Introduction
This article compares passive loudspeaker systems which use a single
power amplifier to drive an entire loudspeaker array through passive
crossover components (capacitors, inductors, and resistors mostly)
between the amplifier and actual drivers, and active loudspeaker
systems, where the crossover components are placed between the
preamplifier (processor) and power amplifiers, so that multiple
amplifiers each reproduce a limited range specifically for a single
driver type, such as just the tweeter(s) or woofer(s). In most cases,
the active loudspeaker has the power amplifier built into the enclosure.
You plug the line-level output of your preamp/processor into the RCA
jack on the back of the active loudspeaker box.
To start with a sweeping generalization, I must say this . . .
In general, all other considerations being as equal as possible,
assuming equal competency in all areas of design, manufacturing, and
value, when it comes to the upper end of the loudspeaker market where
budgets allow more extensive design and manufacturing, if comparing the
hypothetical technical aspects between active and passive loudspeaker
systems,
ACTIVE IS BETTER!!!
I can feel JJ, our Editor-In-Chief, saint of SETs and Behemoth
FET-based amps alike wriggling in a tense grimace, his fingers itching
at the editing keyboard, trying to resist the temptation to alter that
statement.
Well, I can't blame him, or any of those who are immediately taking
exception to that statement. Exception should be taken, because like any
other generalization, while it may be true as a generalization, there
are TONS of exceptions. Even so, that bold, underlined statement is one
heck of a gauntlet to throw down among audiophiles and similar home
theater enthusiasts, most of whom have passive loudspeaker systems into
which they've sunk a fat piggy bank, and are probably very content
having done so. Good for them. By all means, stay happy.
I would never hope to imply that any given active loudspeaker most
likely outperforms any given passive loudspeaker. The most significant
aspect of any loudspeaker is the sum of all components, from raw parts
quality and manufacturing care, to perhaps most importantly the skill of
integration, regardless of specific techniques utilized, be it an order
of crossover slope, the use of particular materials over another, ported
vs. sealed, or even active versus passive design. Get the best stuff on
earth, put it together haphazardly with any methodology you choose, and
you've got a mess. Flat frequency and power response, even dispersion,
low distortion, good dynamic range and responsible transient response
can all be achieved through passive design just as they can be screwed
up in an active one.
However, if the active approach is taken, the most critical 'match'
in the entire path of the audio system (the
amplifier/crossover/loudspeaker chain) can be approached as a whole, far
more efficiently, each component working in its most optimal scenario,
yielding potentially superior results to a more "conventional" passive
approach.
Why? I'm glad even if you didn't ask, because this is currently
a monologue.
Efficiency and Output...
Instead of inserting lossy passive crossover components to filter
frequency ranges and flatten driver response by altering the signal
after the amplifier, an active system performs the same task prior to
amplification. The result is a more efficient system, where everything
put out by the power amplifier goes directly to the transducer,
unhindered by resistors, capacitors, or inductors.
Keep in mind, though, that while the loudspeaker system itself is
more efficient, if you try to express the efficiency and/or output
capability difference in a fixed ratio, you'll fail. It gets a little
tricky because the efficiency differences, and subsequently output
differences, are dependent upon the efficiency of each driver in the
passive design, the nature of compensation in the crossover, the
individual drivers' response curves, the signal spectrum, the power
supply configuration for the active speaker, the nature of the crossover
division in the active system, plus a bunch of things that have skipped
my mind right now.
If we were to make a blind, blanket statement about active vs.
passive system efficiency differences, we could say that active systems
are roughly twice as efficient with their amplifier power than passive
ones. If you were to take the utmost extreme (and dishonest) scenario,
you could set up a test case that would show an active system as four
times as effective with a given amplifier rating as a passive
loudspeaker system, based entirely on how loud you could play that
particular signal before the amplifiers clipped. It's not really an
accurate example outside of the lab, but some of that example would be
applicable as a matter of illustration.
Independent Voltage Swings Over Multiple Bands
Consider a tone at 20 Hz, 20 volts RMS. Consider another tone at 200
Hz, 20 volts RMS. Let's consider a load that's 8 ohms, resistive to keep
it simple. 20 volts RMS is 50 watts across 8 ohms.
