AR's marketing strength was their 12" woofer

[Steve F]

In the 20-20 hindsight division of the "What if" company emerges this thought:

I think AR made two big product blunders during their 1954-1974 Classic period by basing two models on 10" woofers instead of 12" woofers.

In the Roy Allison interview with David Lander that appeared in Stereophile magazine a few years ago, Roy said that "all the [original Classic] speakers were phenomenally successful, except for the AR-5."

Here's a possible explanation: The AR-5 was a 10" 3a. It listed for $175 ($156 unfinished), so compared to the Large Advent ($102-$116, depending on finish), its bass response was markedly less extended and its somewhat smoother, better-dispersed mids-highs were not enough to make up that difference.

What if....AR had done a "12-inch 2ax" instead of a 10-inch 3a? COGS-wise, certainly subbing the 3 1/2" cone mid for the very expensive 1 1/2" dome mid would have at least offset the 10 to 12" woofer cost differential, if not more. AR had the makings of a "12-inch" 2ax with the AR-1x. The 1x was a 12-inch woofer crossing over at 1200 Hz to the 4x-type 2 1/2" cone tweeter in a 3a-styled cabinet. They could have just added the 3/4" dome onto that and voila! There you go. Now you have a speaker that would be the same or lower cost than the AR-5 and be much more competitive in the marketplace, since it would bring the 3a's (and AR's) best performance trait--the 12" woofer's bass response--down into a more affordable price category.

They made the same mistake with the LST-2. It was an AR-5-based speaker. What they should have done is this: Retain the 12" woofer in the same original LST cabinet. Reduce the tweeter count to three (one centered on the front panel) and reduce the mid count to two--one each on the side panels.

This would have resulted in the exact same radiation pattern and tonal balance as the original LST, but with slightly less power handling than the LST-1 (although with 2x the mid and 3x the tweeter power handling of the 3a).

Those 1 1/2" mids were virtually the same cost as a 12" woofer, so reducing the mid count by two would have more than offset the cost increase from the 10 to 12" woofer.

They also would have realized a cost reduction on the LST cabinets, since now they'd be buying one cabinet in larger quantities for both models (just the baffle cutouts would have differed-a minor consideration.) Likewise, the carton packaging would have been common, so there's another cost savings. The original LST was $600 ea, the LST-2 was $400 ea. I certainly would have preferred a 12" LST-2 at $400 ea. to the actual 10" version.

Things are so clear and logical in retrospect when it doesn't matter anymore.

Steve F.

[Carlspeak]

Sorry Steve for going OT. I've hidden the last 3 posts

[RoyC]

Hi Steve,

I agree, the AR-5 and LST-2 were too expensive to manufacture for their intended market, but to presume their failure was in part due to a performance difference with the 12 inch woofer models may be erroneous. Although I agree AR needed something to compete with the Large Advent, I believe new drivers may have been needed to do so. I suppose, otoh, if a lower cost speaker was being marketed with the "same woofer" as the legendary AR-3a it could have been a selling point....at the right price.

It may have been diffcult for AR to produce a viable price-point competitor to the Large Advent using the AR 12 inch woofer and the 2ax mid. A key design component to the AR-3a (and AR-5) is the abilty to cross over to the midrange at 500-600hz. (In fact, the AR-5 actually pulls this feat off a bit better than the 3a because the smaller woofer is more easily crossed over at higher frequencies.) The 2ax mid is only effective down to around 1200hz, and there is no way the AR-12 inch woofer can operate comfortably much above 600hz.

Otoh, the Large Advent woofer cone, which is smaller than the large AR woofer cone and larger than the AR 10 inch woofer cone in actual size, was specifically designed by Kloss to operate well over 1000hz in a simple 2-way design using a proprietary mid/tweeter. The Advent bass response was "close enough" to the 3a for most common folk (like me, at the time:)), and somewhat better than the AR 10 inch designs...and the walnut cabinet and beveled edges looked 3a-ish. Hell, you could purchase 4 Large Advents for less than a pair of AR-3a's and be in audio heaven, according to the hype of the era:).

Another issue with the AR-12 inch woofer is the 2.5 ohm dcr of the voice coil. This woofer is very difficult to drive, and reeks havoc with any amplifier unable to provide reliable output at low impedances. The Advent woofer offers a friendly impedance to any amp. A higher impedance voice coil would have been advisable if the AR 12 inch woofer were placed in a lower cost speaker system.

