Polk LSi25 loudspeakers. (Equipment).
Source: Manufacturer Loan
Reviewer: Howard Ferstler
Up until now, the largest Polk speaker I have reviewed was the RT-35, several of which were included as part of a very workable home theater package I reviewed in Issue 77. At that time, I indicated that with or without a subwoofer they were among the better sounding smaller-room-oriented loudspeaker systems I had encountered. They were particularly impressive once their $400 per pair list price was taken into consideration. I believe this comment still holds, but now we have the LSi25, and in this case we are dealing with an all-new ball game.
A pair of LSi25s will set the serious enthusiast back a cool three grand, although a smart shopper should be able to hunt up some decent discounts. Discounted or not, this is way out of the price/size league that the RT-35 occupied. The question now is: can the more expensive system do as well in this price category as the lower-priced one could do in its price category?
Let's get some details out of the way before we answer this question.
The LSi25 is a substantial loudspeaker system. Floor-standing, it is nearly 50 inches tall with the supplied rubber feet installed, with a cabinet depth of 15.5 inches, and a width of just a bit over 8.5 inches. However, these dimensions need qualifying. The total width varies, because the bottom 3/5 of the cabinet has large wooden panels bonded to the sides (markedly stiffening its surface area down where the woofer driver resides), with the top third not being paneled, and that top section is only 7 inches wide. The narrow section that is flanked by those wood-finished panels is not done in a wood grain at all, but instead has been given a very high-gloss-black finish. The maximum depth of the entire system is actually 1,7 inches if you count the distance from the front of the grill to the back of the heat sinks.
Yep, I said heat sinks. The woofer section (which makes use of a side-aimed and downward-ported 10-inch driver) is powered and requires hookup to a wall outlet to work. The subwoofer amp is rated at 150 watts (which means 300 watts if we remember that each enclosure has a sub), and there is a control on the rear to adjust the low-pass rolloff frequency from 60 Hz to 130 Hz. In addition, there is a variable control for subwoofer level and a 0/180-degree phase switch. There is also an auto/off/on switch, which you would normally leave in the "auto" position, so that input signals could turn on the subwoofer section.
Above the low-bass range, the input is handled by two 5.25-inch middle-bass drivers that work as a team down low, but with one of the drivers attenuated in output as the signal continues to climb in frequency. The high-pass setting for the two of them when they are working together occurs at 80 Hz, with the first unit rolling off at 800 Hz and the second rolling off at 2.4 kHz. From there on up the input is handled by a one-inch Vifa XT25 ring-radiator tweeter, which Polk tells me is identical to one that is used in a $17,500 Krell system. The low-pass and high-pass slopes for the midrange drivers are 2nd order (12 dB per octave), while the high-pass filtering for the tweeter is 3rd order (18 dB per octave).
The LSi25 is rated at 88 dB efficiency, has a nominal, 4-ohm input impedance above the powered subwoofer's operating range (the input impedance is well above that from 80 Hz on down), and the system weighs in at a solid 91 pounds.
Hookup options include speaker level inputs (with the tweeter/mid and subwoofer sections normally tied together with jumper straps), line-level in/out connections, and a special LFE input. The line-level-out jack allows the two subwoofer sections to be daisy chained directly together if the user wants that kind of arrangement. While the line-level input is under the command of the low-pass filter control, the output is a full-bandwidth signal that would be independently low-pass filtered by the second subwoofer.
The variable low-pass adjustment and phase switch for the subwoofer section seems like an odd pair of controls to offer with a system that has the sub as part of a dedicated package. If the high-pass filter for both of the mid-bass drivers is 80 Hz, then it only makes sense to have the low-pass frequency locked in to properly accommodate that roll-off. However, under some conditions (a standing-wave or boundary-reinforcement peak) it might be desirable to set the low pass somewhat below 80 Hz. In other words, the variable control offers up some equalization options that a fixed high-pass/ low-pass setting would not offer.
The phase control is a different kettle of fish, however, and I really cannot see any pressing need for such a control, because the midrange and woofer sections should be properly set in terms of phase as a part of the design of the system. I suppose it could be used to fine tune any kind of low-pass overlap with the high-pass filtering to flatten out a response peak in that range. The user would want a very good measuring device and a good test disc to get that kind of job done right, however, and the existence of the control invites an inexpert user to make adjustment mistakes.