If we want to put either of those signals into a system with a
passive crossover with a crossover point between those two tones
(anywhere really, it doesn't matter in passive systems as far as the
amplifiers are concerned, but to keep it apples to apples, say 80 Hz,
higher order for minimal overlap between drivers), we would need an
output capability of 20 volts RMS, or in other words, something that
could put out 50 watts continuously into 8 ohms.
In a bi-amplified active system, we would need two amplifiers, both
equally powerful, one to run below 80 Hz for the Low Frequency section,
and one to run above 80 Hz for the Higher Frequency section. To
accommodate either of those tones, even one at a time, both must be
capable of swinging voltage (and providing current) to put out 50 watts
RMS into 8 ohms.
If we look at these tones, one at a time, both systems are equally
efficient, but since the active system has an amplifier that isn't doing
something in both scenarios, the passive system is more cost effective.
The passive system requires a total of 50 watts of amplification
hardware, the active system a total 100 watts of amplification
hardware.
If, however, we try to play both tones at one time, the performance
equality disappears, and the value of the active system skyrockets. Even
if the passive crossover only engages in simple filtering with no loss
in the crossover circuit, the passive system will suffer through this
comparison. Considering that we must apply 20 volts at two different
frequencies, because one tone must ride on top of the other at the
output of the single amplifier with the passive system, when the peaks
of each tone line up, that amplifier must output peak voltages twice
that of either single tone, so that the peaks are equivalent to the
peaks required by 40 volts RMS, or 200 watts @ 8 ohms, even if the RMS
(Room Mean Squared, or the mathematical way of saying average) value is
lower and the power output is half that. It's not really putting out 200
watts RMS, just 100 watts RMS. But, the nature of the complex signal
requires extreme voltage swings at times, requiring very high power for
short durations, so that if the amplifier can't source voltage swings
like an amplifier that can source 200 watts RMS, it will most likely
clip and distort.
The active system, on the other hand, will behave just the same as
when it reproduces either single tone by itself, with two amps running
just as if they were running alone. Not really more efficient, in terms
of power drawn for power output acoustically, but able to do more work
with less hardware. When you consider two 50 watt amplifiers next to one
200 watt amplifier, the economics of active systems become more
attractive.
Coincidentally, audio signals usually contain multiple tones that
straddle loudspeaker crossover frequencies. The latter example is the
most extreme possible, but makes a very valid point.
Still, it's not really accurate to add the power output from each
amplifier section in an active system and say you've got at X watts.
This can be illustrated by looking at another extreme, energy focused
instantaneously in a limited range. After all, even through the strength
of the system may be the sum of all parts, the downfall may be that of
the weakest link, i.e., a passive system operating with a single
amplifier capable of 400 watts can put that power at any distribution
within its bandwidth. An active system with four completely separate 100
watt amplifiers can only output 100 watts in a given frequency range at
a time, with a 400 watts maximum total. If we were comparing a single
400-watt amp driving a passive system with four 100-watt amps in an
active system running over separate bands, if the active system needed
to reproduce a signal that only covered one band, all work falls on a
single 100-watt amplifier. In this scenario with a narrow band signal
appropriate for only one driver section, the active system with "400
total watts" would be no better off in terms of output than a single 100
watt amp in passive system.
The only exception where active systems win hands down under any
circumstances in terms of power allocation is when the system's
amplifiers share a single power supply, so that each amplifier draws
what it needs dynamically. In such an instance, the active system has a
total amplifier output capacity which can be assigned just as
effectively as a single amplifier running by itself, in which case an
active system's multiple amplifiers trounce the passive system's single
amplifier when it comes to output capability of signals that spread at
all across more than a single driver or driver complement.
Summing up the respective output advantages in these examples, they
depend on signal distribution. When the signal is spread across bands of
driver operation fairly evenly, all active systems have an inordinate
advantage. When the signal is narrow in frequency content, the passive
systems allocate amplifier resources more effectively than an active
system with completely separate amplifiers for each driver
section. If the active system has a single common power supply
feeding all the amplifier sections for the loudspeaker, it has all the
potential advantages without any potential disadvantages of active
signal dividing.
Some might be catching on that the scenarios of a perfectly divided
workload, where active systems have an extreme advantage, and that of an
extremely singular signal, where a single amplifier driving a passive
speaker becomes more cost effective, are implausible during normal use.
