Soundstage!- The Entry Level

Cheap Maggies

October 1996

A ribbon tweeter and planar-magnetic woofer, all from Magnepan for an even $500? Yes, it's true, the Magnepan MMG offers their trademark sound at an unheard of low price. They're odd, they're quirky, and everyone who visits you will wonder just what they are, but for the right budget minded audiophile they may be just perfect. This month in The Entry Level, you get some background on just why there are so many speaker designs out there, and what sets the Magneplanar line apart from the rest.

It's misguided to feel that any audio product can be described as good or bad without some notion of what you're looking for. If you're looking to buy, say, an amplifier, one that is 100w/ch may be just as "good" for you as one that is 200w/ch, if the extra power isn't needed anyway. Nowhere are the tradeoffs in audio design more apparent then when looking at speakers. There are a ton of different ways to put a speaker together. You can use cone drivers, or ribbon drivers, or any number of other types. You can use a sealed box, or a ported/vented one (those two terms mean the same thing in this context), or maybe even no box at all. If any of these methods were clearly superior to the others, nobody would even talk about the other alternatives; as it is, the distinction between every one of these choices are of the type where "it depends". The classic distinction that some make is that particular speakers are, by their design, particularly suited to classical or rock music. That's a bad distinction to be making, but there's a nugget of truth behind it.

Let's look at some of the major parameters that are affected by the big design decisions. The first thing is efficiency (which also gets called sensitivity). This is a measurement of just what volume a driver or speaker will produce with a particular input power. Usually, this is given as a volume with 1W of input power from 1M away. A typical speaker aimed at the consumer market might be rated at 89dB, 1W/1m. This means that if you use a 1W input signal and measure at a distance of 1M away, you'll be hearing a 89dB SPL (sound pressure level). Sometimes, instead of 1W, you'll see it measured as 2.83V/1m. If the speaker is 8 ohms, this is the same rating scale, because 2.83V into 8 ohms is 1W. If the speaker presents less than 8 ohms of a load, using 2.83V instead inflates the sensitivity rating compared with what it would be with 1W. This makes the efficiency look higher.

Turning efficiency figures into volume levels is a fairly simple process. In order to increase the volume by 10dB, you need 10x as much power. So given our 89dB 1W/1m speaker, if you used a 10W signal instead, the output level at 1m would be 99dB. A 100W signal would produce 109dB. Now, most people listen to their systems further than 1m away from the speakers. For typical listening distances of 6-10', I mentally subtract between 3 and 6dB from the figure given for an estimate of in-room volume. Going back to our fictitious example again, I would expect that, if it were capable of cleanly handling such power, we could expect between 103-105dB in-room with a real 100w/ch amplifier. This all gets more complicated with different speaker loads; for example, while 4 ohm speakers usually end up requiring more power to achieve the same volume as their 8 ohm brethren, most good amplifiers are capable of more output into 4 ohm loads (a very good amplifier will have fully twice as much power with every halving of impedance). There are some other ugly variations to the efficiency figure as it related to load impedance, but as a general simplification you can do some math like that I presented here to get an idea how loud a particular speaker/amplifier combination will play.

The next figure of merit is bass extension. The usual thing you'll see is a figure speaker literature refers to as the f3 point, which usually shows up looking like -3dB@40hz. What this means that, relative to the normal output, bass notes at 40Hz will be playing 3dB lower. The workings of the decibel scale say that 3dB is a factor of 2, so -3dB means that the speaker is playing half as loud at that frequency as it is normally. The thing you usually aren't told is the roll-off speed. The f3 figure shows how much the output has dropped at one point, but it doesn't tell you how fast it drops beyond that point. Roll-off is measured in dB/octave. A typical sealed box rolls off at 12dB/octave. Given our speaker with f3=40hz, a 12db/octave roll-off tells us that at 20hz (one octave below 40hz), the speaker would have an output 15dB below its normal output. Ported or vented boxes usually roll off at 24dB/octave. At 20hz our f3=40hz example would have its output 27dB lower than normal if it were ported instead sealed.

