You are reading the older HTML site
Positive Feedback ISSUE 6
My Polk Surround System: A More or Less Conventional Review
The "Max Surround Project"Continued!
I've been living with my Polk Surround-Sound system for nearly six months now, and I think I now have enough data and impressions to attempt a review in the more or less conventional style (as promised). As multi-channel surround-sound is a relatively new medium, I may have to invent some relatively new and, as of today, idiosyncratically Dudely methods of my own to fill out my impressionsbeyond the usual check points. I hope you will be charitable with me, give me the benefit of the doubt, and try to follow my attempts to overcome the limitations of language. Someone once wrote, "Writing about music is like dancing about architecture." It is even harder to write about how a bunch of optically read computer data is transformed into music by a handful of transistors, some magnets, and some lightweight cones and domes. But I'll give it my best shot.
Setting up my first surround-sound speaker rig I luckily fell into the helping hands of a really good guy at Polk Audio, Paul DiComo. I can't say enough nice words about Paul. This is one of audio's really knowledgeable and thoughtful guys who pops up (like snowdrops) after a dreadful blizzard to shovel your driveway in a gesture of good neighborliness. Paul walked me through the selection process of the rig I listen to, and explained why each of the following speakers were good "mates" for the others in this system, one that is (more or less) customized to my, pardon the expression, listening areathe acoustical equivalent of a black hole in the galaxy from which no good sound escapes. The results have proven more than I could have imagined.
Bout Dem Drivers
To begin, all the speakers come from Polk's LSi series, their top-of-the-line. They also make an RTi series, an R series, an RM series, and a DS system. For more information, click on their website http://www.polkaudio.com/polk.html. This system represents their best shot, and it is built around the Vifa ring radiator tweeter that is often found on speakers that are much more costly, such as the Krell, Audio Physic, and Mission top-of-their-line offerings. The ring radiator contains a series of innovations in tweeter design that make it able to deliver flat response at high dynamic levels without breaking up, and that means openness and clarity of detail without stridency or harshness. This minimal distortion tweeter is able to capture details at high or low levels that signify spatial cues. Such measurable performance translates to the presentation of a wide and deep sound stage, at any listening level, with a high degree of stability. The ring radiator is the speaker designer's ideal, a tweeter that displays exceptional dynamic range without distortion or losing focus or resolving power. I think that's something we've all been eagerly awaiting: a flat, accurate, loud tweeter.
One clue to its performance is that while many tweeters require a 3" circular cutout on a baffle board, the ring radiator requires a 4" cutout. If the area of a circle is pi times the square of the radius, you get nearly a 7" area magnet for a 3" hole, and slightly more than a 12" area magnet for a 4" hole. So, while only 25% larger in area, the magnet may be nearly twice as strong. These are the little things you notice while reading the spec sheets on various tweeters. As you might imagine, for a voice coil of approximately the same diameter, bigger, stronger magnets usually generate greater control.
John Atkinson, Stereophile's editor and resident audio engineer, after testing (March,2003) the LSi 7's tweeter performance in a "cumulative spectral decay plot," described it as "superbly clean over most of the region shown," though "shelving down slightly above 10kHz." If (what is laughingly known as) my memory serves, the Polk website has a nice description of the ring-radiator with a cut-away diagram that explains in greater detail how it works.
The next driver common to all speakers in the LS i line is its aerated polypropylene coned, 5 1/4" mid-bass driver, which Polk heralds as its most "utterly lifelike and completely natural." While this driver doesn't convey the startling "you are there" quality of electrostatics, it seems "fast" and it does offer a very convincing facsimile of, say, the hard-to-get-just-right human voice; especially those with which we've become pretty familiar (say, Ella Fitzgerald's and Louis Armstrong's) and have heard on many different systems, TV sets, car radios, portable CD players with various headphones. I figure, if a system gets the tricky timbres of those two voices just right, Ella's velvety adolescent contralto, Louis' grit and gravel baritone, they'll do alright with strings and woodwinds. As I've said before (see Positive Feedback Online, Issue 4, over at http://www.positive-feedback.com/Issue4/polkaudio.htm), the LSi 9's are a tad on the forgiving side, and I find that is a function of their mid-bass drivers. They are definitely soprano-friendly, and I guess that makes them orchestra-friendly; that is to say, without overmuch emphasis on the irritating spikes and spuriae introduced by the recording process itself, they deliver the sound of instruments nearly as I'm used to hearing them in my favorite seat in my local symphony hall. That's impressive. It means they get the basic tone and its overtone structure right!
