updated the SignalGenerator to generate spikes of varying amplitudes and at a lower rate. Minor incremental changes were made to other files

changed the way that the SpikeDisplayCanvas lays out its children plots. It uses a more dynamic approach that allows plots of mixed types to be layed out in a (sort of) clean manner.

Also fixed various bugs

removed old plotting files (Tetrode, Stereotrode, Electrode) plots. They have all be replaced with SpikePlot.{h,cpp}

Signal Generator now generates fake action potentials!!! The phase dial can be used to scale up and down noise on the spikes

initial commit of SpikePlot.cpp and SpikePlot.h which define a general spike plotting object that can completely replace ElectrodePlot, TetrodePlot, StereotrodePlot. SpikePlot objects can be added to the canvas and render fine, however, eventually the Tetrode variant of the SpikePlot causes a segfault, probably due to the way that spikes are copied around in the SpikeDisplayCanvas.

began porting code from stereotrode plot to tetrode plot. It shows up in the canvas and actually shows spikes after the display is panned or zoomed at least once. Although spikes only show up in 2 of the 3 waveform axes

fixed bug in SpikeDisplaCanvas that was preventing spikes from getting rendered. The variable samplePosition was created but not initialized. I set it to 0

Previously, traces for each channel could not overlap, meaning that
channels would appear to artificially saturate. There are now two
parameters that determine the location of an LFP channel: plotHeight
and interplotDistance. Thus, plots can remain close together while
keeping a large range.

In order to improve rendering efficiency, the screenBuffer was changed
from a Juce AudioSampleBuffer to an array of GLfloats. Instead of specifying
each individual vertex, OpenGL receives a glVertexPointer. Exact timing
improvements have not been measured, but they should be significant.

The SpikeDetector now uses SpikeObjects to pass spike messages,
rather than the GenericProcessor "addEvent" code, as previously.
This makes it easier to distribute spikes to multi-channel electrode
displays, such as StereotrodePlot and TetrodePlot. However, there's
still a problem with setting the axes. "Real" spikes have an issue
with constantly changing the axes, whereas simulated spikes do not.