Just how far below the normal listening level you can hear is a difficult thing to put a finger on. Some say that "useful" bass is still being generated at the speakers -6dB point, while some state you're still getting something added at the -10dB level. Adding to the confusion is a phenomenon known as room gain; at low frequencies, your room tends to bounce bass around and reinforce the notes, increasing their output above what the speaker itself would typically produce. You can alter the degree of room gain by moving your speakers closer to a corner, but that degrades the sound of the higher frequencies. In any case, these measures let you see why both sealed and ported designs exists. Sealed boxes have reasonable efficiency and a high f3 point, but the roll-off is slow below that point. Ports have higher efficiency, and can produce a lower f3 point, but the roll-off is much faster for bass deeper than that frequency. In fact, as you go significantly below their f3, ported speakers start doing funky, damaging to your woofer antics that sealed boxes never go into. So if you want louder bass that goes reasonably deep, ported is the way to go (passive radiators, a variation on ported design, can produce even higher efficiencies but their roll-off is still a teep 24dB/octave). Similar sealed boxes aren't as loud, but they can go much deeper without problems (especially when aided by room gain). There are some qualitative measures as well; just about everyone agrees that sealed enclosures produce a better quality of bass, using words like "tighter".

This brings us to our third factor. Sound quality isn't something you can measure easily like the efficiency or bass depth, but it's certainly just as important. You can't put a number on tonal balance, imaging, or any of the other qualitative things that distinguish between speakers, even those with similar measurements. There's a certain rightness to a good speaker no one has successfully quantified yet. It's those characteristics that distinguish the by- the-numbers mass market material from that we would call high-end.

Going back to the rock vs. classical distinction, it's these three factors that primarily come into play. The stereotyped figures that rock music fans are typically presented by the high-end as are concerned with efficiency, a bit of bass extension, and don't give a damn about that wussy stuff like soundstage width. Our stereotypical classical fan is looking for detail, imaging, and other such qualitative features. Efficiency is not an issue, and they either don't care about deep bass or want it to go way down for things like those big pipe organ recordings. I won't discuss this insulting bit of classification further, but I think the point that's trying to be made is that you can't have all these things at once. You can get deep bass, or high efficiency, or high quality, but if you want all three you're going to pay big time.

Certain types of designs obviously favor one of our three categories. Horns, for example, are known for having high efficiency at the expense of everything else. Try and push the bass deep, and you'll find the horn becomes too large to be useful. And many point out the ugly flaws of your average horn as far as tonal balance goes. In defense of this category, I will state that my favorite all-around speaker at a reasonable cost is the horn loaded Klipsch Forte II that I spend most of my time listening to, but even I will freely admit that most horn designs are junk. Horns are one end of the design spectrum. In the middle, you'll find your standard speaker with cone or dome drivers. You get a tweeter, maybe a midrange, and a woofer, put 'em in a box (maybe sealed, maybe ported), get a good crossover, and you're off and running. Cones offer reasonable efficiency, reasonable bass extension, and if you do things just right they can do a good job at being realistic. Way over at the other end of the market, you'll find speakers that most would classify as exotic. Ribbons are one example. Because a ribbon is so light, it can reproduce high frequencies with exceptional accuracy. The problem with them is that the same ribbon that sounds so good is invariably very inefficient and limited in bass.

With that background in mind, let's take a look at one interesting exotic design, the MMG (Mini-Mag) from Magnepan. Most audiophiles are familiar with Magnepan and their Magneplanar line. Traditionally, the complicated manufacturing techniques used for their speakers have resulted in a very high cost. Magnepan claims that their success and manufacturing scale, combined with direct sales, lets them sell this model at a very reasonable price.

Magneplanar MMG

$500/pair. Two-way speaker system with quasi-ribbon tweeter and planar magnetic midrange/woofer, crossed over at 900hz. Response -3dB@50,24Khz, impedance 4 ohms, sensitivity 86dB/2.83V/1m. 14 1/2"x48"x1 1/4", 23 pounds each. Contact Magnepan to order at 1-800-474-1646.

