Synopsis
		sfdsrtomo < in.rsf > out.rsf reco=reco.rsf grad=grad.rsf flag=flag.rsf mask=mask.rsf debug=debug.rsf adj=n offset=n[1] velocity=y velocity=y shape=n offset=n[1] verb=n niter=5 cgiter=25 tau1=1.e-3 tau2=1. angle=5. thres=0.1 eps=0. tol=1.e-6 rect#=(1,1,...) what=

Parameters
		bool adj=n [y/n] adjoint flag (for what=linear)   float angle=5. angle (degree)   int cgiter=25 number of conjugate-gradient iterations   string debug= auxiliary output file name   float eps=0. regularization parameter   string flag= auxiliary input file name   string grad= auxiliary input file name   string mask= auxiliary input file name   int niter=5 number of inversion iterations   int offset=n[1] offset   string reco= auxiliary input file name   int rect#=(1,1,...) smoothing radius on #-th axis   bool shape=n [y/n] shaping regularization (default no)   float tau1=1.e-3 tau1   float tau2=1. tau2   float thres=0.1 threshold (percentage)   float tol=1.e-6 tolerance for shaping regularization   bool velocity=y [y/n] if y, the input is velocity; n, slowness squared   bool verb=n [y/n] verbosity flag   string what= what to compute (default tomography)