The boxes that house these drivers have been the subject of much Polk thought, trial and error, incremental changes, and were eventually granted patents. The three different models, the LSi 9 (my front channels), LSi 7 (my surround channels), and the LSi C (or Center Channel), each employ the, rear-firing, Polk-patented, Power Port. Compared with a sealed box or a regular old cardboard tube port, the Power Port minimizes chuffing at the mouth of the vent, and disperses the turbulence that often results in back-pressure on the drivers that rob them of bass extension and dynamic range. Paul DiComo tells me the Power Port allows the speakers to do better in the low frequency range (according to Polk's measurements in their anechoic chamber) than with any other configuration they've tried. The Power Port is what allows the LSi 9 to be able to reproduce rock drumming and bass guitar with unexpected authority even without a sub-woofer. While the 7s have only one woofer each, and go down to the mid 50 Hz region, the 9s have a second mid-bass driver to augment the bass frequencies that allow them to go convincingly down to the lower 40 Hz region. Though physically resembling the D'Appolito designs, the Polk's 9s roll off the second woofer as it approaches the midrange.
There is also a front-firing tuned port in each of these speakers known as the "Acoustic Resonance Control." Its function is "to suppress the internal front-to-back standing-wave resonance of the cabinet by resonating out of phase and at the same frequency" (according to Matthew Polk in a letter published in Stereophile, March 2003). These box resonances are smackdab in the mid-range (800-900 Hz) for all the LSi line, as the frequency is generated by the in-box, front-to-rear distance and these speakers have similar dimensions. I'd guess the ports are what make the whole LSi family of speakers so friendly to the human voice by minimizing box resonances that can emphasize the aw, ooh, ah, or eee colorations of the human voice.
Finally, my speaker array closes out with a powered (100 watts) 8 inch sub-woofer. This downward-vented, front-firing, direct radiator comes in a black high-gloss enclosure. It can be placed nearly anywhere in the listening room, and since each room has its sonic thumbprint, with a little trial and error its optimum position will become evident. The PSW-550 also benefits from Polk's standard system of internal bracing and Power Port. I must say this sub woofer didn't knock me out at first, but I came to appreciate its truthfulness over time. I haven't heard all the subs on the market, but I've heard some that were neither as quick nor as nimble in the lower depths. I'd say this woofer is a fit mate for the rest of the system, going low enough to let you know there are organ pedal tones in the mix (when there are), and quick enough to hear the (what I'd like to call) braking horsepowerstopping cone motion before virtually any overhang, or bass ringing, takes place. A nice sized, nice priced, admirable performer. You can't expect more down to 32 Hz, especially if you plan to use it in a small to mid-sized listening room.
By the time you read this, all the Polk sub-woofers will have had their power amplifier output doubled. That will be a minimal price increase for a considerable improvement in performance. Time marches on!
What I listen to, then, is a surround-sound system with front, center, and rear loudspeakers all made of the same Polk woofers and ring-radiator tweeters, with similar crossover knees and slopes, box size, rear Power Ports, and front Acoustic Resonance Control ports, that result in very similar timbre. The front left and right speakers, the 9s, have more bass output than the 7s, but are otherwise the same. The Center is made of identical parts as the 9s, only configured horizontally.