What we get with music or soundtracks are fickle variations that jump
between the extremes, in other words, lots of various frequencies rather
than just one or two.
So, as you can see, while the basic active system will most likely
hold a substantial advantage with anything with more than a few
harmonics, when it comes to output potential, it's not a simple
comparison, and when looking at the previous examples, not truly a
matter of efficiency, but of signal-dependant output limits.
"Lossless" Crossovers...
Where real issues of efficiency come in is the nature of passive
crossover design. While there are resistive losses in passive crossover
components, these are relatively minor.
"Lossless" Response Tailoring
What is more relevant is that each driver's raw frequency response is
rarely what a particular designer has in mind for the particular task,
unless the designer has the luxury of actually designing their own
drivers. Even with in-house driver engineering, further tailoring may be
desirable just to make it a better product. Tailoring networks that
flatten or curve a driver's response need not involve "extra" crossover
components, but may be as simple as altering the crossover curves of the
"ideal" filter network to incorporate the driver response and
compensate. Response tailoring, though beneficial to sound quality, must
selectively attenuate frequencies beyond the simple filter function, so
that a given amount of power in at some frequencies results in less
total acoustic output than without the response tailoring.
Whether the losses are inserted in the passive crossover, or in the
driver design through additional electrical or mechanical damping, the
effect is the same, namely an additional loss of acoustic output from
the driver. In contrast, an active system performs crossover filtering
and response tailoring prior to amplification, so that the power
amplifier couples directly to a driver that works as efficiently as it
can, without the detriment of insertion losses or less efficient driver
design. Maximum efficiency is reached for each driver. In essence, the active system
can achieve flat frequency response without imposing losses of power to
achieve that flat response, allowing the driver to work as efficiently
as possible.
"Lossless" Driver Level Matching
Most importantly for the sake of discussing efficiency losses in a
passive system, is that like the active system, the passive system must
integrate all drivers to operate as a single loudspeaker, not only in
the gradual transition between drivers, but to behave as one in relative
output for a given input level. This means that output level from each
driver must be matched to the output of every other driver across the
board. In an active system, we can simply alter the gain of the input on
each amplifier to compensate for sensitivity differences between
drivers. In a passive system, we have to lower the output of all but the
least efficient driver by sticking in resistor networks to soak up
power. In other words, you make the most efficient drivers less
efficient so that they match the sensitivity of the least sensitive
(roughly least efficient) driver. In even more words, in terms of
voltage sensitivity, all drivers must cater to the lowest common
denominator. For any speaker that is trying to reproduce bass in an
enclosure short of a Volkswagen, this usually means that all drivers
operate at the approximate efficiency level of the typically inefficient
woofer by receiving signals through resistor networks that intentionally
turn the speaker level audio signal into heat. This typically means that
a tweeter or midrange operates 3-6 dB less efficiently than it
could.
One of the beautiful aspects of the "hybrid" speakers with "powered"
woofers is that even though the speaker requires a "full-range" signal
from the power amplifier to pass signal to the woofers, they still
inadvertently reap efficiency benefits. While such a setup isn't as
elegant as an active division of labor through a subwoofer/satellite
system (all speakers set to "small" in the receiver/surround processor
with bass redirected to the subwoofer), the midrange and tweeter can run
with the lowest common denominator between themselves, far more
efficiently, because the woofer operates from a completely different
power amplifier. As a consequence, while a hybrid system with "powered"
woofers still requires the power amplifier feeding it to output the
entire signal, the speaker looks more efficient to the amplifier, asking
for less power at a given SPL. Though I think of built-in powered
woofers as a somewhat half-baked solution, I have to admit that there
are advantages over a conventional, entirely passive "full-range"
loudspeaker.
Are the Output Advantages Relevant?
To be fair, any efficiency advantage of an active design can be
countered easily with a bigger amp behind a passive design. While a
passive design fed by a mongo amplifier may be "wasting" power, most
audiophiles or AV enthusiasts don't really care about efficiency so long
as the dynamic range is maintained. If that requires a hulking outboard
amplifier fed by a dedicated 20 amp line, and the spouse and budget
concede, most consumers don't mind another bigger box, or multiple
bigger boxes. If anything, there seems to be some kind of bragging
rights that come with the extensive use of resources, particularly among
the "high-end" connoisseurs. I wanted to address the efficiency/output
argument because it has been raised in the past, but I personally
consider it a dead issue for most, beyond the scope of a technical
discussion. In either a passive or active loudspeaker scenario, the
available amplifier power should remain higher than the required
amplifier power. If that's not the case, we need to reevaluate the
entire system, or maybe our listening habits.
