When performing hyper-high resolution mass data integrations, the bin
size as computed by the res/da/ppm precision object is not small enough
to plot nice spectra. This is particularly true when generating
theoretical isotopic clusters with lots of points for the shape of the
individual peaks, for a given resolution. The best bin size divisor is,
empirically, 6. That is, the bins must have a size corresponding to the
bin size computed by the precision object divided by 6.

Update to upstream version '0.9.6'
with Debian dir 23edcd2565cb799dd695942ac6a03883ea74975a