What should I upgrade next?

It's Sunday, Tiger's on the tube, and I feel in the mood for writing a manifesto.

Ah, the classic divergence of opinion on how to best approach the upgrade situation has emerged here: Front end first? Amplification first? Or speakers first? As can be seen from the above posts, a seemingly good case can be made for all of these solutions. Front end: If this is not providing a high-quality source signal, nothing else that follows will be able to make good sound. Amplifier or Preamplifier: These are the heart of a system - if they are choking off the source or shortchanging the speakers, you'll never get the performance they may be capable of. Speakers: The most important, the most variable, and ultimately the most personal determinant of what kind of sound will be produced, they'll make the biggest change the most quickly. And you know what? All of these presumptions are absolutely true (and that's before even getting into cables)!

So what to do? The conundrum is, that although the whole ball of wax must work together as a system where each element is clearly important, you (and for the most part, all of us) can only upgrade one step at a time over an extended period for practical reasons. One implication that this "system paradigm" portends for the three approaches outlined above - and it is equally true for all of them - is that no one element that you may upgrade along the way is going to be able to give of its maximum improvement until the whole job is completed. And since that is so, then the arguments that go, "You must upgrade this before that, otherwise you won't get any benefit/be able to live with what you hear" become somewhat less than useful to contemplate, in that because they are substantially true in all of the examples, they are therefore by definition also fairly circular in nature.

So we need a better philosophy to guide us in how to proceed. Fortunately, there are some mitigating factors here that will work in our favor. Firstly, while it is true that a lone improvement may either show up a weakness somewhere else, or be largly masked by one, it is also quite unlikely that any legitimate upgrade will actually make the whole system perform worse overall. Secondly (and this contains the exception to the preceding rule that proves it), certain subsystems within the larger overall system are more important than others.

The subsystems, in order of importance: 1) The listener/family (if applicable)/dwelling space subsystem; 2) The listener's musical-sonic preferences/listening room size and layout/speaker type and size subsystem; 3) The speaker/amplifier/speaker cable subsystem. (More subsystems follow: 4) The amplifier/preamplifier/interconnect subsystem; and so on.)

The exception I referred to, which results directly from the ordering of these subsystems: You'll notice that each subsystem, in order, contains within it factors that exert a determining influence over the next ranked subsystem. (For instance, subsytem #1 contains "Who is the listener?", which determines the musical-sonic preferences contained in #2, as well as "What is the home like, and who lives in it?", which in turn determines the listening room factors also contained in #2, and so on.) You'll also notice that choice of hardware is subservient to higher-ranking factors in this hierarchy, and that subsystem #3's ranking implies that the interaction of the speakers with the amplifier (and their connecting cable) is the most universally important subsystem having to do strictly with hardware - which it is (it's always in play, and is the most highly interdependent). So the exception to the rule (about no supposed upgrade making the whole system worse) has to do with upsetting the overall hierarchy of importance of these subsystems, or violating compatability within a subsystem. (Some examples - Choosing: a listening room that interferes with one's family life; a speaker that isn't suitable for use in the room available; an amplifier that isn't suitable for the speaker chosen, etc.)

There is one other factor that makes a rational decision about how to proceed easier to arrive at. Since, in all liklihood, you are going to eventually want to upgrade every piece of gear in the system, it is wise to take into account the future need to be able to meaningfully audition components in your own home, within your own system context. This means that a clear and revealing window, through which one can view unobstructed prospective upgrades elsewhere in the chain, will be of great help in assembling the total system as it progresses. This, in turn, argues against heeding the "Don't choose downstream components that may reveal weaknesses upstream" cautionary warning. If your downstream components aren't capable of revealing the flaws of your upstream gear, how can they be of any help in auditioning new gear to correct those hidden flaws? (Again, a clearer downstream window is unlikely to make the musical experience worse overall, provided there is not a debilitating speaker/amplifier mismatch created, even if upstream flaws are made more clear along with the music.)