More substantially on the topic of comparison, active systems tend to
have . . .
More Stable Crossovers
Passive crossover components rely on the complex reactive (and
changing) impedance of drivers, in addition to the impedance of each
other, to perform their filtering. As a consequence, passive crossovers
can be far more difficult to implement precisely, and more
limiting in the filter's precision, than their active counterparts.
Unlike passive crossovers, active crossover components are completely
isolated from the fickle impedances of loudspeaker drivers, "buffered"
not only by the power amplifier, but in active active (vs. passive
filter "active") crossovers, by their own amplifiers. Designers of
passive systems can apply impedance compensation components to counter
the reactive properties of the drivers themselves, should they wish to
incur the cost, making the filter design easier, but they're still not
out of the woods, even with today's fantastic computer modeling. Huh?
Consider.
At high output levels, drivers heat up. The hottest part is where the
heat begins, at the voice coil. As the voice coil heats up, the coil
resistance increases, and maybe the suspension compliance changes (they
can temporarily become looser, which in turn affects the resonant
frequency and damping.) The point is that the passive components are no
longer dealing with the same scenario they started with, and the driver
impedance changes. The filtering characteristics of the passive
crossover begin to shift, and so by definition either work less
optimally at low output levels, or work less optimally at high output
levels. The same applies to the loading of bass-reflex systems, which is
why good sound reinforcement gear is designed so that it doesn't
actually sound right until it's obnoxiously loud, but we're getting off
track.
Active systems, because they're buffered from the changing impedance
of the drivers, are completely immune to this behavior. It is true that
passive systems with high power handling through and through will have
less of a problem with this, as the coils heat up less to begin with,
but that's nowhere near as nice as the complete immunity offered by the
active system.
But wait, there's more, such as . . .
Loudspeaker/Amplifier Matching
There is one more point on behalf of active systems that I'd like to
make. The active loudspeaker system has a practical performance
advantage that one should point out to those who could consider multiple
dedicated amplifiers an extravagance or even hindrance when one would
do, particularly if that one could be more expensive per channel. This
point is the relationship of the loudspeaker to the amplifier, and it's
got nothing to do with passive crossover components. When a
manufacturer builds a high-performance outboard amplifier, the
amplifier MUST accommodate a wide range of loudspeaker impedances, some
of which may be very unusual. In order to do this, the manufacturer
needs not only good design, but to massively overbuild for extreme
variances in loudspeaker loads. This usually translates into big, heavy,
and expensive. Unfortunately, this overbuilding usually yields
benefits not in proportion to the effort.
For example, my own big, outboard, muscled-up, stereo amp of choice
and budget can source 20 amps continuously into each channel, or in
other words, cruise into 2 ohms at 800 watts, twice, for 1600 watts of
continuous output, assuming my wall socket can hang in there. If we tie
that kind of output with premium sound quality, that ability costs mucho
dollars.
My particular amp is from a no-frills manufacturer, so it's still
relatively cost-effective for what it is, but not so cheap as I'd like
so that I could buy a few more. The extreme low impedance output
capability is nice to talk about, and handy when in use as a reference
amp so that it's ready to drive any review sample that comes in the
door, but until I actually use a loudspeaker that gets to 2 ohms over
any substantial part of the higher energy portion of the audio spectrum,
i.e., mid-bass, low bass, maybe midrange, I can't get the full output
potential out of that amplifier in terms of milking the power supply for
everything it's got, even if I can drive it into clipping distortion by
exceeding it's substantial voltage swing capacity. It was 'optimized'
for the most strenuous current delivery scenarios possible.
If the designer of my particular beast of an amp had a specific
loudspeaker in mind, he could have designed the amplifier to deliver
that full 800 watt capability for real, without worrying whether
something stranger might happen along. However, what I'm actually
getting with most speakers is an amplifier capable of an easy and stable
400-500 watts/channel, because that's what the impedance of most
loudspeakers allows the voltage swing to push through. I may be better
off than an amp that was 'optimized' for power delivery of 400-500 watts
into that load since the power supply remains more stable when it never
goes beyond 60% of its real capacity, but that's only because it
can't.