The system offers three kinds of hookup options. One involves the previously mentioned jumper straps between the speaker-level inputs for the entire system, with a standard, single-amplifier hookup to the connectors. This is the easiest and most straightforward hookup and the manual that comes with the system is clear about this being the preferred way to connect the speakers--and I concur.
A second hookup involves driving the mid/ tweeter sections with the output from an amp part of a preamp/amplifier combination, but driving the subwoofer with a line-level output directly from the preamp.(The jumper straps between the speaker-level connectors would be removed and Y connector would split the signal to the mid/tweeter section's outboard amp and the LSi25's own subwoofer amp.) This also will work fine, and one advantage is that it eliminates having low-bass signals be run through two power amplifiers (main and subwoofer) in series.
In theory, this should give better results, but in practice it probably does not add up to that. Because it will also involve running shielded cables from the preamp to the subwoofer, there are cable-length-related issues under some conditions (like the speakers being a long distance from the preamp) that would not exist with the first, more conventional hookup option noted.
The third hookup involves using the sub-out signal from a receiver or processor to power the LFE input. (Unlike the line-level input, the LFE input is not filtered.) One would use a Y connector to split the LFE signal and then it would be routed to the two LSi25's sub inputs by standard shielded cable. This kind of hookup would mandate configuring the processor or receiver to send all the system bass to the LFE output and not just the LFE part of it. (In other words, configure all satellites to the "small" setting.) If the processor or receiver had stereo sub-out jacks you could run their outputs to each sub and this would insure that much of the low bass would stay in stereo form. (This is a nonissue with me, but some people believe that stereo low bass is better than mono low bass.) Of course, you would get the same stereo bass behavior with the standard hookup that I first mentioned.
One disadvantage of the LFE hookup is that the midrange driver section of the LSi25 is already high-pass filtered at 80 Hz. However, a processor will have its own high-pass filtering when the satellite sizes are set to "small"' and so you would have two high-pass networks in series. There is no way to know for sure whether the low-pass filtering of the processor would dovetail properly with the redundant high-pass filtering in the speakers.
As best I can tell, the best hookup option remains the first one. It minimizes the number of errors that can happen.
Oh, yes, the manual offers up two other hookup options. One involves bi-wiring. While this may seem cool, there is no situation I know of where bi-wiring will deliver superior results. Sure, bi-wiring reduces wire resistance by hall but it only does so at and near the crossover point. Above and below that it has no impact over what you get with a single set of wires.
The second hookup option mentioned in the manual involves biamplification. Well, while this has its advantages with conventional speaker systems, it makes no sense at all to do this with a system that has a built-in, powered subwoofer to begin with. After all, the system already has biamplification if you use the second and third hookup options I noted above. While a separate tweeter/midrange amp can do just fine, the subwoofer amp simply does not need a second and separate external power amp feeding its input. The current draw with that input is tiny, and a second amp would deliver no advantages whatsoever. Indeed, you would have two amps in series feeding the subwoofer driver, and so the result would be no better than what you would get with the hookup I mentioned at the very beginning.
External amplifier biamplification would be silly with this system, or with any system that has an on-board, powered subwoofer. Why the manual mentions it at all is a mystery.
Of course, the manual I received with this pair of speakers was actually the weak point of the whole package. It is a preliminary item that hopefully will have been replaced with a more coherent, better-written version by the time this review gets into print. Not only does it offer up those needless and misleading hookup options previously noted, it does not mention the LFE hookup option at all. Indeed, the LFE input jack is not even pictured in the hookup diagrams. Even the web-site version of the manual I looked at was in error.
Anyway, enough of hookup options and manuals that have hopefully been rewritten and reprinted. I connected the speakers the old-fashioned way (option number one), and that is the way I suggest you do it. Let's now see how well the LSi25 systems performed during a series of measurement and listening sessions.
