Ever perform a preamp bypass comparision test?

Would you care to elaborate why you would want to spend a lot of time doing this comparison test? Do you listen to music always at unity gain of your source components? Have you considered that unbuffered outputs of source components usually do not interface well with power ampfliers inputs?

First question: I'd want to spend time doing this test in order to learn about my preamps and what they are doing, not only in comparision with each other, but to the unadulterated source.

Second: Of course not.

Third: Not true. It's passive resistive volume controls and long cable runs that source outputs may (as opposed to "usually") not interface well with, not amplifier inputs in most cases.

1) Are you inferring that volume controls, buffer stages, and output switching in preamps do not affect their sonic characteristics? When you listen to rec out or unity gain outputs of a preamp, you are by-passing all of the above mentioned.
3) Not true is not good enough. It depends on the buffer stages of the source components.

Before I respond, let me say up front that I truly appreciate AV_specialist's interest in this thread if it is sincere, but either I am missing something or I need more explanation or justification of his seeming position, so here goes:

1) Umm...Duh! That's kind of the point of the test. (It's actually input switching, not "output switching".) Please go back and reread my description of the test and why I set it up the way I did if you did not pick up on this.

2) You show me the source component with an output stage incapable of driving a typical 1m or 2m cable run into an imput impedance of at least 20K ohms with least 1v to 2v, and I'll show you a source component that can't 'drive' a preamp any better than it can a power amp. Almost any digital player, DAC, or separate phonostage made will provide a low enough output impedance with enough voltage to interface with a typical amp input. Almost all amps (excepting some pro-type gear) will provide high enough input impedance, voltage sensitivity, and gain to be driven by a such a typical source. Of course, for listening you may desire more or less gain, but in general, unless you are restricted to employing very long interconnect runs or use a very low-gain amp, almost anyone could do my proposed test at unity gain and get instructive results. (In fact, even in the case of a setup where the interface was clearly disadvantaged by not employing a preamp, the results would still be equally instructive as far as ascertaining the effects of the preamp. Whether the results are in favor of the preamp or not doesn't negate the value of the test itself.)

AV_specialist, I don't mind being critically questioned, but I don't see what you're trying to get at. Either you're just trying to poke at me for the amusement of it, or you haven't really read carefully what I've written above. Why don't you just come out and tell me your agenda here? You obviously are in some way 'against' the test procedure I've outlined, but I have no idea why. I mean, if you simply prefer not to try this, then don't - but that wouldn't explain your coming on this thread to challenge me (your perfect right to do, and which I welcome), so I must assume you take a greater interest in this subject than merely not caring to bother with it.

If I had to guess why or what that agenda is, then I would suppose this: I believe many audiophiles tend to recoil at the idea of tests such as this because they don't really want to be made aware of the implications of the results. I am not saying that control over attenuation, switching, output buffering, or overall gain is unimportant in a system. I am not arguing against the use of active preamps. I use them myself, and will continue to do so as far as I can see. I do not normally listen to my system without a control center in the loop, and I have not owned a passive volume attenuator (yet). If I were to try one, I have no doubt that it would also suffer in comparision to the sound of the straight source, but not at unity gain as in my test, since that would involve zero attenuation, unlike an active preamp (for whatever that's worth).

But I do not for a second believe that putting the source signal through additional stages of gain and attenuation can in any way "improve" that signal, necessary as those factors may be for system control. However, it is true that a preamp may change the signal in ways that an audiophile could find pleasing, but without some sort of reference, the listener will never know for sure how to attribute those characteristics. The reference could be the sound of other preamps, and that is valid and is often done, but I am suggesting that the reference also include the unprocessed sound of the source itself, because that is the real baseline for determining how intact a preamp is capable of transmitting the input signal.

None of this means that other methods for controlling overall gain (volume) or driving long cable runs will necessarily be superior to a good preamp - they often won't be for any given source component or system layout. It also doesn't mean that in some situations, a preamp may not fortuitously balance out some sonic characteristic of a source component, amp, speakers, or cable, but complementary colorations do not equal transparancy to the source. That preamps exist at all in these days of all-digital rigs and stand-alone phonostages is mostly for ease of system configurability and convenience of control. They are no longer a necessity in the same way a power amplifier is. The saying "The best preamp is no preamp" didn't come into existence for no reason.

You can claim none of this really matters since we are not actually going to listen to music without our preamps in the loop in most cases, and that point is perfectly true as far as it goes, if that is all you are saying. But I think audiophiles ought to be much more skeptical of claims made by manufacturers or in reviews suggesting that XYZ preamp is going to confer all these wonderful properties upon the sound of your system. The best that we can actually hope for is that a preamp do as little harm as possible, while at the same time providing the most good in return when it comes to system flexibility and control.

I advocate that audiophiles would do well to learn more about what their preamps are really doing in their own systems, and remain unconvinced by any protestations from technically aware folks who ought to know better than to argue against the potential worth of performing a simple home test. Such preampophiles will often have a proprietary interest in maintaining preamp fiction, which in turn can lead to the fiction about source components needing all kinds of help in order to interface with amp inputs, when in fact it's cables and passive attentuators they may in fact have difficulty interfacing with. (And on that last item, it's as much [or more] the output impedances and insertion losses of passive attenuators creating difficulties for amp inputs and cable runs, as it is their input impedances creating difficulties for source components, that can cause them to suffer in comparision with active preamps. Despite those potential problems, many audiophiles still report better results using passive attenuators, given appropriate conditions, so I take all the Chicken Little hoopla about the dangers with a large grain of salt. I still want the availability of extra gain myself however.) A little personal knowledge about the performance of one's gear should be nothing to act scared of.

P.S. - If you have problems with what I am saying here, you should click on my threads and read my post about a test setup for looking at DAC performance and the effects of upsampling as implemented in an upsampling-switchable DAC. You ought to love that one...

I don't want to take much more time in this thread as it does not relate to reveiwing products. However, your exercise here seems a bit pointless as 1) You admit that you do not listen to your source components at unity gain, therefore, whatever your findings here would mean that you would need to have some sort of volume control inserted in the path. Perhaps this test would be more revelant if the assumption is that you would hear an improvement because you are by-passing switching, volume control, etc. However, what if one finds the opposite? That is without proper buffering, your source component does not interface well with the power amp. 2) Your assumption that all source components will drive short cables and drive most consumer power amps is questionable. Yes, many audiophiles have reported that passive volume controls work well for them. My experience is that they don't, I can hear what adding passive resistance does when passive resistance of various types are added to the signal path. If the end result is that you will need to use volume controls, switching, and/or buffers, then they should be included in the test.