Allow system to produce arbitrary number of points in a trial

[raacampbell]

This is largely in place now: f5e8d72 (and: 8991e7a).

• To enable this, I have used a convolution instead of the ad-hoc interpolation to smooth transition between points. This does not produce identical traces to before, but they are very close indeed. I need to verify that the new approach does not alter the required blanking parameters. If so, we can updated the test waveforms and go with the new system.
• Also need to check that the laser power is being calculated correctly for smaller number of points.
• Put in place a system for automatically limiting the number of possible points based on desired laser power, max available laser power, and the blanking time.
• Get rid of the stimulusDuration option from the stimConfig file. It's going to be complicated to handle, since its max value will depend on the number of stimuli.

The current stimulus pattern is 40 Hz stimulation because we are presenting the stimuli over a 25 ms period.

[raacampbell]

Looking on the scope, I see no difference in blanking behavior between the ad-hoc and filtered versions of the command waveforms. Good!

[raacampbell]

The convolution is good and does not alter the waveforms in any significant way eca2ede

[raacampbell]

See bug #150 related to this new feature.

[raacampbell]

It is now very clear in the instructions that the user is responsible for measuring power per point. So maybe we don't need to limit max points.