The code on Github can now read the rotation data from the accelerometer and adjust the amplitude of the buzzing sound accordingly. I also changed the way the accelerometer and ATtiny connect to the Nano like I described in the previous update. The accelerometer connects to the I2C interface on pins A4 and A5 and the ATtiny connects to the hardware SPI on pins D10, D12, and D13. 

Roughly 100 times per second, the Nano reads the gyroscope data from the accelerometer. The magnitude of the rotation is smoothed using a simple low-pass filter, and then converted to an amplitude value, which is sent to the ATtiny. The smoothing step makes sure that the amplitude does not fluctuate rapidly due to noise in the measurement

Let be the angular velocity around the x-axis and be the angular velocity around the y-axis. The first step is to apply the low-pass filter to get the smoothed value , which is done using the formula:

The next step is to calculate the buzz amplitude , which can range from 0 to 255. This is done with the formula:

The constant offset of 50 ensures that even when the lightsaber is held still, some sound can still be heard. All of the calculations are done using integer arithmetic, and the divisions are implemented using bitshifts.

When testing the program, I noticed a slight staticky noise in the buzz when I moved the accelerometer. I suspect this is caused by the signal suddenly changing from one level to another when the amplitude is changed. I will likely program the ATtiny to slowly ramp the amplitude from its current value to its new value to minimize these effects.