Flag for adaptspec not converging output

[shicks-seismo]

Hi, was wondering whether it would be please possible to add a switch to mtspec.multitaper for somehow suppressing the output adaptspec did not converge messsages from adaptspec.f90.
I'm not exactly sure why these messages appear, but even when they occur a sensible spectrum is still calculated.

[krischer]

I'm not really that deep into the mtspec library anymore. But something like this should be able to catch the output: https://github.com/minrk/wurlitzer