I'm a little unclear about two points in your letter, so I'll have to back up a step and answer you -- perhaps -- in greater detail than you were expecting. First, there is no reason you can't buy a receiver and a pair of speakers now and then add more as you need them or can afford them.

However, it will be a lot easier if you do some planning now -- and if you can pre-wire your walls to accommodate your future plans, so much the better.

As faithful readers know, I'm just nuts about the single-box Linn Classik radio/CD player/amplifier, which costs about$1500. If you're looking for simple and special, it's the way to go.

In-wall and in-ceiling loudspeakers have come a long way in the last few years, and they can sound quite good. I think in-wall speakers -- mounted in the wall your favorite chair faces -- can sound quite good, especially if you get the kind with "aimable" tweeters. In-ceiling speakers won't image properly, but unobtrusive may mean more to you than better sound (or, more properly, my take on better sound). Niles, Sonance, Atlantic Technology, NHT, and Polk Audio all make high-quality models -- some, like the Niles HD650, even include built-in equalization to compensate for their on-boundary placement.

Now two details about the outlying rooms, The first is that you might as well put in pairs of speakers in each room, rather than a single speaker -- speakers are sold in pairs, after all, and a pair of small speakers can play softly and keep the sound uniform over a larger area than a single speaker, which needs to be played louder to fill the whole room.

My second point is that you're better off running long runs of interconnect than speaker cable -- it suffers less signal loss and can even be cheaper. That way, you can put small amplifiers, such as Russound's $150 DPA12, and drive in-wall or on-wall loudspeakers in as many rooms as you like.

I am still using the Slink-e, but I have to say that my appreciation of Nirvis has slipped a bit since they discontinued their production of their DXS digital switching unit, which allowed several megachangers to run through a single DAC. If the company had chosen to discontinue it because too many other companies were competing with them, that would be one thing, but, as far as I can determine, there's no other similar device available and Nirvis' discontinuation of it removes one of the biggest reasons I originally recommended it -- its expandability.

Thanks for the nice words. I agree with you that one of the most satisfying aspects of the digital revolution is the degree of control it gives music lovers over their recorded music.

I never cease to be amazed at how narrow a world some audiophiles construct for their hobby. It's getting to the point where everybody is convinced that only they know the one true musical truth and everyone else simply isn't serious about "music."

It seems to me that there's no wrong way to appreciate music. Now that doesn't mean I like everything I hear or that I will value a piece of music simply because you do -- and you're under no obligation to enjoy anything I happen to enjoy. Nor does it mean that some music just doesn't have more depth or meaning than other forms -- but it doesn't matter how much depth or meaning a piece of music innately has if it doesn't speak to you.

The same thing applies, I think, about the means by which we listen to that music. When the violinist Ignaz Schuppanzigh complained to Beethoven about the difficulty of his violin concerto, Beethoven scornfully responded, "When I composed that, I was conscious of being inspired by God Almighty. Do you think I can consider your puny little fiddle when he speaks to me?" I feel the same way about issues such as tubes vs. transistors, analog vs. digital, FETs vs. bipolars -- what do I care about these insignificant details when the issue is Beethoven's ability to speak to God?

Sometimes I want the full-bore audiophile experience. I want to hear my Nu-Vista 3D CD player/Conrad-Johnson Premiere 17LS/MF Nu-Vista 300 system and I resent the slightest loss of resolution -- but there are also times when I'd rather chew glass than go through the whole routine to cue a disc, change it, cue another disc, repeat, repeat, repeat as necessary. That's when I love my MP3 player -- and who can gainsay me that?

Well, the Stereophile recommendation means that the product was covered in the magazine and that Stereophile's reviewers liked it (1) enough to recommend it and (2) even better than the other products they liked only well enough to place in Class D.

I'm not trying to be flip -- when you hear as much hi-fi as Stereophile's writers have, it takes a special product to stand out at all. Actually, I'm one of the writers who voted to put the Active/20 there, so I would think that means something, wouldn't I?

In my experience with the Active/20s, if they sounded too hot in the highs, they probably aren't very broken-in (they might also have been toed-in, making the tweeters point directly at your ears, which is not the best choice in most rooms).

This is not to put the Heybrooks down -- I'd place them in Class C, too. While they do have the slightly softer "British" sound, frequently described as "polite," they don't lack transparency one whit. The Heybrooks have a less brash energy than the Active/20s and you might prefer that sound, especially if well-played-in Active/20s sound too tizzy for you.

I think the Linn would be a great match for the CDM1s. You certainly could put more power on 'em and they probably would open up a bit, play a bit louder -- you know, essentially give you slightly more than the Classik alone, but for more money. But if simplicity is your goal, why complicate matters unless the results are spectacularly better?

And, even though I have praised the Classik quite lavishly, every time I return to it after an absence, I wonder if I have adequately described just how special it is. No, it isn't perfect, but it makes so few sonic sacrifices and does so much so well that I find myself completely ignoring the tiny audio critic in me that manages to boss me around most of the time. As a result, I forget all that hi-fi stuff and simply experience the music. You might react differently, of course, but I have noticed that many of the audiophiles I respect most -- guys like Art Dudley and John Atkinson -- just happen to own Linn Classiks in addition to their "he-man" audio systems.