By now, anyone reading (who is still awake and following me!) should be able to tell where I am headed. If you combine the necessities created by properly observing the natural hierarchies (of the subsystems relating to lifestyle, musical-sonic preferences, and gear-interdependency) with the provision for auditioning future upgrades, then you will conclude, as I do, that the first choice to made is the speakers, followed by the amplifier (and cables) to drive them, and so on back to the source.

In the case of the original forum question above, I would assume that the #1 subsystem (listener/family/dwelling) has already determined which room is the listening room. The #2 subsystem (listener preferences/room specifics/speaker appropriateness) then comes into play, determining which speakers in general are consistent with both the room and the music that will be played in it. To take the case of the above-mentioned Maggies as a for-instance, the room would need to be large enough to place dipole radiating speakers well away from the walls, small children running through would not be advisable with large lightweight panels present so far out on the floor, and the listener might well be better off preferring chamber music or acoustic jazz to grunge or electronica. If, after narrowing down the choices at dealers or friends, and hopefully being able to audition some of the possibilities at home, the Maggies are indeed chosen, then subsystem #3 (speakers/amplifier/speaker cable) would be the next to receive attention. To promote harmony within the subsystem, the needs of the afore-chosen speakers would have to be catered to - in this case, a low-efficiency speaker mandating a higher-powered amplifier, which would probably mean solid state at the buyer's price point. Then, speaker cable could be auditioned to find the best synergy within the subsystem.

Now that these first steps have established the "clear window" needed to proceed with choosing the rest of an upgraded system, focus can be put on auditioning preamplifiers next, then maybe digital front end components (which form a subsystem of their own if separates are chosen), cables to connect each step, and so on. When a system that is an order of magnitude more revealing than what it replaced has been established, more productive attention can be paid to the various "tweaks" available, now that the tools exist to better evaluate their performance cost/benefit ratios within the system.

So, speakers first and work back from there, but not because speakers make the biggest difference - rather, because choosing them most directly involves the factors highest in my hierarchy (the home and the people who live in it, the music the listener prefers, the properties of the room the system is in), and because this ordering best facilitates making intelligent further system choices through auditioning down the road. (BTW, all of this does assume that the top factors - the home, the family, the preferences, or the room - will not be changing in relevent ways during the time the system upgrading is in progress. This is obviously not always the case, and could be a major cause of a total system reconfiguration.)

Ironically, Audiogon itself, which helps so many of us "live beyond our means" when it comes to improving our systems, can also be responsible for some interesting detours regarding system planning. I know there's been a couple of times I've deviated off point when a piece of gear I was going to look into later popped up on the 'Gon at a hot price. After beginning with my speakers in 1997 and slowly working my through three amps, two preamps, two rounds of cable upgrades, a new CD player and then a new digital separates front end to replace it, two new cartridges, a new standalone phonostage, a new tuner, a CD-R recorder, two new power conditioners, new headphones and an outboard amp to drive them, and different supports and racks, I'm getting near to beginning again at the speakers and amp, but have delayed that upgrade pending a move (while still jumping on a new remote preamp deal recently to audition against my older manual one). I've definitely found that this approach has aided in giving restraint, direction, consistency, and predictable results to my upgrading process, as well as permitting me to enjoy both my music and my gear buys the whole while. Best of luck!

P.S. - Of course, this baby took me so long to unload that now I see Subaruguru has gone and posted basically the same advice before me. As Gilda Radner once said, never mind...

I want to make one additional point as to why, IMHO, I think an audiophile looking to upgrade everything should start at the speakers and work back from there (see my lengthy post above). Kt88 isn't incorrect in his GIGO argument (as I acknowledge above), it's just that a source-first approach, while maybe in some instances sheilding one's ears from stuff about your present system you'd rather not know, does not lend itself as well as the approach I'm advocating to the total process of system improvement. Above, I go into detail on two primary reasons why this is so, summarizable as: 1) The speakers chosen are going to have the most influence on, and the most interaction with, those elements of one's listening environment and listening preferences that are least amenable to errors in the selection, or least likely to undergo change during the system-building process, such as family considerations, the properties of the listening room, and types of music listened to; 2) Ideally, system-building will be best accomplished if one can perform meaningful auditions of gear under consideration in the context of one's own home listening environment and system, and this requires a "clear window" provided by revealing and resolving end-of-chain gear to aid in evaluating preceding components in the chain. I also stressed the importance of making a proper match between the speakers and the amplifier chosen to drive them as the most important subsystem within the chain when it comes to gear selection, with the speakers coming first, and then an appropriate amplifier chosen to complement them.