As an aside, here is an excerpt from an early 70's Consumer Reports speaker review:

"AR-3a:

In April 1965, CU check-rated the Acoustic Research AR3 loudspeaker and noted that it had the deepest bass response of any loudspeaker we had ever tested. Naturally, we were curious about the quality of its successor, the AR3a, so we decided to test it even though its price ($250) put it well outside the scope of the accompanying report. The AR3a delivered just as good bass response as its predecessor, but we heard and measured a distinct exaggeration in the middle frequency register that made the speaker sound thick and heavy. (On investigation, we traced the peak to the AR3a's new mid-range unit, combined with a similar problem in the woofer.) We found a problem in the treble, too. Even with the tweeter control fully advanced, the tweeter could not match the relative loudness of the woofer. That made the treble in the AR3a sound a bit distant, although the tweeter frequency response was very smooth and extended. Had we ranked the AR3a with the other speakers in this report, it would have fallen between the high- and medium-accuracy groups. Thus, in CU's view, the AR2ax is a better speaker than the AR-3a, and the AR2ax costs only half as much."

Although the AR-5 addressed this issue somewhat, it was probably doomed from the start as long as the lower cost AR-2ax and Large Advent were in town.

Attached is the original text of the CU report in a Word file.

Roy

CU 3a CommentWord File.doc

[Steve F]

Perhaps I shouldn't have clouded the issue by positing the 12" '2ax' as an OLA competitor. Perhaps I should have just compared it to the 5 and asked,"Which speaker would have had a greater market appeal?" IMO, the 12" '2ax,' because AR's 12" bass response was far and away AR's most compelling attraction.

Re: "the AR 12" woofer couldn't get effectively above 600-700 Hz" issue. I don't buy it, not completely, anyway. Somewhat, but not completely. It's mostly popular myth, not supported by either historical AR speakers or actual loudspeaker science.

The AR-1 and AR-3 12" woofers went to 1000Hz. The 1x went to 1200 Hz. CU said the 3a sounded "thick and heavy" compared to the 3, yet the 3a x-o'd at 575 at the time of CU's test. So the 575 Hz x-o didn't magically make the 3a a perfect speaker. In fact, if the 3a was used with the amp's treble control up about 5 dB, it sounded great--no thickness, no heaviness. Advancing the treble control doesn't affect the w-m x-o point. With the 3a, it was a treble/tonal balance/spectral tilt issue, not a woofer-to-mid crossover frequency issue.

Further, there's nothing 'special' about the AR 12" woofer that makes it uniquely incapable to playing effectively above 700 Hz. Also, it's a "small" 12-inch woofer (about a 9.5" piston), so its mid dispersion is fine up to 1400 Hz or so. Do the math. (13560 divided by 9.5 = 1427.37 Hz--that's where it will become directional).

What are your thoughts about the LST-2 as I've laid it out above? Do you think a 12" LST-2 at $400 would have sold better or worse than the actual 10" woofer LST-2 for $400?

Steve F.

[RoyC]

I really don't think there was a place between the AR-2ax and the AR-3a for any version of the AR-5. I believe the AR-5 with the big woofer would have presented enough design problems to bring the cost to nearly as much as the 3a to manufacture. I've been knocking around speaker design and repair for nearly 40 years and there is no way the 2ax midrange and the AR 12 inch woofer will work together, much less to AR's standards of that era. The large woofer's bass response may be somewhat greater, but other aspects of the response would have suffered. Would it have sold better? maybe..

In fact, there have been some measurements and experimentation with the ceramic magnet 12 inch AR woofer in a two-way design in another audio forum showing significant response issues as it approaches 1000hz. It has nothing to do with directionality, simply frequency response issues. It should also be noted the earlier AR 12 inch cloth surround woofer used in the AR-1, 1x, 3, and early 3a is much better at handling higher frequencies than the foam surround, ceramic magnet version. I am sure this is why a larger inductor was used in the 3a's equipped with the later woofer. There are real reasons why speaker designers crossed over large woofers at lower and lower frequencies as time went on.

The AR-5 was an excellent sounding, well designed speaker system with no market. I personally prefer it over the 3a for all but deep bass-laden high volume level music. The primary reason, imo, has to do with the smaller woofer-to-mid transition. The greatest drawback to the AR-5 is that it is the least sensitive of all the AR systems of the day (in spite of stated specs). At low to moderate levels it is a clean, delightful speaker to listen to, but volume level is limited compared to its brethren (including the 2ax). It has to do with a higher system impedance, steeper crossover slopes, the woofer's lower crossover point (compared to the 2ax), and the heavy padding of the midrange for proper tonal balance.

