Acoustic Research AR-303a Speakers

tmallin

WBF Technical Expert
May 19, 2010
1,071
452
1,655
72
Chicagoland
For background on the new second system in which I'm using these speakers, see this post and this post, both of which are in my overall discussion of the Lyngdorf TDAI-3400. Like my primary system, this new second system in my Blue Room uses a TDAI-3400 for amplification and streaming. This second system, unlike my primary system, is also used for watching broadcast TV and streaming video services.

This is the second time I've owned the AR-303a speakers as vintage speakers. The first time around I wrote comparing their sound to the earlier-yet classic AR-3a speakers. See my "Vintage or Less Vintage: The AR-3a vs. the AR-303a" thread. The first time I used the 303a speakers in a large living room, bi-amplified by four channels of an Arcam AVR600 A/V receiver (highly rated at the time by Peter Moncrieff of International Audio Review [IAR]). The speakers were on 24-inch high stands, aimed straight ahead (no toe in) and I sat back about 10 feet from them. The first time around I had issues with forwardness/brightness in the 2 - 5 kHz region.

This time, the speakers are single-amped, single wired with Benchmark speaker cables from the Lyngdorf TDAI 3400. The speakers sit on lower 18-inch-high stands (these stools, actually), are toed in to face their respective ears, and I listen in the near field, about 50 inches from my ears to the drivers. As toed in in my set up, the center fronts of the baffles are about 62 1/4 inches apart. Thus, the subtended angle between the speakers from the listening position may be viewed as somewhat greater than the "standard" 60 degrees. However, since I have the speakers set up with the offset midrange and tweeter on the inside, the subjective separation may be somewhat reduced. My ears in my unmodified present chair are an inch or two below the tweeter level. The top center front of each speaker is 23 3/4 inches from the sidewall, 30 1/8 inches from the wall behind the speakers, and the woofer centers are about 26 inches above the floor. The tweeters are 39 inches above the floor. The Blue Room is very small, about 9 1/2 feet wide along the speaker wall by 9 feet, with an 8 1/2 foot ceiling. The room has plaster wall and ceiling construction and the floor is carpeted with thick cut-pile wall-to-wall carpeting over a thick pad. The speaker baffles are a few inches in front of the 42-inch flat screen Sony OLED TV.

Cutting to the chase, this time around not only are the bass and midrange extremely fine sounding for a vintage speaker of this cost (you can find them online these days for around $1,200 a pair asking price), but the highs are also now correctly balanced and very fine sounding as well.

I'm not sure which factors made the difference in my current vs. past reaction to the sound: biamping vs. single amping, no toe in vs. toed in, high stands vs. lower stands, ear height relative to the tweeter, far field vs. near field--or maybe a combination of two or more of these factors. But my reaction is now more in line with most of the other online and print reviews of the sound of these speakers.

As far as reviews go, I refer you to John Atkinson's Stereophile review and the attached review by Robert E. Greene done for the print version of The Absolute Sound in July 1996 but which is not available online as far as I can discover.

I can verify by adjusting my vertical seating position that Atkinson is right about needing to sit on or just below the tweeter axis to get the best high frequencies from these speakers. My current small Drexel/Heritage chair was still about an inch too low even though I placed a Target shelf beneath the seat cushion to raise my ear height a couple of inches. I have ordered a pair of Penny Mustard barrel chairs for this Blue Room which I should receive by mid-January and which I know from experience (I have three of these same chairs in a downstairs room already) will raise my sitting position to 39 inches, which is right on axis with the tweeter with the speakers on their 18-inch stands. This ear height is more crucial in my current near field listening positon, of course, for simple geometric reasons. For the present, I have achieved this 39i-inch ear height by adding this gel cushion below the seat cushion of the chair instead of the Target shelf.

I hear the "roughness" REG refers to in his review but, oddly, this quality seems to diminish with proper listening axis (toed in and ear height relative to the tweeter) and near field listening. And while I hear a slight foreshortening of the depth of field compared to that from other speakers mentioned by both Atkinson and Greene, in this near-field set up, there is still enough depth and three dimensionality to satisfy, even on large-scale orchestral music.

What I now have is truly excellent vintage speaker sound. And the bass! I again find myself marveling at its depth, power, fullness, cleanness, and overall rightness. Even larger stand-mount speakers like the Harbeth 40s can't match these acoustic suspension ARs in the mid- to low-bass department.
 

Attachments

  • The Acoustic Research 303 Loudspeaker July 1996.pdf
    548.4 KB · Views: 21
Last edited:
I am not yet using the Lyngdorf's Room Perfect in this second system. I have tried the Voicings and occasionally they can be helpful. But, as set up, the tonal balance is quite pleasing on a wide variety of material from large scale classical, to chamber music, symphonic band, classical vocal, classical choral, organ, as well as all sorts of jazz and pop.