While I consider it better to err on the side of clipping the voltage
rail a little sooner (only a problem for the duration of clipping) by
under-winding the secondary of the power transformer for greater current
capability vs. exhausting the power reserves (a problem until the
transformer can refill the capacitors) by over-winding the secondary of
the transformer for excessive voltage capability and impressive "peak"
output levels, a "perfect" match is more ideal than either
situation.
In real world scenarios, outboard amplifiers themselves are often an
inefficient allocation of manufacturing resources, and therefore money,
when it comes to cost/performance gains. Surely we're better off than
using an under-powered receiver, but by using massively overbuilt
products, we often use and pay for more than we need for a given
application, and get less from what we put in than if the same amplifier
were designed as a dedicated unit for a particular loudspeaker, be it
passive or active. I'm not saying that big outboard amplifiers don't
reap substantial rewards compared to using the onboard amplifiers found
in typical receivers. In many cases, particularly at higher output
levels, the extra juice pays off big time, not only in sheer volume, but
clarity.
What I'm trying to say is that even if the active system uses smaller
power supplies and less output devices, etc., because amplifiers in a
fully active system are truly part of a known whole, they
can offer the same raw performance as big, huge outboard amplifiers when
used in that specific scenario for which they were
intended, or for that matter, even possibly perform better than the
'bigger' outboard counterpart. Does that mean that a dedicated amplifier
in an active system necessarily performs better than another amplifier
with its own chassis? Of course not. If there ever were a universal
truth in audio, it would be, "It depends."
Not Necessarily Sunny
Most of this has been pretty one-sided, in favor of the active
systems. However, there are some myths about the superiority of active
loudspeakers, as well as some potential downfalls.
Myth- Active Loudspeaker Systems Have Better Damping of the
Woofer, due to a direct electrical connection between the
amplifier and the driver, as opposed to Passive Loudspeaker Systems,
which require an inductor or two in series, adding electrical
resistance.
Well, sort of, but it's misleading to imply that an active system
will necessarily have tighter bass because of "better" damping from the
power amplifier.
First, more damping is not necessarily better. Rather, there is an
amount of total damping in the system that is considered the target area
to maintain flat frequency response, good transient response, and
healthy output levels at the lowest range of the woofer. Too much
damping, and the frequency response will prematurely decline, leaving
the sound thin and anemic. Too little damping causes a peak right before
the low frequency limit, causing the bass to sound fat, "slow", and
muddy. In sealed systems, the "ideal" target range of the system "Q", a
number that describes a system's resonance, and inversely damping, is
usually considered between 0.5 (critically damped for fastest settling
time) and 1.0 (most extension for a particular box size.) In that range,
the optimal target depends on the design goals. With ported or passive
radiator bass-reflex systems, there are multiple resonant devices that
are all combined, so it gets more complicated, but the basic principles
are the same.
Placing a passive component, such as an inductor, between the
amplifier and the driver doesn't mean worse damping, and so far as the
inductor is linear, requiring that it's air-core versus an inductor that
uses a metal slug inside it to increase inductance for a given loop,
doesn't create more distortion. It simply means less electrical damping
by the power amplifier, which can easily be compensated for by the
designer via more mechanical damping of the driver suspension, or within
the enclosure. Less electrical damping itself may even be beneficial in
particular contexts if it serves the total design.
What is relevant is that the damping in a passive system creates
another variable, which is not only by definition variable (the
resistance of the coil can change a little with current) but provides a
factor by which the variation of a more variable factor (the resistance
of the voice coil) will translate into more dramatic changes in system
damping. Is it a big deal? If the inductor has a reasonably low
resistive component, probably not a whole lot. But, if the inductor is
of the air-core variety, the low frequency nature of the crossover for a
woofer will require the inductors to be quite large, and because of the
length of wire needed, even if it's of relatively heavy wire gage, the
resistance will be more significant, and the inductor will be more
expensive. Big deal? I don't know. As I said, probably not a whole
lot.