I first saw the ad back last October. A real set of speakers from Magnepan, using most of the same technology that their expensive models have, only smaller. Cheap, too; just my style. They've got an intriguing ad. See, they claim that once music fans become hooked on their speakers, they just keep buying bigger and better ones in the future, ignoring all that cone junk. The problem, then, becomes how to get people addicted to that Maggie sound in the first place. So here's what they're doing: buy a set directly from the factory, shipped UPS to your house, and if you don't like 'em, send 'em back within 60 days and they’ll refund your money. Simple enough, right? I've always wanted to hear a pair of their speakers, but the closest dealer has always been too far away. With this offer, I could try out a set and get a real extended listen without any financial risk. Sounded good to me, so my credit card followed its frequently trodden path out of my wallet and I called the factory up.

You get a couple of choices for wood style and fabric color. After taking care of that, the sales rep did caution me that I needed a good power amp. Check out the specs; that's a nominal 4 ohm load, more difficult than average to drive, and with that low 86dB efficiency you're gonna need some juice. I assured her that I had a number of big beefy amps that would laugh even at a 2 ohm load, but it is a valid concern. Many cheaper amplifiers are not happy at all with anything that drops below an 8 ohm load, and these speakers will not be pleasant for such an amp.

It took about a week for everything to show up. The speakers come packaged together in a tall, slim box. That's one of the strengths of this speaker; they don't take up a lot of room, you can store them very compactly, and they are light enough to move easily. There's a touch of assembly to do before you can use them. The MMGs are designed to be floor standing, and they have a set of "feet" that get attached to the bottom for support. It sure beats setting up stands, that's for sure, but it does take a couple of minutes to thread the screws through. I got to spend that time staring at the back panel, and I noticed my first problem. The binding posts were very non-standard. They use a proprietary system I've never seen before. There is a circular hole designed to have bare wire put in it (you could also use most pins, but forget about spades and bananas). Then you take the Allen wrench they include and use it to lower a sort of screw down to clamp the wire into place. A very strange system, and not one I am at all fond of. It took me a while to find some bare wire I could fit into it, as I am normally a spade and banana wielding cable guy. And I'd hate to think what kind of trouble you'd run into if you lost the slim post tightening tool; I keep mine in a big sealed plastic bag so I can't misplace it. Ultimately, I ended up buying an adapter the company sells for $35 that plugs into their odd connector and has screw posts on the end. This is OK for spades. It's still not great, because now I need a flat screwdriver, and I'm still out of luck for my banana terminated cables. I rate their binding posts a major pain. The manual makes it obvious that it's designed for big gauge bare wire. I gave up bare wire a long time ago because I was getting sick of the ends oxidizing in nasty ways, and I have no intention of returning to those unpleasant cables.

Speaking of the manual, it's decent, but no prize winner. You get a stapled together photocopy of five pages of instructions they printed out from some computer there, and it all looks slightly less professional that the stuff that comes out of my printer. But, hey, the speakers are only $500, so I can't complain too much that the manual looks cheap. You get the usual warnings about keeping the polarity correct, yet another warning about blowing up cheap amplifiers, and some decent suggestions for placement and room acoustics. No recommendations for break-in, but they do talk about some of the options you have, like tweeter attenuation (resistors are included to drop the tweeter output) and fuses. I found the suggestions for room placement to be very accurate, i.e. put the speakers 2-3 feet from the rear wall as a starting point. I didn't notice any placement irregularities that weren't covered there.

Enough of the background already, how do they sound?

So I found some bare speaker cable and got 'em hooked up. When I dragged out my usual torture test material and started listening, I was totally stunned. In some ways, these little $500 speakers easily outperform cone speakers costing over twice as much. Normally, I expect inexpensive speakers to give me a good presentation of the music, but these MMGs went way beyond that. There was a level of realism to the sound matching the best I've ever heard at any reasonable cost.

The main thing that struck me was how well they managed to be precise and alive without becoming harsh or irritating. Everything sounds simultaneously smooth and detailed, which is quite a trick to pull off. I was especially impressed with how accurately acoustic sounds were portrayed. Whether it's a guitar or a violin, you can hear the strings resonate more realistically on these speakers than many far more expensive traditional designs. That authentic sound is there for just about every other instrument as well, except for those that dip deep into low bass territory.