I first set the systems up in my main room and did my usual series of pink-noise room curves. As I have noted in the past, this consists of using the 20-second accumulation mode of my AudioControl SA-3051, 1/3-octave RTA while slowly moving the measurement microphone over a 1 x 1 x 5 foot area at the listening couch. While some individuals have indicated to me that a smaller measurement area would be better, they miss the point that what I am trying to come up with is a reasonably decent room curve that works for a broad listening area and not a one-location curve.
During this series, I located the speakers at various distances from the front and side walls in order to work out a compromise with the cumulative bass-reinforcement and cancellation effects from both the floor and wall boundaries. While the subwoofer crossover point is normally low enough for boundary artifacts to not be a problem with the sub driver, boundary cancellations and reinforcements could be a problem in the midbass with the drivers operating from 80 Hz on up.
I found that workable results were obtained in my 3,400 cubic foot main room with the speakers anywhere from 3 to 4 feet out from the front wall, putting the midrange/midbass driver centers about 4.5 to 5.5 feet from that boundary. Spacing between the speakers (driver centers, again) ranged anywhere from roughly 7.5 to 10 feet apart.
It did not pay to locate them so that the half the distance between the systems (driver centers) equaled the distance from the left and right speakers (driver centers) to the front wall, or to the side walls. With any speakers, and the Polks are not exempt from this, adjusting positions wrongly will compound any boundary-related suckout notches--the Allison Effect notch. Within reason, the further out from room boundaries the Polks were placed, the less likely it would be for the midrange/midbass drivers to have problems with compounded boundary-related artifacts.
The best placement in my main room delivered a room curve that was fairly flat from 300 Hz on up, with a mild peak at 500 Hz and a broad dip in the 1.6 to 2.5 kHz range. The peak would be driver-related and the dip would involve the interface between the midrange driver and the tweeter. From 3 kHz on up to 12.5 kHz the ring-radiator tweeter was very flat, although it performed no better in that respect than the tweeter in the NHT ST41 reviewed in Issue 90. (This tweeter is also used in the more upscale M6 model that has become one of my five reference systems.) Nor did it surpass the uniformity of the tweeter arrays in my Allison IC-20 systems.
Indeed, the NHT tweeter and Allison tweeters were both a bit flatter responding up to 12.5 kHz and both continued extending ruler flat out to 16 kHz, with the NHT continuing that way out to 20 kHz. The Polk tweeter was 5-6 dB down at 16 kHz, and more than 10 dB down at 20 kHz. Interestingly, the tweeter in the Polk RT-35 1 reviewed in issue 77 measured just as flat over its range as the LSi25 and was stronger out to 16 kHz.
Of course, ruler-flat tweeter response beyond 12.5 kHz is no big deal with just about any musical material, and so for all intents and purposes the Polk's ring radiator is right up there with the best.
Below 250 Hz, the response of the system began to climb and it was 8 or 9 dB above the average midrange level between 200 Hz and 100 Hz. This is not uncommon performance for a lot of speakers, although most that behave that way tend to be smaller models and not big, floor-standing jobs that do not require an elevated response in the middle bass to compensate for limited abilities in the lower-bass ranges.
Below 100-200 Hz frequency range, the level could be adjusted by means of the subwoofer level control. The sub itself delivered uniform output down to about 31.5 Hz, with a rapid rolloff below that frequency. At 25 Hz, it was about 11 dB down in relation to the output one-third octave higher up. The rise in the 100-200 Hz range was strictly related to the behavior of the midrange/midbass drivers, because it persisted even with the subwoofer cut completely off. I got the best subwoofer blend and least emphasis over middle-bass range if the subwoofer low-pass control was set to about 60 Hz.
These curves indicate that the speakers would tend to exhibit an overall warm sound, with an additional tendency to sound a bit distant, compared to systems with overall flatter response. It also indicates that it could deliver the bass-response goods with the vast bulk of program material. However, I would not expect the Polk subs to performance match the better outboard subwoofers in the range below 20-25 Hz. Bass sweeps with the Delos Surround Spectacular disc (DE 3179) demonstrated audibly clean and flat response to 30 Hz, with the output well attenuated by 25 Hz.
Using the imaging and panning signals on the same disc, the Polks exhibited uniform and well-focused images at all locations, including those at half left and half right. In this series, the Polks performed as well as any number of other fine systems I have auditioned, but as with those other systems the soundstaging only worked to perfection when listening from the sweet spot.
