth it, but we should be wary of negative values.
This is one way we could fix this, its what I do whenever I'm concerned about negative values during resampling:
Some tools don't take kindly to negative values and break in unknown ways (nu_correct and inormalize come to mind). The use of sinc/bspline interpolation during resampling is still worth it, but we should be wary of negative values.
This is one way we could fix this, its what I do whenever I'm concerned about negative values during resampling: