Doesn't much matter at audio frequencies.
You can solve for the equivalent capacitance and inductance of a cable, based on its impedance, velocity factor and length.
Leq = len * mu_0 * Z/Zo
Ceq = len * e_0 * Zo/Z / vf
where Z is the characteristic impedance of the cable, Zo is that of free space (~377 ohms), and mu_0 and e_0 are the permeability and permittivity of free space (1.257 uH/m, 8.84 pF/m), respectively.
This assumes velocity factor is due to dielectric constant of the insulator, which is correct for almost all cables. There are some specialty cables that are magnetically loaded, in which case vf contributes to Leq as well. We'll just ignore those.
Incidentally, Zo == sqrt(mu_0 / e_0) so some simplification of this can be done. It's about as much memorization either way.
Use unit conversions as needed. If you input units into Google Calculator, you'll get the right answer out.
https://www.google.com/search?q=6+feet+*+permittivity+of+free+space+*+377+ohms+%2F+50+ohms+%2F+0.67The last sanity check is to verify that this is the correct calculation to make. Take the highest frequency you're using (say, 200kHz, a typical output from an audio generator), and find its wavelength, λ = c / F (speed of light divided by frequency). 200kHz is 1500m. If the wavelength is >20 times the length of the cable (say, >75m), you really don't care about transmission line effects.
You can in turn compare these reactances to your system impedance. At 4 ohms, the capacitance isn't ever going to matter; but the inductance may. 4 ohms and 10uH rolls off (lowpass) at 64kHz. At 600 ohms, it's the other way around: 600 ohms and 4nF rolls off at 66kHz.
Incidentally, it's not a coincidence that those two frequencies are close. I picked L and C such that they would be characteristic of whatever length of 50 ohm cable (uh, ~60m worth, apparently). It just so happens that 4 ohms is almost as many times below 50 ohms, as the times 600 is above 50. That is, 600/50 = 12, and 50/4 = 12.5. This illustrates another handy trick, we can take the 1/4 wave cutoff at 50 ohms and scale it up by the impedance ratio. And the cutoff will always be inductive if below, and capacitive if above.
For audio testing, you usually have high impedance inputs, so it won't matter if you use the 4 or 600 ohm setting, as long as you get the voltage you need. If you need accurate measurement, always verify the output voltage from the signal generator. (You may need to use an oscilloscope to measure this -- common DMMs typically drop off in the kHz range.)
Tim