I really don't understand why more modern speakers are not voiced this way. Most commercial material on Tidal and Qobuz, as well as CDs, is at least a bit thin on the bottom and bright on the top. Internet radio tends to be worse yet. Speakers should be full and rich on the bottom and a bit down-sloping toward the top to counteract this reality. Instead, most speakers seem balanced just the opposite: at best flat in the low end, but tilted up in response in at least the top two octaves (5,000 Hz up) if not from around 2 kHz up.

This seems to be a preference for a lot of modern listeners. The changes made to the Wharfedale Linton are a recent example. The Wharfedale designer acknowledges addressing the minor criticisms of buyers of the overall very well received Linton 85th Anniversary Edition in designing its new Super Linton. Buyers wanted to be able to place the speaker closer to the wall behind it without incurring bass boom, wanted less reticence in the midrange, and wanted a more open sound in the top end. Comparing the Klippel measurements of both speakers posted on Erin's Audio Corner, you can see how in the Super Linton the bottom two octaves of bass were rolled off a bit, the mids became more forward/less recessed, and the middle high frequencies increased by about 1 dB. Erin likes the end result, but while I might appreciate the greater dynamic ability of the low end, I don't think I'd like the changes further up. (It's easy to add dynamic capability to a bass driver if you roll off its lower octave extension--it just is not working as hard at any given frequency at any given SPL.)

The AR-303a, as I think the Atkinson review summarized, are "big-hearted" in the low end. Atkinson meant it as uncomplimentary or at best a warning to those who want more neutral sounding speakers. But I take this characteristic as a partial remedy for what ails most commercial recordings.

The Dynaco A25 vintage speakers I had in this room before the ARs were even more natural or even in overall tonal balance, lacking the "big-heartedness" of the AR's low end. The Dynacos are very easy on the ears and dynamically responsive enough for any kind of music even at considerable playback levels in this small room. Their weaknesses are purely in a slight lack of transparency and lack of bass below 50 Hz and lack of high frequency sparkle--subjectively rolling off just a bit from 5 kHz on up, I'd say. The lack of bottom and top balance each other and sound quite "right" together, however, especially since the entire range is not as transparent as the best speakers. Gordon Holt's 1969 review of these Dynaco A25 speakers in Stereophile is right on the money, as I hear them.

The midrange of the ARs seems quite natural in balance, but is not as ultimately clear and revealing as excellent modern speakers like my Watkins Generation 4 or Graham LS8/1.

The AR-303a sound is wide open and has an excellent sense of the height illusion; no miniaturized instruments and voices here. And as you turn the volume up from moderate to life-like levels, there is nothing impeding the low end from staying full and low in distortion. Neither do the upper ranges evidence any sense of strain. The sound just gets bigger and the bass more powerful sounding as you reach levels where Fletcher-Munson effects are minimized. The big 12-inch acoustic suspension woofers really deliver the goods!

Listened to in the near field as I have them, there is a considerable sense of immersion, probably from side wall reflections in my acoustically untreated room. But as the Stereophile test report shows, the lateral off-axis radiation from the AR-303a is very even and smooth up to at least 6 kHz, and above that rolls off fairly smoothly. The sidewall reflections thus blend nicely with the stronger direct near-field sound and give a nice sense of the original natural or artificial venue, even though the wrap-around effects I hear in spades from my Watkins speakers in my other acoustically treated room are substantially diminished.
 
Last edited:
Here is a frequency response measurement taken by Robert Greene of his Acoustic Research AR-303 speakers. The AR-303 is basically the same speaker as the 303a I have except that the 303 uses only a single pair of different binding posts and thus is not biwireable or biampable. As notes to this graph he said that the room effect narrow dips in the lower frequencies are not a serious problem in audible terms and that the speakers actually sound quite convincing. He said they are maybe not quite as uncolored as he can get out of BBC models and the Spendor A4, but still good. And the bass is good indeed without a sub. He said he could have notched out the boom down in the bass but it seemed not too important in musical terms.


AR 303 Response2 per REG.jpg

From this graph you can see what I mean about the response being "big hearted" and generous in the low end, tapering down a bit in the top two octaves. This response complements many commercial recordings very well indeed.
 
Last edited:
Making Your Sonic Presentation Sound Larger

Many music lovers and audiophiles want their home systems to present a "big" sound, as in spatially more expansive. Many mega-bucks systems can do this and this is one factor that impresses about such systems. What are the factors which make for a spatially more expansive presentation?

