sfistpad (2.0-git)
n-D IST interpolation using a generalized shrinkage operator and zero-padding

        sfistpad < in.rsf > out.rsf mask=Fmask.rsf pad= verb=n pow2=n niter=100 pclip=10. normp=1. n#= mode=
Note: Acquistion geometry specified by mask operator

string mask=
auxiliary input file name
string mode=
thresholding mode: 'hard', 'soft','pthresh','exp';
'hard', hard thresholding; 'soft', soft thresholding;
'pthresh', generalized quasi-p; 'exp', exponential shrinkage
int n#=
size of #-th axis
int niter=100
total number of iterations
float normp=1.
quasi-norm: normp<2
ints pad=
number of zeros to be padded for each axis [dim]
float pclip=10.
starting data clip percentile (default is 10)
bool pow2=n [y/n]
round up the length of each axis to be power of 2
bool verb=n [y/n]