As for the 3a being acceptable by turning up the treble control...all receivers and preamp tone controls have unique characteristics, varied shelving, slopes, etc, so the effect of using the control varies. If the flagship system requires this adjustment to sound as good as its lesser brother, what does it suggest about the design? In fairness to the 3a, however, none of its brethren of the era with the exception of the LST, can play as low or as "big" cleanly at high decibel levels.

Your large woofer-ed LST-2 may have been a more successful system, as long as a more robust cone midrange was used. A cabinet volume equal to the LST would also have to be used to accommodate the woofer's requirements, making it a larger speaker...and the low impedance of the system would still have to be taken into consideration. If it could have been pulled off for significantly less dough than the LST, I would agree with you on this one.

Based on serial numbers collected over many years and Tom Tyson's information, there were:

-Just over 100,000 AR-3a's sold (2/3 of them with the ceramic magnet, foam surround 12 inch woofer)

-Around 350,000+/- AR-2ax's

-Over 400,000 AR-4x's

-50,000 +/- AR-5's

Roy

[Pete B]

It is also possible that they were reserving the 12" for their top of the line only models.

I think they could have gotten around the design issues that Roy points out but what

if Steve's design ideas left no significant market for the higher priced models?

On the other hand AR really needed a low cost model to compete with the LA rather

than just ignoring the threat.

Who designed the AR-5?

[RoyC]

It is also possible that they were reserving the 12" for their top of the line only models.

I think they could have gotten around the design issues that Roy points out but what

if Steve's design ideas left no significant market for the higher priced models?

On the other hand AR really needed a low cost model to compete with the LA rather

than just ignoring the threat.

Who designed the AR-5?

I agree with all your comments, Pete. (I just don't think AR could have gotten around the design issues using an off-the-shelf 2ax mid.)

The AR-5 crossover layout and design are identical to the AR-3a, as were the AR-5's 8 ohm versions of the AR-3a tweeter and mid. Since it was introduced in1969, I'm guessing it was the same Allison/McShane combo that designed the 3a. Tom or Steve may know for sure.

Roy

[Steve F]

"I really don't think there was a place between the AR-2ax and the AR-3a for any version of the AR-5."

This is probably the most cogent comment of the entire 10-inch vs. 12-inch '2ax' discussion (including all of mine!). Thanks for summing it up very nicely, Roy.

Re: the 12" LST-2, a "more robust" midrange" would not have been needed, IMO. My proposed LST-2 would have had double the midrange PH of the 3a--a very significant improvement. If the customer wanted or needed more, they could buy the LST-1. But the line was incongruous the way it was:

3a (-3dB at 35, $250), LST-2 (-3dB at 44, $400), then LST (-3dB at 35, $600).

Better to not have a drop-off in LF FR when going from $250 to $400, and still maintain a significant, meaningfull increase in power-handling and dispersion as you step up the line.

Steve F.

[Steve F]

Oh, and yes--Roy A. headed the design effort on the AR-5. You can tell that it was "his" by his vociferous defense of the 5 when he objected to Julian Hirsch's findings of an inexplicable mid-range frequency response abberation compared to the 3a.

Roy insisted that in all AR's measurements and listening tests, the two speakers were absolutely identical above the very low bass.

Julian found a mid-range response dip in the 5 that he insisted was not an artifact of his particular testing environment or test set-up. Julian said the 5's FR anomaly persisted even when the speaker was measured under different conditions.

Roy and Stereo Review went back and forth a few times on this, and to SR's credit, they published Roy's letters.

But certainly, a questionable (even though mostly positive) review of the AR-5 in what was then the industry's most influential magazine by the industry's most widely-read reviewer didn't help the 5's market fortunes--in addition to the 5's "man-without-a-country" pricing and LF performance position.

Steve F.

[RoyC]

"I really don't think there was a place between the AR-2ax and the AR-3a for any version of the AR-5."

This is probably the most cogent comment of the entire 10-inch vs. 12-inch '2ax' discussion (including all of mine!). Thanks for summing it up very nicely, Roy.

I agree... guess I should have left it at that, Steve. Could have saved myself alot of typing.

(...even though I still don't think the big woofer and the 2a mid would be a good combo in any model due to frequency response limitations. )

Roy

[genek]

I think the idea of the AR-5 was to provide a speaker with AR-3a quality mid and high end but with less demanding amplifier requirements that would be better suited for smaller listening spaces, rather than a speaker to sell at a lower price point. It was a reasonable engineering goal, but it ran headfirst into consumers who looked only at the prices.