The Classik probably is better-sounding than your megachanger -- some changers, especially in combination with outboard DACs, can probably offer even greater resolution, but that's beginning to complicate matters.

Yes, I do have a Panache in house, and I am in the process of writing a review of it to run in December. Actually, I'm in the process of comparing it to a few other integrateds right now, so check back in on the first and find out just how much I like it -- oops, did I show my hand?

The Rega Planar 3 is, perhaps, the greatest bang-for-the-buck component in all of audio -- it's nearly impossible to conceive of anything else that delivers so much of the unalloyed musical pleasure of the really expensive format leaders. With any step up you make from the Planar 3, you pay a lot more money for refinements on the basic sound you're getting now. Stepping up to a Sondek LP-12 is a big step, but you get a much wider soundstage, tones with more detail and inner voices, and, to my ear, further insights into the sweep and architecture of the music. (Linnies tend to call this pace, but I think what it actually does is reveal the bones and ligaments that connect and support music's muscle system.)

Of course, you don't have to buy a fully tricked-out LP-12 -- there are about a quarter million used Linns out there, and you can buy one and bring it up to current spec one upgrade at a time.

I also like the VPI HW-19 Jr., which you can buy for $700 (sans arm) and which can be upgraded to HW-19 status over time. This deck sounds quite different from the Linn, and you should hear both, since few people find tables equally persuasive.

Music Hall's MMF-7 ($999 including a Project Nine tonearm and a Goldring Eroica high-output MC cartridge) is a sleeper that doesn't get nearly the praise it deserves. I think what it does best is keep everything so nicely balanced.

Richard Shahinian is one of audio's unsung heroes -- his designs remain fairly stable for decades (the Obelisks have been around for 20 years and have been changed only to keep pace with raw driver improvements). I have never heard a Shahinian design that wasn't balanced, musical and engaging -- and this, together with the stability of his designs almost guarantees that he gets little press.

Let's face it: Audio goes through trends continuously, but music itself doesn't change much. Neither do Shahinian loudspeakers. They're also relatively easy to drive and really easy to place (they roll). I'll explore the whole story of Mr. Shahinian and his speakers in the upcoming year, but you should definitely listen to the Obelisks if you get the chance.

Now for the technical part of your question. When electronic circuits are called "balanced," that refers to a circuit topology known as differentially balanced. This is used a lot in pro audio applications, such as recording studios, where issues like cost take a back seat to protecting the signal from extraneous noise and interference.

To transmit a signal (information) electrically only requires two wires (conductors) -- in audio, one conductor is usually connected to a component's chassis (ground). This two-conductor transmission system is called single ended and it's cheap and relatively simple. It is also extremely vulnerable to interference, since any noise induced in the ground will be carried just like the signal is -- and the next product in the chain won't be able to tell the difference between the signal and the noise.

If you use two cables to carry the signal (called the transmission leg and return leg) and connect a third cable to ground, but not to either signal-carrying cable. Any noise transmitted to the cables will be in phase, while the two signals will be out of phase with one another. The shared noise signals are called common mode currents and the signal is called the differential. The receiving end of the connection will have a differential amplifier (a device that reads only the differences between the signals -- all common-mode noise is ignored). Thus, devices that employ this type of balanced circuit are called differentially balanced and the rejection of noise is called common-mode rejection.

Balanced cables, obviously, have to have the three conductors and typically use XLR locking connectors.

The purist approach to balanced circuits uses two sets of identical components for the two halves of the signal -- which obviously adds to parts cost and the complexity of these products. Some manufacturers get around the added cost by using signal splitters rather than fully differential circuits and some audiophiles think this is akin to cheating.

Other audiophiles (and audio manufacturers maintain that balanced circuits make sense in the electrical environment of stage, screen, and studio, but the short signal runs and relatively tame home environments most hi-fis live in don't require the "overkill" of balanced operation).

I've had mixed results, myself, but using balanced connections has never made the sound worse in my experience.

As for amplifier classes, that's a form of short hand used in describing how amplifiers operate. Amplifiers can use vacuum tubes or several different kinds of transistor, but they really all do the same job -- the big differences lie in how they derive the muscle to do it.

Basically an amplifier doesn't really magnify a signal so much as copy with a bigger, more powerful copy (accept a low-level signal and re-create it as a signal with greater power). Over the years engineers have divided the amplifier types into classes: Class AB, class B, and class A are the most common. For the sake of argument, we'll use an amplifier that uses two identical output devices per channel (technically known as a half-bridge push-pull circuit).

Pure class A's output devices remain fully on whether conducting current or not, which means 80% of their power gets dumped into the air as heat. The loss of efficiency does pay off, however, since crossover distortion does not exist in class A.

There's also a class C, used mostly in AM radio transmission, and class D. Class D, or pulse-width-modulation amplifiers, have existed for years (Infinity came out with an entire line of them in 1974), but are only now becoming both practical and impressive -- many observers believe they will dominate all consumer electronics within 20 years. There have been so many recent developments in this arena of amplifier design that I will devote an entire feature to class D amplification soon.

I get more questions about the Hsu than any other subwoofer line. I've examined them and I've copyedited reviews on the Hsu subs (say that fast five times!) by writers I respect, all of which were incredibly laudatory. But I've never auditioned a Hsu myself. Obviously, I need to get my hands on some.