Past reading about speakers of this design left me expecting the bass to be wimpy. Wrong! While it certainly doesn't go very deep, with a -3dB point at 50hz, the bass that is there is very tight and powerful. Sure, the extreme bottom end is missing, but they certainly don't sound like a pair of anemic mini-monitors. The sound is full-range and goes as deep as most speakers in this price class.

Magnepan tells you over and over that you'll need a high quality amplifier to deal with the load these speakers represent. They're not kidding. I was glad to notice that these speakers don't seem to be especially sensitive to amplifier interaction. The technical literature suggests that the drivers in this speaker present essentially a resistive load, without all that nasty capacitance and inductance that makes amplifiers react differently. Trying out some amplifiers of similar quality that normally sound very different from each other, I found that the MMGs sounded just about the same with all of them. These are not speakers that are going to give you amplifier matching headaches to get good sound of them. This is a good thing, because if you needed an amplifier with excellent sound and several hundred watts of power, you’re talking serious money. I was quite content using some standard inexpensive solid-state muscle amps, without feeling I needed something more refined to make the speakers sound good.

After hearing some of my favorite songs playing on these Magnepan speakers, it was very difficult to drag myself back to those big cone and horn designs I normally use. When I listen to less expensive models like these, I usually end the session glad to return to my regular system. While they have their flaws, I could be satisfied with a pair of MMGs as my only speakers, and that’s not something I often say about $500 speakers. It's worth pointing out just what those flaws are.

The Bad News

Just because I've raved about them so far doesn't mean these speakers are for everyone. You'll get no free lunch here. Remember before when I was droning on about sensitivity and power requirements? There's a reason I brought that up. The Magnepan MMG has a rated sensitivity of 86dB/2.83V/1m. I said that manufacturers cheat a bit sometimes with these figures, and here's a perfect example. People are used to seeing sensitivity figures rated with 1W of input power, which is 2.83V with 8 ohms. Putting 2.83V into the 4 ohm rating of the MMGs possess requires 2W of power, so the rating is more like 83dB/1W/1m. Since most amplifiers have a higher power rating into 4 ohms, that cancels this out a bit. Still, though, compared with a typical speaker that might be 89dB/1W/1m, you're losing either 3dB or 6dB of power based on how you slice the numbers around.

Let's put this into perspective for you. My Proton D1200 power amp has big, cool power meters on the front that are fairly accurate. I was routinely pumping over 50w/ch into these speakers in routine listening; no way would you want to have less than that and expect to play a set of MMGs at all. When I cranked things up, I could easily drop over 200w into each speaker. At one point, I was using an original Adcom GFA-555 to power the MMGs. For those not familiar with this amp, it's a solid state brute that is rated for 200w/ch of continuous power (even more into this 4 ohm load), and has good headroom to boot. There's two clipping lights on the front of the amp that brighten when you're pushing it too hard. In the five years I've owned this amp, I've seen those lights once, that was during some testing where I purposely beating up on a pair of speakers just to see what the limits where on the amplifier and the speakers were. Hook up these power hungry Magnepans to it, turn the volume up, and those lights start blinking like a Christmas tree. That's right--these speakers easily soak up the entire output of a big Adcom amp and leave you wanting for more. The useful power handling capacity is somewhere around 200w/ch; it was about that point that the planar magnetic woofer started making ugly "quit it already" noises. But I was surprised at how pleasant the clipping noise is on these speakers. Cone drivers normally sound horrible when you start pushing them to their limits, and I personally am very sensitive to that distortion. When I go to parties where the host has some little system they turn up way louder than it was meant to, I always leave with my ears ringing even if the volume wasn't all that loud. I didn't hear any of that bothersome distortion near the output limits out of these speakers. When you push them too hard, you hear an obvious rattling out of the woofer and the frame, but back off a bit and they are just fine. The output sounds very linear up to the clipping point, and I certainly appreciate that on a speaker that I'm more likely to push up against its limits than my usual fare.