[ar_pro]

OK, so what makes the AR-5 about $50 per box more expensive than the AR-2ax?

It shares a cabinet, woofer, tweeter, and associated hardware with the 2ax, so that's a wash.

If the cost of the 2ax midrange is subtracted from that of the AR-5's mid, how much could really be left?

I guess I'm wondering why the 2ax wasn't simply replaced by the AR-5, taking advantage of (mostly) sharing the AR-3a's drivers, and perhaps reducing the per-unit cost through economy of scale.

Was there no way to build a $135 AR-5?

An AR-5 at that price-point would've rendered the 2ax obsolete, and would have been a much more refined system when compared to the Large Advent.

[Steve F]

Perhaps the 5 was priced higher than it need to be, but probably not by much. If I remember the parts pricing correctly, the 1 1/2" dome mid was very close to the same cost as a 12" woofer. The 2ax's 3 1/2" cone mid was about the price of a transistor radio speaker, so the delta in cost between a 2ax and 5 was not unsubstantial. I don't think the 5 could have been priced much below where it was--maybe $150-160--and it would still have been at a huge disadvantage to the OLA for bass extension and showroom "brightness," as well as value.

I think Roy has it right--the 5 was not needed, and its sales proved it. A great-sounding speaker, however.

Steve F.

[RoyC]

I've had the opportunity to do some side by side comparisons between the AR-2ax (early and later vesrsions), 5, and OLA. The subjective differences between the 2ax and the 5 are not easily heard. Actually, I believe the earlier AR-2ax with the cloth surround woofer and larger tweeter would be preferred by many people in this line up. It is easy to see why the showroom would have been a problem for the 5. It is also the least sensitive of the 3 models, so simply switching to it without compensating for the difference would have posed yet another disadvantage.

Another thought...Not only was the cost of the midrange higher, but the 5's crossover had many more components, adding to the parts and labor cost to manufacture.

Btw, while I know what CU was referring to with regard to the character of the AR-3a midrange response, I have not heard anything to suggest the AR-5 actually had a greater midrange response "issue" than the 3a. In fact, I've always preferred the 5 in that area. I wonder what Hirsch was detecting as compared to the 3a.

Roy

[Pete B]

My read is that AR was founded on the Acoustic Suspension woofer innovation

and later known also for the wide dispersion dome mid and tweeter innovations.

The AR-3a provides the best of everything in a bookshelf speaker.

The 2ax gets you the 10" woofer and dome tweeter with a low cost minimal crossover.

The AR-5 provides all the best in the way of wide dispersion and best crossover

design from AR at a higher impedance but with the 10" woofer.

It seems that the 2ax was good enough such that buyers did not see

the value in going to the 5.

The 5 would have made more sense if there was never a version of the 2 with

the .75" tweeter.

Who designed the 2ax?

[Steve F]

I'm pretty sure Roy A directed the development team for the 2ax, as he did all the Classics of that time period. Did Roy choose the cap values and do the A-B's in the lab of all the early prototypes? Probably not, that was probably an eng tech. But RA no doubt "signed off" on the final design.

Re audible differences between the 2ax and 5 (and 3a) through the midrange: I have very vivid memories of A-B's done with my (2nd-gen) 2ax's and my cousin's 3a's as well as a friend's 5's. There was a great Weather Report album on Columbia called Sweetnighter. The opening track Boogie-Woogie Waltz was a great speaker test track--lots of layered Latin percussion and a very, very deep bass line.

The differences between the 2ax and 3a in terms of midrange detail were nothing short of striking. Striking. The 1 1/2" dome revealed so much more detail and subtle percussive effects than the 2ax's 3 1/2" cone, it wasn't even close. It was not subtle or hard to hear at all. Same with the 5 vs. the 2ax. My cousin lent me his 3a's for the summer of 1973 and all summer long I A-B-C'd my 2ax's, his 3a's and my sister's 7's. (My Kenwood KA-7002 integrated amp had A-B-C speaker terminals.) I knew those speakers inside and out.

But in terms of tonal character, the 2ax had, for lack of a better descriptive phrase, a less "colored" tonal character than the 3a, which was kind of honky or nasal by comparison. However, the 3a was far, far more detailed, even though it was honkier. The 5 was just as much more detailed than the 2ax, but less honky than the 3a.