I guess it is time for me to confess I like my music to reach facsimile proportions, or, I like my music system to play back at (as near as possible) the level at which the music was recorded. Usually that is louder than my wife likes to listen tothose gender differences in perception, again. It is at these volume levels that I hear the surround-sound most vividly. The lower the level, the less you will hear the surround in action. When I crank the system up, and sit in the sweet spot, I hear a sound very similar to my killer stereo rig in timbre, vocal reproduction, and spatial arrangement of the sound.
How do we know what we think we know? I'm not going post-modern on you; I'm trying to work this out, again, for myself, and for you, with regard to sound reproduction. I've had the same set of big speakers for a long time (since 84), Dynaudio Consequences. They were Dyna's top of the line eighteen or so years ago, and they were quite excellent as things went, then. I've been "upgrading" them over the years. One of the features was an Isobarik loaded 12" woofer (their 30W-100, with a 4" voice coil), which means two 12" woofers; the first firing into the room, and the second (in its sealed chamber) firing into the back of the first (in its sealed chamber).
Please pardon me if I get didactic and offer a quickie refresher course on "Isobarik." The Isobarik idea is to get the second woofer to act as a governor and keep the first from going non-linear (which had the mirror-image effect on the second as well). Theoretically, it is by keeping a constant pressure on each other. Let's say the first woofer is at its maximum excursion of 1 inch. Without a second woofer to keep the pressure constant ("iso," means "same," as in isosceles triangles: "barik," weight or pressure, from the same root as barometer), the air is compressed and rarified thirty times per second on a low organ pedal tone. When the sealed box air is compressed (without a second woofer) the air acts like a spring, pushing the woofer cone out; when rarified, it acts (somewhat) like a suction cup. With a second woofer, firing in phase, the spring-like effect of the compressed air is minimized in both directions. I'd say it is a very good theory, because in practice Isobarik loading keeps the first woofer from getting undisciplined, especially at its outer limits. I'd have to admit that it is pretty expensive way to go, what with a second driver and its enclosure. But, I liked that part of the design, allowing the woofers to do all they are designed to do without box constraints. In my room, it is hard to tell the difference between twin Isobarik 12s and twin 18" Klipsch horns (100 Hz and down) on most music.
The rest of the original loudspeaker was a series of Dynaudio's best, in those days, which amounted to a four-way speaker: a 6" mid-bass driver (17W-75), a 2" mid-range dome (D-54), a 1.1" dome tweeter (D-28), and a 0.8" dome super-tweeter (D-21). The crossover had 1st order (6 dB) slopes, with the 3dB down points in most of the usual places. However, the least loud of the drivers was the 6 inch mid-bass driver, that made less than 86 dB (measured) for one watt, and each of the other drivers had to be padded back. I decided I would replace the Consequence's 6" driver with Dynaudio's 9" woofer (21W-54) that would make a measured 89 dB for one watt. This swap would allow me to adjust the whole of that speaker module to 89dB. That was a little louder than the 12 inch woofers (30W-100), which were measured at 88 dB. To correct the 9 inch woofer and drop it back an accurate dB (so I could match it up with the 12" Isobariks) was a task too daunting for me in the days before computer-assisted crossover design.
My solution was to purchase a Shadow electronic crossover that featured an adjustable bass lift of up to 3 dB, and drive the woofers by their own dedicated amplifier. This swap-out gave the whole system 3 dB more sensitivity than I'd had and meant I could use a relatively low powered tube-amplifier from 100 Hz up.
A note about the 9" woofer. The one to which I was first introduced was made in Denmark and was marketed as Sen-Lab, and had the same designation 21W-54. Later, Dynaudio marketed its 21W-54 on the same cast metal frame. It had an early paper version, and by the mid 1980's had upgraded to poly-propylene. Later, in the '90s, Scan-Speak would offer a driver on the same cast metal frame, looking very similar, and known as their 21W-8555. Assuming there were incremental improvements (carbon fiber filled paper cone, butyl rubber surround), it represented the fourth iteration in a series of loudspeakers on that same frame over twenty years by the same group of designers. I wound up using the Scan Speak 21W-8155.01 (long throw) driver. It's used in David Robinson's Nova, and it is still considered The King of the mid-sized bass drivers. It is frequently used in double configuration.