In response to the other suggestions about starting with sources, though, I'd like to point out one other advantage of my approach. An average system has, usually, just one end-of-chain subsystem (speakers, amplifier, speaker cable) and one middle-of-chain subsystem (amplifier, preamplifier, interconnects), but may have more than one source subsystem present (DAC, transport, interconnect) (phono preamp, cartridge, turntable, interconnect) (tuners) (A/V processors) (analog recorders) (digital recorders). Since an audiophile will basically always be listening through the same end-of-chain/middle-of-chain subsystems, it makes more sense to choose one's source components through auditioning that will account for the sonic traits of the rest of the system, rather than the other way around. If one began with the source subsystems individually, by the time the chain was completely upgraded to the end, the listener might well find that system synergy between the source subsystems chosen and the middle- and end-of-chain choices that followed was somewhat hit-or-miss. In other words, this ordering could promote good matching between, say, the CD subsystem and the rest of the system, but fail to accomplish the same with the analog vinyl subsystem - both source subsystems haven been already chosen prior to getting around to, say, the speakers. By starting at the end and working back, one is much more in control when it comes to selecting only gear that will ultimately create an equally synergistic balance between all subsystems present throughout the chain when completed.

We'll argee to disagree, Kt88, although I certainly acknowledge the possibility that you built a satisfying system in your own way - mine is not the only way (and mine is also definitely not the best system out there, and never will be!), but I do think it's the most logical and predictable way. Listen on, brother!

For the record in this thread, Magnepans are not electrostatic speakers, nor are they ribbons, in the case of the 1.6's. They are planar-magnetic speakers, combined with ribbon tweeters only in the models with the "R" suffix. The confusion tends to arise because all of these types are frequently (but not always) configured as boxless, dipole radiating, panel-type speakers, but they do employ different operating principles. (The Maggie models with the "QR" suffix are designated "Quasi-Ribbon" for the tweeter, which is roughly speaking like a cross between a planar-magnetic driver and a true ribbon driver.) The following is my admittedly non-expert attempt to explain these differences (corrections welcomed where needed).

In a true ribbon, the best-known examples of which (aside from the Maggie "R" tweeter) are probably the late, lamented Apogee's, the driven membrane is a metallicized film that is itself the conductive element of the driver, much as the voice-coil functions within a regular dynamic driver. In a planar-magnetic design, such as the mid and bass panels of all Maggies, a thin-film membrane is embedded with a wire running through it that functions as the conductive element. Both of these design types employ fixed magnets to drive the membrane when current is applied to the conductor, again much as in a dynamic driver. In an electrostatic design, like a Quad ESL, Martin-Logan, Sound-Lab, or InnerSound, no magnets are used. Instead, a thin-film membrane is suspended between two perforated (for sound propagation) conductive metal elements called stators. The speaker must be plugged into the wall AC, to supply power that is used to impart the membrane with an electrostatic charge when operating, which does not vary. The stators then carry opposite phases of the variable alternating music signal from the amplifier (after being converted from current to voltage by an integral speaker input transformer), which causes the stators to alternately attract and repel the statically charged membrane, producing the diver's motion.

For reasons that I don't fully understand, panel drivers of the planar-magnetic and ribbon varieties are typically not run full-range, but crossed over using separate, frequency range-optimized individual drivers to handle the complete spectrum, while electrostats are in principle single-driver, crossover-less, full-range transducers, but in actuality are often crossed over to a dynamic-cone bass-range driver for practical reasons having to do with maintaining a manageable panel size. Also, ribbon tweeters have often been employed to handle the treble range in box-type, otherwise dynamic speaker designs, such as many historical Infinity models that utilized their EMIT ribbon tweeters, or newer models from Piega or Red Rose, for example.