Anders Petersson

@pdales, I fixed the typo in the user's guide. Note that the seismic moment reported by SW4 on startup uses the correct formula. Thanks, Anders

Once your new feature is properly implemented all of the current test need to pass when skip_precursor is enabled. Follow Bjorn's advise above. To properly compute the gradient, the adjoint...

@wliuphd I am quite concerned about your changes. In moptmain.C I see code with a lot of ad hoc scaling factors based on t0, but nothing about the source frequency....

The fundamental problem with your approach is that it doesn't take the spread of the source time function into account. For this reason there will be cases where the truncation...