tproc v2 updates

[meeg]

• In tProc v1, when using a generator with a digital mixer you need to specify your frequencies relative to the mixer frequency, as if you're setting an IF. For complicated reasons, you also needed to use relative frequencies when setting the frequency of a readout that was frequency-matched to such a generator. In v2, frequencies are always absolute.
• Some tweaks to avoid crashes in the case where you're using firmware that has no generators or no readouts.
• It turns out that the definition of the "sigma" parameter for Gaussian and DRAG pulses has always been wrong - the standard deviation of the resulting pulse would be smaller by a factor of sqrt(2). This is fixed for tProc v2, but tProc v1 retains the old behavior, so as not to mess up people's existing programs.