However, few manufacturers use air-core inductors exclusively,
particularly with woofers, due to the aforementioned size and cost
required, in which case you're now dealing with the non-linearity of an
inductor of an iron-core or similar variety, which can become
detrimental beyond the extent of changing system damping, but can
saturate, and dynamically change their inductance value, screwing with
the crossover in a transient manner.
Consider that the non-linearity of metal-core inductors is worse at
high current levels, that most woofer's impedances are lowest right
above and below their low-frequency limits and so draw more current at
those frequency ranges, and that the spectral distribution of energy
peaks are often heavily weighted in the mid-bass on down, and you've got
a very difficult situation. You can minimize this problem with higher
capacity metal-core inductors, though it starts becoming an increasingly
expensive band-aid, as they get heavier and larger, which only minimizes
the problem instead of solving it.
This is starting to sound like another benefit of active design.
Well, it is, and it's most substantial in the ranges prime for woofer
crossover frequencies. However, we don't have to go completely active to
completely avoid this pitfall of saturating inductors, as we can use a
typical subwoofer/satellite combination, where the subwoofer/satellite
filter uses an active crossover (provided with every unit that offers
bass management,) and the higher crossover frequencies used with
mid-bass/mid-range drivers and tweeters make air-core inductors
extremely plausible and cost-effective. Vance Dickason, author of the
wonderful "Loudspeaker Design Cookbook," goes into inductor
saturation much more thoroughly, and actually recommends this type of
"hybrid" approach, as an easy method of side-stepping inductor
non-linearity in DIY loudspeaker design.
Hmm . . . . I think I remember at least one manufacturer out there
advocating just that approach before receivers even had bass management,
so much that they provided the line-level low-cut filters to make it
happen. HP-80 anyone?
Despite the seemingly overwhelming advantages of a fully active
system, there is a serious potential pitfall . . .
Whatever You've Got, You're Stuck With It.
Even if the manufacturer can either design or OEM active electronics
that more or less work, with an active loudspeaker system that uses
dedicated, on-board electronics, from input to the electronic crossover
to the output stages of the power amplifier feeding the driver(s), be it
a fully active system, or a hybrid job with "powered" woofers, there is
no upgrade path. Not enough power, low parts quality, or just poor
circuit design, what's there is all we'll ever have. Sorry, go to jail,
and don't pass go.
Stacey reminded me that there is an exception when the manufacturer
offers an upgrade to a newer revision, but that's their choice, not
yours.
While I think that the typical audiophile or AV enthusiasts takes
mixing and matching, from speakers and amplifiers to cables and
shoelaces to an extreme, the convenience and advantage of having a
completely integrated package that comes with a self-contained active
system can be a hindrance, in two scenarios.
The first example of where the complete system approach of a fully
active loudspeaker system can be a disservice is where the manufacturer
may have gotten some parts right, but fell short in other areas. In the
audio industry, it's unfortunately common.
The typical example is a loudspeaker designer who really doesn't have
a clue about amplifier design, or simply underestimates the value of
putting resources into high quality electronics, and as a result ends up
with a product that performs more poorly than passive equivalents mated
to a higher quality front end.
Some "high-end" loudspeaker manufacturers, particularly the small
boutique companies, don't have engineering resources much beyond a few
texts documenting generic projects, a few guys handy with a soldering
iron when they're not engaged in terrific carpentry or lacquering work,
and a lot of time to listen. These people, though perhaps fantastic
"artists" when it comes to designing and customizing the sonic
colorations of loudspeakers through trial and error of swapping passive
crossover components, aren't comfortable or even competent with
electronics design. If you screw up a passive loudspeaker, it sounds
screwy. If you screw up an active one, things could blow up. These
manufacturers should absolutely stay out of the active market.
Just like other audio components, fully active systems aren't created
equal. Way back when I tagged along with my friend Andor at an AES show,
where active monitors aren't such a rarity, although the Genelec
demonstration did well in the dynamic race on the showroom floor,
nothing I ran across at the time (KRK, Tannoy, Alesis, etc.) convinced
me that active loudspeakers were appropriate for high fidelity
reproduction. They seemed more just like a vehicle to get small speakers
to play really loud with lots of bass. In fact, with some of the blatant
response peaks so obvious even with passing listening, it made me wonder
if they actually made their speakers less accurate so as to make them
sound more superficially detailed. That's not to say that the real
quality stuff wasn't there, but I just didn't find it. It was also
a few years ago, and maybe all of the problems I perceived were acoustic
issues, so all of those companies may actually offer truly good products
now.