And Then Came Morel...
No sooner had I gotten the speaker up and running, with a new woofer amp, and a tube amp (Sid Smith pooged Marantz 8B) for everything above 100 Hz, when I read an article by the Southern Michigan Woofer and Tweeter Society claiming they'd had very good results with a 3" Morel mid-range dome (MDM-75), and its sister piece, a horn-loaded 1.1" tweeter (MDM-27)both very similar in design to the Dynaudio pieces of that time. The article had oscilloscope photos of sweeps and rise times 'n stuff. When I heard them, I decided I had to go out n get em. They were a bit of a revelation. So I swapped out the Dynaudios for the Morels, getting more chest tones than head tones on female vocalists. [After hearing the ring-radiator at work, I am about to swap out my fifteen year old Morel tweeters for the Vifa ring-radiator tweeters, with anticipation of greater clarity and a larger sweet spot for the listener. More on that sometime soon.]
Jumped Up Consequences
So, now I had my Consequences, with their Isobarik 12" Dynaudio woofers (driven in parallel), powered by an Adcom amplifier (GFA-555) at 200 watts per channel, though when faced with a low, 3.3 ohm load, the amp became a 400 wpc amp. It sounded real good, especially when I popped in a new, up-to-date chip (Linear Technology's LT 1361 dual unity-gain op amp) in my Shadow Electronic Crossover. I eliminated a humongous crossover choke that used to roll off the woofers and put a haze on everything else in the original Consequence. I gained 3 dB of sensitivity, and a world class 9 inch mid-bass woofer. Everything else remained the same: Scanspeak 21 W-8555-01 (long throw) 9inch woofers, the Morel MDM-75 horn-loaded 3" midrange domes, Morel MDM-27 horn-loaded 1.1" domed tweeters, and Morel MDM-29 flat-flange 1.1" dome super tweeters. This was now mixed by a 2nd order crossover with premium parts (Goetz ribbon chokes, polypropylene caps) computer generated by Terry Chaika at Madisound.
This was becoming one hell of a system. I thought I had hypnotized myself into thinking I was getting really outstanding resolution until one of my pals came in and upon listening to something we both have known a long time, Peter Maag and the London Symphony Orchestra's (Decca SXL 2060) reading of Mendelssohn's A Midsummer Night's Dream, he said, "I never caught that low frequency going on before. Was that a subway?" Of course I had. It was a test of subtlety of the listener, not of the speakers. I'm sure I've heard a bug land on Jerry Garcia in the middle of "There Ain't No Bugs On Me," but the system could never differentiate, or I was never sure if it was a mosquito or a small black fly; and if a black fly, a Cremonese or Venetian black fly. Hear Grisman and Garcia, Not For Kids Only (Acoustic Disc ACD-9). I was now getting very high class sound. Maybe not world class, but not chicken salad, either. Maybe high second tier, or, if you're charitable, low first tier. Accurate timbre top to bottom, somewhat mellow yet very detailed, with lots of "zotz" (or dynamics) without sounding like PA, and super sound staging, presenting wider images than the speakers width when that is the engineers intent, and multi-layered depth, when that is in the recording (say the LP, Bach Music For 4 Harpsichords, [London, 15075]) with two soloists up front and two more even in front of them, strings and woodwinds behind, brass in the next layer, and bass viols and percussion behind them). I've been tweaking them for over fifteen years.
