I said I'd share the experience if I did any experimenting - now there's something to report. Having just removed my old-but-trusty Adcom GFA535 (100w into 4 ohms) and re-installed my two nOrh L2 monoblocks (>350w, same impedance) I tried a half meter of Tara Labs Quantum 66 from the Yaqin buffer in place of Kimber PBJ. Sound: smooth and laid back; some loss of sibilance and presence on male voices on PBS Friday evening talking heads shows. Maybe it was the nOrh amps, which have been out of the system for over a year (don't get me started on why)? So I turned off the Outlaw, unmuted the Panasonic plasma, quickly swapped the Kimber PBJs back in, and turned on the Outlaw again. UNMISTAKABLE improvement in male voice quality, and the credits music suddenly had more space, plus the sense of power the nOrhs can give. So can cables make a difference? They sure did for me!

I think if you did an analysis with lab equipment you'd find that that the high end is being rolled off slightly due to the higher output impedance of the tube buffer interacting with the capacitance of the cable.

Nothing bad about that, but it's just something worth knowing.

Altec, I have no doubt you're right, but that leads to two other questions:
1. Are there any cable characteristics that do affect the audio signal from equipment with low output impedance (i.e., solid state), since capacitance apparently doesn't?
2. With tube equipment should one look for cable with the lowest capacitance to avoid high end rolloff?
Thanks as always for your knowledge.

The interaction of output impedance with cable capacitance is always there - it's just a matter of degree. Solid state has less, but it still has some, especially if the cable is long enough.

Cables can also have inductance, which will also restrict high frequencies. Inductance is typically pretty low however with most cables. Ironically (or not) the high end boutique cables tend to have higher amounts of inductance and capacitance - this is why they tend to "color" the sound to a noticable degree. Again, nothing wrong with that, but the same effect can be had with standard cable and judicious use of resistors and capacitors.

With tube gear, the thing to do is keep the cables as short as possible. If the tube preamp is a cathode follower (like the tube buffer probably is) the output impedance is higher than solid state, but not extremely so. If the signal is taken off the anode, the output impedance is going to be very high, and an extremely short cable must be used.

I use plain RG-59U cable for all my interconnects. It's capacitance is reasonably low and it's bandwidth is guaranteed to be very good (it is, after all, made for 6 Megahertz video signals). The equipment I design makes use of BNC connectors for all analog inputs and outputs because this connector is extremely reliable, with high bandwidth capability. Good quality RCA connectors from companies like Canare are great for connecting to typical consumer gear. They can be had in 75 ohm impedance for use with digital S/PDIF signals.