The proposed method (place a resistor across the preamp output, listen for volume change) for measuring the preamp output impedance is a blunt instrument, at best, absent proper instrumentation. Here's why: Suppose that a preamp is able to produce an open cicuit voltage Vs. If its internal impedance Rs is much less than the load impedance Rload, then Vload = Vs. When a resistor Rtest << Rload is placed in parallel with the preamp output (and therefore, in parallel with Rload), then Rtest dominates the load seen by the preamp. Thus the voltage delivered to the load is Vload = Vs*Rtest/(Rtest + Rs). This means if Rtest = Rs, the voltage drops to 0.5 Vs, which is a 6 dB drop in power to the load (power delivered to Rload is P = (Vload**2)/Rload). Let Vratio = Vload/Vs, and we can rearrange the equation to get Rs = Rtest(1-Vratio)/Vratio. For a 3 dB drop in power, Vratio = 0.707, meaning Rs = 0.41 Rtest. For a 1 dB drop in power, Rs = 0.12*Rtest. So, depending on how sensitive you are in detecting a volume change, there can be very wide variability in your estimate of source impedance (in this example, a 1-6 dB volume range corresponds to an 8:1 range of estimated source impedances). Bottom line: to measure source impedance, careful instrumentation is needed. The proposed test, without measurements, can only given an order-of-magnitude estimate at best.