The Impact of Surround Sound
When I'd gotten used to my Polk LSi 9s, I took them downstairs to my big rig and substituted them for my 100 Hz and up section. Of course I had to place them on pedestals. At first I was blown away at how similar they sounded to my DIY speakers (Maybe that's what drew me to them initially?): almost their equal on most parameters, timbre, zotz, soprano friendliness; nearly equal in detail, clarity, sound-staging, imaging, freedom from harshness; if you like the woody sound of the violin (say, of the Stradivarius) as opposed to the brilliant sound (say, of the Guarnerius), then superior. But very near the very best in many aspects, like freedom from obnoxious box colorations. These Polks are terrific performing speakers. Playing over my Isobarik dual Dyanaudio 12" woofers, the differences between my speaker and the Polk LSi 9s are real if not great, which makes the Polks a true bargain in the audio market at $1040/pair MSRP.
When I took the 9s back upstairs, and played the same SACDs through the system all set up for surround, with the rear speakers properly placed, all the relationships were the same, except the sound-stage was scaled in proportion to the size of the room. I had heard the Polk 9s fill a much larger room, downstairs, so I knew they could do it. I disconnected the speaker leads on my surrounding 7s and played the music without those channels (stereo, plus center and sub). The roundness, or fullness of the sound stage diminished a noticeable notch, but it didn't do irreparable harm. I started out listening for sound staging and imaging, but I thought I heard a tad better clarity on solo harp. I put the 7s back on line. The clarity was somewhat improved. There then followed a march of the comedian, me: upstairs to downstairs, downstairs to upstairs, dragging speakers up and down, dragging pedestals down and up. Me in my clumsy, klutzy way, kvetching, whining at my dog, irritating my old athletic knees, for what? What was going on?
When I'd compared the 9s to my DIYs in my downstairs room, a large room (30' x 15' with 10' ceilings), I heard small difference in clarity. If anything, I thought that was one area in which my DIYs had a clear edge. I inveigled Corno de Bassetto to stop by and listen when he was in town. We spent some hours with this riddle. First one sounded better on one special passage of music, and then the other. After trying to understand what had happened, I concluded that since the SACD players and the amplification were so low in measured distortion, this difference in clarity was likely some form of distortion, or smearing of direct sound by reflected sound introduced by the room: slap back echo distortion, or room node induced distortion, etc. Something like that. This difference in clarity led me to make the inference that a good surround-sound system could "correct" room anomalies in problem rooms. Not necessarily, not absolutely, not in each and every room. But likely in smaller rooms like my "black hole."
I've discussed this in greater detail in PF Online, Issue 5, in a record review of Yolanda Kondonassis's harp music in "MAXimum SACDs: One from Decca, Two From Telarc." (See http://www.positive-feedback.com/Issue5/sacd.htm.) The surround channels get the information to the ear/brain for processing a little faster and louder than the natural room reflections. As they do this in a coherent and controlled way, they pre-empt the time smearing found in most untreated rooms. We hear the reflections as they occurred in the recording venue. We do get greater focus and clarity. The point here is, all things being of more or less equal quality (SACD players, amplifiers, interconnects and cabling) my Polk surround sound system is a good enough resolver of acoustic detail to be a reference tool. That is not to say it is a reference system, and it does most things better than most all other systems. It is to say that it helped me diagnose my big room as needing more acoustic treatment. It captures the subtlest of subtleties, small blurring of transients. Of that there is no doubt in my mind. And that realization was a surprise to me. More than surprised, I was astonished. And I concluded my Polk surround system is way cool enough a diagnostic instrument with which to review new SACD or DVD-A multi-channel releases.
It has often been said that a surround system with Brand X speakers seems somehow more satisfying than a comparable stereo-only system of the same Brand X in the same room. I had thought the larger number of speakers was likely covering up speaker defects, or nulling them out. Now I see (well, think I see) that the time aligned surround channels (if done accurately) actually can pre-empt the random reflections of a room, particularly an audio unfriendly room. So the system acts to mask the defects in the room, not defects in the speakers. With my Marantz SR 9200 front end, my Polks do this very well, indeed. If you're in the market for a surround system, and your front end (multi-channel CD player and Home Theater Receiver) is as good as mine, the Polk LS i speakers will deliver the goods every time.
