True bypass, cables and buffers
A recent addition to "conventional wisdom" for guitar players is the notion that you should add a buffer at the head of your pedalboard (but after wah and fuzz, which like to "see" the guitar) if you're running a bunch of true-bypass pedals between your guitar and amp. The rationale is that you should isolate the guitar from changes in capacitive load caused by switching engaging any of your pedals. As with most things about which guitarists obsess, there's a grain of truth behind the reasoning, but the applicability of the observation depends upon the specifics of your rig; there is no hard-and-fast rule to apply.
The real problem with daisy-chained true-bypass pedals is that even the short interconnects add up pretty quickly because the connectors add capacitance. On the brand of cables I use (a common brand, but I'm pretty sure they're all in the same ballpark) the pair of connectors adds about 50pF of capacitance; that's equivalent to a couple feet of decent cable. So it you look at each of those six-inch interconnects as if they're adding two or three feet to your cable run, you begin to understand the problem. On top of that, each pedal adds its own capacitance when it's in bypass.Let's look at specifics. Assume we have four or five true-bypass pedals hooked together with half-foot jumpers, plus a 20-foot cable from guitar to board and a 10-foot cable from board to amp. Do the measurements and the math on this rig and you'll find that there's the equivalent of about a 40 foot straight run between guitar and amp when all the pedals are in bypass. By itself, that's not a terrible thing. What it will do is to push the resonant peak of the guitar's pickups to a lower frequency and cause the guitar to lose more highs when you roll off the volume. Again, that's not intrinsically bad. Some players even use long (or high-capacitance) cables to tailor their sound; it's a great tool for taming a shrill guitar/amp combination or for giving the upper mids just a bit more body. The alleged problem comes in when you engage a pedal. Suddenly, instead the equivalent of 40 feet of cable the guitar "sees" a shorter 20-foot (or more, depending on which is the first pedal not in bypass mode) run of cable to the first engaged pedal; the engaged pedal drives the rest of the cable run with a low-impedance signal so it really matters very little what happens downstream, at least with respect to cable loading.
Cutting the length of the cable run (or capacitance) in half will change the frequency of the guitar pickup's resonant peak by appoximately a half octave (a flatted fifth). That's certainly enough to hear, in a musical sense. Whether it's significant or important depends upon the rest of your gear and the room (the shifted frequency may coincide with a peak or a bump in the frequency response and exaggerate the effect) as well as how critically you're listening; what you can get away with in live performance may (and that's may, not will) mess with your instrument's position in a studio mix.
My own philosophy is that true-bypass is good because it eliminates all of those not-really-transparent buffers from your signal path. But I don't worry about how the changing cable load affects the tone of the guitar; as soon as you engage a pedal - any pedal - you're changing the tone of the guitar anyhow. I run just a few pedals with a 20 foot cable from guitar to pedals and a 10 foot cable from guitar to amp, so I'm in pretty much the same situation as I described above. I set up my EQ with all pedals bypassed so the guitar is driving that approximately 40 foot equivalent cable run, and have no problems getting a very pleasing tone. I don't detect an objectionable change in my guitar's tone when I engage any effect, so I don't bother with a buffer.
Technically it seems like a good thing to add a buffer, but my ears tell me otherwise. Even a good buffer will degrade the signal in subtle but detectable ways. Yes, it's almost certainly subtle enough that the difference would be masked in a live performance; the question, then, is whether the change is the guitar's sound is noticeably different when you engage effects. If I had ten or fifteens pedals and a twenty-foot run from pedalboard to amp instead of four or five pedals and a ten-foot run, I'd bet that the change in going from all-bypassed to first pedal engaged would be significant and I'd probably need a buffer. Like I said, though: trust what your ears tell you.
If you're more concerned about the effect of cable loading on the behavior of your guitar's volume control, a buffer at the head of your pedalboard will certainly limit the amount of treble rollof that you get when you back off the volume control. However, adding a compensating cap (commonly called a "treble bleed") across the guitar's volume control will mitigate the effect of cable capacitance without introducing active circuitry into your signal chain.