I also gave the paired subwoofer sections some maximum-output checks. This involved keeping the speakers where they were and locating the microphone at the listening-position location I use for all such tests. The speakers were roughly 14 feet from the microphone.
At 31.5 Hz, the Polks could cleanly achieve 101 dB, and at 20 Hz they could hit 83 dB. The latter figure again demonstrates that the systems are not tuned to go that low. At either frequency, if I tried to achieve higher levels the port emitted an unacceptable amount of turbulence noise. As a point of reference, a pair of Dunlavy Cantata systems could hit 102 dB at 31.5 Hz and 100 dB at 20 Hz, and my Allison IC-20 systems could hit 112 and 103 dB. A Hsu Research VTF-2 subwoofer (17 feet from the microphone, located in a corner) could hit 106 and 90 dB.
For most of my listening work, I had the speakers out about 3.5 feet from the front wall and about 10 feet apart. As with the test sequences, I used a Samsung DVD player as a CD sound source and fed its analog-out input to a Sunfire Theater Grand III processor that was operated in its two-channel, "source-direct" mode. Amplification was by means of a pair of Sherbourn 1/300MB monoblocks that I reviewed in Issue 92. To keep things upscale all the way I connected the speakers to the amps with Dunlavy LCR Ultra speaker wires.
I tried a variety of program sources, both new and old. One of the most impressive-sounding was a superb recording of Heinrich Biber and Johann Schmelzer instrumental music entitled Seventeenth Century Music and Dance from the Viennese Court (Chesky 173). I make a habit of using this disc as a reference demo tool, because it is almost unique in terms of realistic and perfected soundstaging, clarity, and depth. I must say that the Polks did a very good job with it. The mid-bass elevation imparted a warmth and richness to the sound that set the systems apart from some of the other models I have auditioned, and the mild dip in the midrange added a degree of depth and sense of larger-room space to the sound.
Another fine recording that highlighted the performance of the Polks was a new transcription of Gustav Holst's The Planets (Naxos 8.555776). This music covers the range from full bombast to extreme subtlety and the LSi25s covered that span with complete aplomb. Again, the sound had a somewhat warmish cast, but this was not altogether objectionable, and the subjective result reflected a solid richness on some passages. I also gave a new recording of Vaughan-Williams' Symphony Number 4 a go (Chandos 9984), and as with the Hoist work, the result was rich, full, and warm. This disc also includes six of Vaughan-Williams choral songs, and the Polks handled the vocal textures extremely well.
This listening and measuring session was rather lengthy and at near the end of it I checked out the sub-amp heat sinks on the back of the speakers and they were quite warm to the touch. Not extreme, mind you, but warmer than what I have encountered on a number of other subwoofers, including those that had conventional class A/B amps like the Polks.
Of course, measurements and single-presentation listening can only deliver part of what one might want to know about a pair of speakers. One other technique is to compare the pair under review to a pair of known high quality. In this case, a few days after my initial listening sessions I selected a pair of Dunlavy Cantata systems that I had already reviewed in issue 87. (While the Cantata is now officially discontinued, the upgraded SC-IIIa, which has the same $5,500 per pair list price, is pretty much electromechanically the same.) I consider the Cantata to be a reference standard, of sorts, at least for speakers designed to be placed some distance from wall boundaries. While it is priced well above the LSi25, I think that it is better to use a pair of reference standards for comparison purposes than systems that are sonically only just OK.
Of course, comparing speakers this way is problematical, because if they are arranged AB/BA the inside pair will not have the soundstage width of the outer pair. Consequently, I stagger them AB/ AB, which keeps the units the same distance apart. The down side to this technique is that the center of the soundstage shifts. This makes it tricky to determine center focus, but I believe it to be the lesser of two evils. To compensate, I try to shift my position slightly to the left or right as I listen. The location also somewhat compromised the side radiation of one of the LSi25 woofers. However, I did some measurements and found the problem to be inconsequential.
Power to the Polks remained as before. The Cantatas were hooked up to the amplifier section of a Yamaha DSP-1000 integrated amp fed through its aux inputs. The inputs to the two amps were controlled by a Radio Shack tape-switching control. This kind of arrangement allowed me to level match easily and get the average outputs of the speakers pretty close, and also allowed me to switch rapidly between each pair.