In the room itself, I found that even with a few hundred watts of power feeding in, the output level from these speakers peaked around 95dB. For comparison sake, when I was testing the RA Labs Mini-Reference ($224/pair) speakers with a little 6.5" woofer, I was able to get a similar 95dB speaker clipping output in the same room with about 40w/ch. That seems about average; you can buy cone speakers that play louder than a pair of MMGs for half the cost, using 1/5 the amplifier power, but they won't come close in terms of realism or detail. The tradeoffs here are very obvious.

Snob in the box

One of those things I've always heard from those high and mighty audiophile types is that typical box speakers sound boxy. This is one of those comments containing no useful information; just what does that mean, anyway? After spending some time listening to the MMGs, I started to notice what they are talking about. Typical box speakers vibrate in time to the music you play on them. Better speakers vibrate less than cheap ones. Controlling and dampening the box resonances is one of the major factors that separates good speakers from bad. It's tough to accurately describe what the vibration does unless you've heard speakers without that coloration included. The basic effect is that midrange information becomes smeared and distorted by the bass notes played. When you listen to speakers outside of a box with a design like the MMG, things like vocals are more present and believable because the box isn't inserting that unwanted information into the music. Now that I've gotten a firm grip on what "boxy" sounds like, I can't say that it's all that important to me. Generally, any time you have enough bass moving to get the box adding a bunch of midrange garbage, I find that the midrange was already obscured just by virtue of having that bass going. Sure, it's nice to be free of the box sound, but it's not an issue in itself big enough to justify boxless speakers if the other tradeoffs involved become a problem. Strictly opinion here; your mileage may vary.

Get me a real woofer

I'm never satisfied with speakers that don't stretch down into the bottom of 40hz territory. On the rock music I spend most of my time listening to, there are too many bass players that like to hit the big 42hz note that typical bass guitars have their bottom string tuned to. On speakers like the MMG, where the bass is very rolled off by 42hz, I am not nearly as satisfied with the music because the impact of notes in that range is so diminished. When you consider the lack of deep bass, along with the power limits I was running into, it seemed obvious that what I really need to make a system built around the Magnepans move was a subwoofer. Between adding in the deeper bass and relieving the MMGs from some of the strain of the bass content, it would seem an ideal solution.

One of the pieces in my audio component collection is a Proton AP1000 preamp. One thing I find very handy is that has an electronic crossover at the preamp level you can use at 75hz or 150hz. Perfect for typical subwoofer applications, relieving the bass from the main speakers and feeding it to the sub. My first shot at improving the MMGs used that preamp to feed the bass to the ever popular $125 Subwoofer, an inexpensive sub from the early days of my speaker building. I thought this would go fairly well. Wrong! Remember those comments about box coloration? The big subwoofer has it in spades (like I said, it was an early design, before I knew that much about dampening such things). When you add it to the Magnepans, all that midrange distortion comes right back out again, making them sound considerably worse. Obviously a bad match.

When the $125 sub just won't do, I pull out the big ammunition. Right after they were introduced, I bought a set of Hsu Research HRSW10 subwoofers. These are a terrific design, with a pair of enclosures made out of thick cardboard tubing and a 10" woofer at the bottom. The tubular design leads to very low box coloration. After hooking them up, I thought I had it all--ribbon tweeter, planar magnetic midrange, and that nice cone slam on the bottom way down to below 20hz. Relieved of the bottom woofer duties, the Magnepans played a bit more cleanly; not the huge increase in clarity that's the typical result when you remove the bass from a cone speaker, but it was a bit nicer. I was therefore expecting to be able to play them a little bit louder. Wrong again. The normal limit on how loud you can play a pair of MMGs is the woofer excursion. What happens if you remove that limit is that you only get another dB or two before the tweeter fuse blows. I never actually damaged the tweeter, but things certainly didn't sound very good after the fuse went. While using a separate subwoofer does increase potential volume a bit, don't expect that you'll suddenly get a huge increase in dynamic capability. The woofer and tweeter on the MMG run out of stream at around the same volume level.