In their brochure, of the LSi series the Polk copywriters say, "listeners will marvel at their accuracy and resolution... and remarkable reproduction of even the finest details." At first reading, I was almost moved to laugh at this hyperbole. But reviewer after reviewer has similarly chimed in about the virtues of these speakers: Anthony Cordesman at The Absolute Sound, Robert Reina and John Atkinson at Stereophile, Doug Schneider at Home Audio Equipment Review. The LSi speakers have won awards and citations in the Italian and American audio press. I was a doubter, but here I am, a convert, using similar phrases to describe the virtues of this loudspeaker system, actually the whole nine yards of Polk's surround sound system with sub-woofer. It is quite something.
Meditations on Surround Sound
How can I explain this surround-sound phenomenon? I'm not sure. I'm pretty sure I heard various differences between the speakers upstairs and downstairs, between my DIY speakers and the Polk 9s, between SACD stereo only and SACD multi-channel. I'm not an electronic engineer, nor a research designer, so I don't have the background with which to test my hypothesis. But, out of all my schlepping and hauling, kvetching and whining, listening to various speakers with different amplifiers, with different listening buddies, in different rooms, and trying to control for variables, just so I can give you, gentle reader, the benefit of my experiencesand, I think I can say without too much equivocation, SACD multi-channel seems to promise most improvement in acoustically difficult smaller and mid-sized rooms. I'm not sure if SACD multi-channel works a little better, or a lot better than SACD stereo in a larger [mildly acoustically treated (live-end/dead-end) to whole-hog treated (electronically diagnosed and rebuilt)] room. I'll have to wait till the deep pocket guys start reporting back on their dedicated rooms.
How does my Polk rig sound? It sounds great! Really. Detailed but not aggressive highs; fast, but not etched middle registers; adequate, if not bowl-you-over lower registersthough that is a well known limitation of room size and configuration. It doesn't make my small room sound infinitely large; there are still the constraints of size and walls; but it makes my "black hole" sound considerably larger, by seeming to make the walls disappear.
Does it sound as big as my downstairs room that is about twice as large? Well... no. Would that it were so. But it does hang an image of a symphony orchestra that's believable, if scaled down. On Carl Orff's Carmina Burana (Telarc, SACD-60675), Atlanta Symphony Orchestra & Chorus, Donald Runnicles, cond., it had Corno de Bassetto scratching his head in wonder. When baritone Earle Patriarco let go with Omnia sol temperat, we both got to chirping about "in the room" illusion, and later when the chorus joined in, it seemed the image was "wider" than the speakers, maybe wider than the walls. It certainly seemed deeper.
When I played Antonio Vivaldi's The Four Seasons, First Impression Music (FIM SACD 052), which is a bold, upfront recording of mostly violins, there was minimum scraping or scratching sounds that we've all come to put up with over the years. Old Corno was visibly blown away, saying things like, "I already have six or eight versions of this music that I've come to consider boring. But this recording, on this SACD technology makes it fresh again," or words to that effect. I'd add, this recording "on this rig." The Polk loudspeakers are obviously violin-friendly as well.
Conclusions about the Polks
So how do I rate the Polks, as a speaker system on which to listen to SACD multi-channel music?
I haven't heard the entire range of products out there, beyond little glimpses at the shows. I know my system of six speakers retails for about $3,200 MSRP, and it gets a large portion of the magic I've heard at the Chesky corporation's cost-no-object mastering studio. Which is not to say, you can't beat it. I'm sure you can. I'm saying my Polk speakers are good and plenty, and they do what they are called upon to do most excellently, as in Bill and Ted's Excellent Adventure. When you consider the "law of diminishing returns," you'll have to spend a lot to better these speakers.
As most folks familiar with Polk gear wind up saying; "Great performance for the money." They don't embarrass me. As a matter of fact, quite the opposite. The best way to explain it is with a rhetorical device, the chiasmus, an example of which might be: "Since I've owned my Polks, my Polks have owned me."
Catch ya later!