The 2ax had the most natural overall tonal character of the 3a, 5, and OLA. But it didn't have anywhere near the mid detail of the 1 1/2" dome mid AR speakers. I remember those listening sessions like they were yesterday.

Steve F.

[RoyC]

<<Same with the 5 vs. the 2ax.>>

Really? "Striking" would not be the word I would use in my recent experience, though I agree about the 2ax having a less colored character than the 3a (not so much the 5). In fact, I happen to prefer the 2ax (and 5, for that matter) midrange response over the 3a in the same way as the CU report cited above did...the "necessity" for tone controls notwithstanding. I agree, the overall sound of the 5 is more detailed than the 2ax, especially compared to the later version of the 2ax with the 3/4" dome. Of course the room and type of recording can have a dramatic effect on impressions...not to mention, at this point, such things as 40 year old tweeters...so who really knows.

Roy

[Steve F]

One man's "striking" is another man's "subtle." Certainly no implication on my part about either one of us being 'right' or 'wrong.' I'm simply recalling my experiences from 40 years ago and those of several fellow audiophile friends. I can hear the opening of Boogie-Woogie Waltz as clearly in my mind's ear as if it was yesterday. Switching from the 2ax to the 3a was like removing a heavy wool blanket wrap from around the speaker, that's how much more midrange percussive detail there was on the 3a (and 5) vs. the 2ax.

No amount of level control fiddling could make up the difference. These were all new speakers (less than a year and a half) and everything was in perfect working order.

Including my 19 year-old ears!

Steve F.

[Pete B]

What I noticed most about the 2ax in terms of flaws was a "phasing"

sound as I moved my head most likely caused by shifts in the frequency

response with position as a result of the simplistic crossover. This suggests

to me that the 2ax does not have very good polar response. Does anyone

know of a test report showing polars?

[Steve F]

High Fidelity Magazine reviewed the 'new' 2ax in their Nov 1971 issue. In 1971, the collective knowledge of the industry did not include much about near-field intereference, phasing effects, diffraction, etc. More complete knowledge came later, most noticeably in 1978 with Tim Holl's AR-9.

But to paraphrase the late, great Zilch, the Classic AR's (2ax, 5, 3a) were a "cacophony of near-field interference and phase cancellations." AR relied on far-field power response measurements and near-field individual driver measurements. They did not do near-field complete system measurements and didn't think those were particularly relevant to how a listener perceived the sound of a speaker 8 feet away in a normally-reverberant living room.

But HF Mag did those measurements and the 2ax is exposed for the near-field driver-placement/phasing flaws that is has, just like you heard.

HF's published curves are very revealing: Their "front hemispheric" curve--essentially a power response curve taken into the front 180 degree solid angle--is very smooth and slightly down-sloping. It conforms exactly with AR's design goals for the speaker and is a remarkably excellent measurement.

However, their "on-axis" curve is much rougher. The pertinent characteristic of this measurement is the very sharp 5 dB downward "V" notch centered at 5 khz--precisely the 2ax's mid-to-tweeter crossover frequency. The speaker was measured by High Fidelity with the 2ax's cabinet in the vertical orientation. So placed, the 2ax's mid and tweeter are perfectly side-by-side and all of their phase-related overlap and cancellation effects are in the horizontal plane.

This is precisely the destructive audible effect that the AR Vertcials--the 9, 90, 91, and 92--successfully avoided, because their vertical driver orientation preserved perfectly unfettered, interference- and cancellation-free FR response in the horizonal plane. The horizontal plane is the more important one, because listeners can and do sit at widely-differing angles horizontally relative to the speaker.

The 9 and its brethen were such well-designed speakers from so many standpoints and they were light-years ahead of the Classics and even the ADDs. To this day, virtually every serious speaker that has accuracy as its goal uses vertically-aligned drivers. Every one, from every company, regardless of bass loading technique or anything else. We can thank Tim Holl and the AR-9 development team.

But, yes. your observation of near-field "difficulties" of the 2ax are spot-on and borne out with documented 3rd-party proof. Good ears on your part.

Steve F.

[RoyC]

One man's "striking" is another man's "subtle." Certainly no implication on my part about either one of us being 'right' or 'wrong.' I'm simply recalling my experiences from 40 years ago and those of several fellow audiophile friends. I can hear the opening of Boogie-Woogie Waltz as clearly in my mind's ear as if it was yesterday. Switching from the 2ax to the 3a was like removing a heavy wool blanket wrap from around the speaker, that's how much more midrange percussive detail there was on the 3a (and 5) vs. the 2ax.