There was one issue brought to light during this experimentation. While the Proton preamp I was using for crossover duties is a decent performer, it's no high-end component. Typically I'm pretty satisfied with its quality. With the revealing MMG speakers as the ultimate destination, it really showed off its limitations. The general ugliness it adds to the sound was brought into sharp focus, and I was very glad to return to my usual Rotel preamp. This is a concern; while a pair of MMGs isn't very expensive, you're not going to get the quality you should from them unless the associated components are of good quality. Pass the sound through something cheap, and you will hear it on these. I’d recommend using a simple capacitor as the high-pass filter on these speakers instead of the high output from a standard two-way electronic crossover, because that will degrade the sound considerably less.

So what are these good for?

I started off this review talking about how you need to consider what any potential addition to your system is good for. The Magnepan MMG represents an extreme to the sound quality spectrum. Efficiency concerns are thrown out, and what you get is something that has far more of the qualities revered in high-end audio than any similarly priced competitor I've heard.

Let's consider one of those competitors. The B&W DM602 is a speaker I've recommended to many people before. For $550, you get a pair of speakers that are the best traditional cone design I've ever heard in that price range. The bass depth is about the same, but the MMGs soundly whip these B&Ws in the realism and detail department. Here's the rub: the sensitivity for a pair of DM602 speakers is 90dB/1W/1M, and these are 8 ohm speakers. You easily need four times as much power to make the volume levels competitive, and the maximum SPL is at least 3dB higher with the DM602. Because of the lower amplifier power requirements, you could spend much less on a power amplifier than what you'd need to drive that big planar magnetic load and afford a better CD player or preamp instead. On the flip side, you need a set of stands to make the DM602 work correctly. For those that just want a set of simple speakers that are going to play loudly without a behemoth amplifier, the DM602 is the obvious choice. This is why I'll still recommend that my friends without an audiophile compulsion buy those speakers from B&W. You have to want that last bit of quality very badly to justify purchasing a demanding set of speakers from Magnepan. But if you're on the path to high-end audio nirvana but have that silly budget getting in your way, I don't know of anything else you can buy for $500 that satisfies that craving quite like a pair of MMGs.

Try it, you'll like it

As a parting note, a bit more information on the offers Magnepan has associated with these speakers. The in-home trial lasts 60 days. If you keep them, there's an offer to trade-up to bigger models in the Magneplanar line with up to a full refund (you don't quite get full trade-in value for the smaller models).

What will the extra money for a larger pair buy you? The MMG speakers are two-way; there's a quasi-ribbon tweeter and the trademark Magnepan planar magnetic midrange/woofer. I'm fuzzy on exactly what makes the tweeter a "quasi" ribbon, the only real performance difference I see is that the true ribbon used on their more expensive models has a full 360 degree radiation pattern, which they don't claim for the quasi-ribbon. Looking at the product lines, you'll notice that Magnepan really has only two types of design. The less expensive models ($500-$1995) all use drivers very similar to the MMG (some do have separate midrange and woofer panels). The true ribbon models start at $3150. Essentially, all you get as you upgrade to the bigger models within each line is deeper bass response. There's no radical improvement in the drivers on the more pricey models until you clear the $3000 mark, you just get more of the same basic sound. If you can live with the bass and volume restrictions of the MMG, there's no reason to be eyeing those more expensive models.

There's another promotion as well, where if you sucker...err, convince three friends to buy pairs of MMGs as well, you get a kickback...oops, I mean refund incentive. The whole things smells vaguely like some sort of pyramid scheme ("make Magnepans fast"), but it's a cool idea. I paid for my pair with my own money (that is, I will pay for them when I finally clear off that silly credit card bill); it's not like they give me the things for free. If you end buying a pair and keep them, let me know and maybe I can get dealt in on the whole refund setup. Most of my friends are too busy having kids for me to recommend they buy a pair of these; the thin planar magnetic material is not a good choice if you have poking fingers around.

A free trial, great sound, and you'll freak out everyone who visits ("those are speakers?"). What more could you ask for, right? Magnepan is really doing the budget audiophile a favor with the introduction of the MMG. If they sound like what you're looking for, it's certainly worth your trouble to check them out.

(Greg has a number of additional mutterings about audio and music available on his home page).