This predicament of lop-sided system design need not be a substantial
problem in practice IF the manufacturer takes care to address ALL
aspects of the product, from electrical signal input to acoustic output,
with equal attention. Even for the consumer who'll always want "better"
electronics than the "standard" design, if the original electronics are
really good, but he or she wants really, really good, the consumer is
still better off with really good electronics in an active system than
really, really good electronics in a passive scenario. If the product is
really good due to careful deliberate design from the ground up, just
leave it alone. If a person really needs to fiddle, then the problem is
of the second variety.
Limited Tinkering Capacity
The second, and more relevant disadvantage of active loudspeaker
systems has nothing to do with objective performance advantages or
disadvantages, but rather the need of many consumers to twiddle with
their systems to make it more theirs, more to taste, to wiz on it, so to
speak. A fully active system takes away choices. If we want to soften
the sound with a more mellow, warmer amp, or crisp it up with a slightly
spicier alternative, we're out of luck. If we want to put some fat,
fancy, after market speaker wires to roll off the top end for extra
"refinement," too bad dude. From the hobbyist perspective, a
comprehensively designed, all-inclusive system is a nightmare. If you
sweat about after market power cords, "premium" digital cables, or
Mping-pong discs more than the furniture and acoustics of your listening
room, a completely integrated system probably isn't your bucket of
tea.
Most high-end enthusiasts crave gizmos, more boxes, and a long chain
of purist components, rationalizing performance advantages through
specialization and isolation. For those who can remember a few years
back, consider the fancy CD transport/premium digital cable/fancy
anti-jitter device/premium digital cable/fancy upsampling device/premium
digital cable/fancy outboard DAC/premium analog interconnects/fancy
analog preamp combination that was touted as the ultimate in CD playback
to precede the power amplifier.
In truth, a simple CD player with a volume control is the most ideal
for such a task, and anyone who had a clue about how a CD player
actually worked knew it. ALL CD players incorporate a FIFO buffer
following reading the transport, as required by the jumbling done by
Sony's error correction on the disc. In fact, a single box CD player
slaves the servo of the transport motor to pick up data at the speed
appropriate for the master clock that feeds the onboard DAC
conversion, making jitter levels dependent ONLY on the quality of the
master clock, without any need for jitter reduction. Actually, single
box CD players have an architecture with lower inherent jitter than any
multi-box solution.
While much fuss was made recently about "upsampling" and sample rate
conversion accessories so that we could feed either a 88.2 kHz sample
rate or 96 kHz sample rate to our new outboard DAC units with a full 24
bits from the "inadequate" 16 bit information sampled at 44.1 kHz
gleaned from the CD, the truth is that every modern DAC, including the
cheapest units sold in CD players ten years ago, "upsampled" to a
minimum of 176.4 kHz, and usually 352.8 kHz sample rates, with more than
16 bits of data, not to increase resolution, because such is
mathematically impossible, but to apply the digital reconstruction
filter. Back then they called it "oversampling." Let's not even mention
the shorter signal path possible within a single chassis, and just
consider how much those premium digital cables and extra boxes added to
the end price while degrading the ultimate performance potential.
I'm sure that somebody out there is going to hold up a truly poor
single-box CD player as proof of the opposite, and to that, I've only
got a few numbers to counter it, 508.24.
Summary
I don't want to put the vibe out there that people really need an
active loudspeaker system, or that passive loudspeakers are substandard,
because there are so many truly terrific passive alternatives.
What I do want to emphasize is that if you don't at least seriously
consider an active alternative when it comes your way, and follow the
typical bias of conventional "wisdom," you might be doing yourself an
incredible disservice. While it would be nice to win over the market and
the industry to buy and produce more active loudspeaker products, as I
do consider it the most sensible solution to high-performance design
dilemmas, I must be realistic. Most truly great loudspeaker designers
I've talked to know the benefits of active systems and refrain from
supplying such products. Why? The bulk of the high-end consumer market
isn't ready. But, when it gets ready, manufacturers will begin to
supply more, and better options. I'm just asking consumers to be ready
by the time they get their next itch, because it is likely to be very
good stuff.
I just wanted to share. Thank you.
- Colin Miller -