No amount of level control fiddling could make up the difference. These were all new speakers (less than a year and a half) and everything was in perfect working order.

Including my 19 year-old ears!

Steve F.

Actually I know what you mean by the "percussive detail" attributes of both the 3a and 5 compared to the 2ax. It is the 3a's tonal balance as it pertains to vocals I have a problem with. The "heavy wool blanket" may be removed relative to percussive detail, but the 3a has the heaviest blanket of all when it comes to vocals. This problem was addressed with the AR-11.

I have always found the 3a to be least desirable with vocal heavy music, and the AR-5 to be the most desirable of the speakers we are discussing...and whatever midrange attributes and anomalies they share are not as apparent to me. I completely agree with CU's assessment regarding the "thick and heavy" aspect of the 3a's lower mid response relative to the 2ax. I also remember reading a review of the day commenting that AR "got it right this time" in reference to the AR-5 midrange response compared to that of the 3a. This is my impression as well.

Regarding the 2ax, I suspect the simple, gradual 2-way type crossover plays a significant part in some of its issues. There is much overlap between the drivers in the 2ax, and obviously its recessed fiberglass covered midrange cone could never have the dispersion of the domes of the 5 and 3a.

Roy

[Steve F]

"Actually I know what you mean by the "percussive detail" attributes of both the 3a and 5 compared to the 2ax. It is the 3a's tonal balance as it pertains to vocals I have a problem with. The "heavy wool blanket" may be removed relative to percussive detail, but the 3a has the heaviest blanket of all when it comes to vocals. This problem was addresed with the AR-11."

I agree completely. The 11 was the 3a done right--no "if only's."

The founder and President of Boston Acoustics--formerly head product person at the original Advent Corporation--said to me on many occasions that in his opinion the 5 was AR's best speaker, and he had a grudging respect for the 2ax, which he called "sneaky."

He had a low opinion of the 3a, its tonal balance, its price, and its midrange driver, which he said "clipped" (his unusual but verbatim wording) too easily and didn't have good power handling due to its too-low crossover point. "No way a 1 1/2" dome should be crossing over at 575 Hz. I hate dome midranges--too fragile, too inefficient, and 4-6" cones have more than wide enough dispersion from 500-3000 Hz."

Steve F.

[ar_pro]

"No way a 1 1/2" dome should be crossing over at 575 Hz. I hate dome midranges--too fragile, too inefficient, and 4-6" cones have more than wide enough dispersion from 500-3000 Hz."

Didn't AR lower the midrange crossover frequency on the AR-11, compared to the AR-3a?

I know that it was raised to 700 Hz for the 3-way AR-91.

[RoyC]

"No way a 1 1/2" dome should be crossing over at 575 Hz. I hate dome midranges--too fragile, too inefficient, and 4-6" cones have more than wide enough dispersion from 500-3000 Hz."

Didn't AR lower the midrange crossover frequency on the AR-11, compared to the AR-3a?

I know that it was raised to 700 Hz for the 3-way AR-91.

The AR-11's published specs say the mid crossover point was 525 Hz. The midrange crossover components are not really much different than that of the 3a...in fact, the cap is smaller. Perhaps the resistive characteristics of the new 3-way switch and the use of ferro fluid in the mid's voice coil gap accounted for the difference.

Roy

[ligs]

"Actually I know what you mean by the "percussive detail" attributes of both the 3a and 5 compared to the 2ax. It is the 3a's tonal balance as it pertains to vocals I have a problem with. The "heavy wool blanket" may be removed relative to percussive detail, but the 3a has the heaviest blanket of all when it comes to vocals. This problem was addresed with the AR-11."

I agree completely. The 11 was the 3a done right--no "if only's."

The founder and President of Boston Acoustics--formerly head product person at the original Advent Corporation--said to me on many occasions that in his opinion the 5 was AR's best speaker, and he had a grudging respect for the 2ax, which he called "sneaky."

He had a low opinion of the 3a, its tonal balance, its price, and its midrange driver, which he said "clipped" (his unusual but verbatim wording) too easily and didn't have good power handling due to its too-low crossover point. "No way a 1 1/2" dome should be crossing over at 575 Hz. I hate dome midranges--too fragile, too inefficient, and 4-6" cones have more than wide enough dispersion from 500-3000 Hz."

Steve F.

AR 58 is essentially a 3-way AR3a/11 with a 4-inch cone midrange. How does AR 58 sound like?