Emitter voltage undershoot and other creepy bits

[asgard]

Enclosed here I have put a zip file with an LT-Spice netlist, and associated plot of the simulation.
This circuit represents a portion of a project I am working on which has a fairly large array of LED's in
a Charlieplex configuration.  The top power is the driving source on the collectors of the switching transistors.  The bottom two are configured to drive the transistor bases as a two-bit binary counter.
I have some questions regarding the plot of the Spice run.

1. The current switches are turned on only when the two bits of the counter differ (XOR function).
    The voltage at the emitters don't seem to be similarly limited.  Each one will experience a one-bit time
    period where the voltage at the emitter is elevated past the LED turn-on level, yet the indicated
    current is near-zero.

2.  The simulation shows the emitter voltages have an undershoot of nearly -0.3V at the falling edge of
     the pulse when simultaneously the diode current pulse is also turned off.

3.  Correct me if I'm wrong, but the the emitter voltage drops about 0.7V when the diode current is turned on.  Is this the result of the base-emitter junction forward voltage drop?

You all can probably tell I am not normally an "analog" guy, right? 

[alm]

Posting a screenshot with the relevant details makes it easier for people to see your problem without having to download your circuit. I attached one for convenience.

1. The current switches are turned on only when the two bits of the counter differ (XOR function).
    The voltage at the emitters don't seem to be similarly limited.  Each one will experience a one-bit time
    period where the voltage at the emitter is elevated past the LED turn-on level, yet the indicated
    current is near-zero.

If both the base and emitter are at approximately 3.3V, no current will flow through the base, and hence through the emitter.

2.  The simulation shows the emitter voltages have an undershoot of nearly -0.3V at the falling edge of
     the pulse when simultaneously the diode current pulse is also turned off.

Capacitance of the diode junction? Putting a small cap in series makes it worse. This would be basically a tiny version of a switched cap circuit as used for charge pumps.

3.  Correct me if I'm wrong, but the the emitter voltage drops about 0.7V when the diode current is turned on.  Is this the result of the base-emitter junction forward voltage drop?

Yes. The 3.3V when both are turned on is probably collector-emitter leakage through the transistors.