to illustrate what AudioGuru is talking about, look at the figures below. in a "real-world" application of the circuit you are testing, we will use a similar emitter follower in an amplifier. you will see two traces, one with 10 ohm emitter resistors (R15, R16) you will notice that the output clips at +/-4V. this is because the amp itself is swinging almost rail to rail, but your 10 ohm emitter resistors are acting as a voltage divider with the load, reducing the output voltage. in the second trace, the emitter resistors are 0.22 ohms. the full output swing (almost all that is...) is now across the load, and not being wasted in the output circuit. the input signal level and the overall gain of the amp remained the same. also notice that there is a driver stage, and a bias transistor. the bias transistor Q4 performs the same function as the diodes in your circuit, with R9 and R10 being used to set the conduction voltage of Q4. the outputs are actually wired as darlingtons, so Q5+Q7 act as one composite transistor, and Q6+Q8 act as a composite transistor.