I initially tried the previously auditioned recordings, and I felt that the Polks sounded a bit thick in the middle bass compared to the Dunlavys, although with some passages they sounded richer, with the Cantatas seeming to be a tad thin.
I have a fine recording of Vivaldi concertos (Claves 50-9706), and with that material the Polks were definitely competitive. There is a lot of clean treble and upper-midrange sound on this release and both systems sounded quite similar at those frequencies. The Cantatas were perhaps a bit tighter sounding, and a bit better focused. However, the LSi25s seemed to be a bit more spacious, no doubt due to the wide, smooth dispersion pattern of those Vifa XT-25, ring-radiator tweeters. (The Cantata's Vifa D27TG-3506, a very high quality unit in itself, is flanked all the way around by felt damping material to control wide off-axis radiation.) The Polks also have smaller-diameter midrange drivers than the Dunlavys, which allowed the midrange to also be a bit more widely dispersed.
On the other hand, at the other end of the frequency spectrum the Cantatas were clear winners. On a superb recording of Mendelssohn Organ Works (Argo 414 420) the Cantatas did a better job of reproducing the deepest, sub-25-Hz pedal notes, although neither system was in the same class as the better subwoofers I have on hand.
I also tried some pop-music transcriptions. One of my favorites remains Dire Straits On Every Street (Warner 22680), and here the Polks seemed again to be a bit thicker sounding, with the Dunlavys a tad leaner, overall. The disc has some pretty low bass for a pop release, but it was not deep enough to show up any significant differences between the speakers. Another good jazz recording I recently obtained was Dear Dorothy: the Oz Sessions, performed by the Chad Lawson Trio (Summit 330). There are a lot of clean triangle and cymbal sounds on this release, and here the ring radiator in the Polks easily held its own against the tweeter in the Dunlavy systems. Throughout the midrange the Polks sounded quite good, bringing the piano sound on the disc more forward than what I heard with the Cantatas. However, the overall midrange balance was not as smooth as what the Dunlavy systems could deliver.
I also did a series of comparisons with the Delos Engineer s Choice disc (DE3512), which I like to use, because I have heard the assorted and well-recorded passages so many times. In this series, the Polks actually did quite well. On the guitar passages they easily matched the Cantatas and on some of the orchestral material I felt that they also matched the violin-reproducing abilities of the more expensive Dunlavys. In some instances I felt that the Polks even had absolute edge, particularly when it came to simulating a sense of hall space around the ensembles. On the female vocal passages, the Polks were terrific, easily being at parity with the Dunlavy systems.
Any speaker system is going to have its hands full when being compared to the Dunlavy Cantata, and I felt that the Polk pair held their own quite well. While not as smooth overall, the differences were usually slight, and in some respects (the ability to generate ambiance around the performers) the Polks were often superior. And while their bass response was not a match for the more expensive Cantatas down really low (both were equal with test tones down to 30 Hz, but down lower the Cantatas pulled away), the Polks were clearly able to hold their own with 95% of the music I auditioned.
The visual style of the systems will appeal to some listeners more than others, although there is little doubt that the side panels are functional and do add rigidity and mass to the woofer enclosure. And while some will wonder about the supposed advantages of the powered subwoofer section and the assorted hookup options (and might be put off by having to plug their systems into wall sockets), there are situations where the flexibility might pay off.
In terms of overall sound-reproducing abilities, the systems are certainly competitive with others in their price category. I'd say that the LSi25s are well worth their asking price, and if a typical dealer discount is factored in they would be a very good deal, indeed, or even a steal.
|Printer friendly Cite/link Email Feedback|
|Article Type:||Product/Service Evaluation|
|Date:||Apr 1, 2003|
|Previous Article:||Panasonic DVD-RV32 DVD player. (Equipment).|
|Next Article:||David Russell Plays Bach (Prelude, Fugue and allegro, BWV 998; Chorale Prelude," Wachet auf, ruff uns die Stimme," BWV 645; Partita No. 2, BWV 1004;...|
|Cables: Perception v. Reality.|
|Des Moines makes the switch to soy. (FYI News & Notes).|