I believe that there are several aspects to sound which can contribute to our perception of the "size" of the presentation:

Actually large physical size and/or subtended vertical angle--very tall speakers--tend to sound vertically big, whether they are planars or cones and domes. The largest non-stretched images I ever got from non-planar speakers were from the D'Apollito array of the Legacy Audio Whispers. I have never heard speakers reproduce that aspect of space any better or even quite as good. Really life-sized musicians as if heard from close up, but never a sense of vertical stretching. The speakers sounded not like a pulsating sphere with the implied compact size, but like a pulsating ellipse or egg with real height to very vertically focused images. Those speakers were quite tall and weighed 240 pounds each with their four 15-inch compound dipole woofers apiece.

Ambiance--how much of the size, shape, and boundaries of the recording venue are portrayed. This is one area where adding subwoofers or otherwise extending the bass to the very bottom of audibility and maybe even below can help. This is actually at least as important a contribution of adding subwoofers as any actual bass extension, given the lack of 20 - 40 Hz frequencies in most actual music.

Image height/size--are instruments squashed vertically, are they life-sized in height, or are they stretched vertically? Planars often stretch image height beyond life sized. The Sanders 10e were not bad at this--just a little vertically as well as horizontally stretched. Maggies tend to be quite vertically stretched and horizontally imaging challenged in terms of focus and stability. Three-way speakers where the tweeter is on top, mid in the middle, and woofer on the bottom can often produce life-sized images, as can good D'Apollito arrangements. Two-ways can be squashed but this is mitigated by careful listening position and perhaps angling the speakers back a bit so more sound hits the ceiling.

Envelopment--how much wrap-around effect there is to the stage. Is most of the sound just between the speakers? Does the soundstage extend to the walls? Does the stage wrap around toward the listening position? When phasing effects are recorded or created in the studio, is there a sense of wrap-around of staging past your sides and to your rear or even full envelopment, as you can have with surround sound set-ups? With two speaker stereo, much of this in my experience depends on how well your listening room's first reflection areas are treated, how exactly positioned your speakers are with respect to the listening seat, and how well the drivers "speak" in phase. First-order crossovers or DSPed crossovers where all the drivers' sound arrives at the same moment greatly enhance envelopment. The Watkins and Grahams are great at this; the Watkins because of their time-aligned first-order crossover and the Grahams because there is no crossover at all until the mid-treble.

Depth of stage--the deeper, the larger the apparent sound source--the same considerations apply here as with envelopment, I find.

Near field vs far field listening. In my experience, the closer up you can listen without losing inter-driver coherence from your speakers, the larger the direct sound will appear to be. Harbeths are the best at this in my experience. The M40 series can be listened to from as close as 20 inches from drivers to ears and still not hear out the driver positions as long as your listening axis is just an inch or two below the tweeter center. You can enhance this effect somewhat by carefully increasing the distance between speaker centers to beyond 60 degrees, just being careful not to develop the dreaded "hole in the middle sound" or making the resulting sound too bright. The more the sound comes at your ears from directly to the side, the brighter the sound. You can combat this by not toeing the speakers in to aim directly at your ears, but then you will get more side wall reflections--it's a balancing act. If you don't mind the "claustrophobic" feel of such large speakers being up close and personal (I did eventually mind and that is one reason I moved on from my M40.2s), you will certainly hear very big sound indeed with fully extended bass as well and no vertically stretched images, just life-sized sound.

Another factor which can influence how large a sonic presentation seems is the degree of ceiling reflection. Few speakers intentionally engage the ceiling as a reflector, but there are a few that do and they can sound quite huge even though they are short. The old EPI 201/Epicure 20 or 20+ is one. One Burhoe module faces front, the other faces the ceiling with just a slight angle forward toward the listening area. I keep looking for a decent pair of the Epicure version to hit the used market again so I can try them in my new vintage room. More recently Burhoe is doing basically the same thing in his Direct Acoustics Silent Speaker II model with just one module and it facing the ceiling the same way.

The other company which comes to mind is Shahinian. Several of their models feature a significant portion of ceiling-facing sound. The big Diapason/Double Eagle model used to sound magnificent on classical music at AXPONA before Shahinian died and the Shahinian room was always an oasis of classical music civility and was usually packed for that reason. He would play what you wanted to hear and seemed to appreciate the offerings from attendees. The speakers are still being made.

I suppose true omnidirectional speakers qualify, but I think that with true omnis the side wall reflections detract from the presentation unless the room is either very heavily padded or unless the room is very large with the speakers placed many feet from the walls. With short speakers and a relatively high ceiling, the reflections from the ceiling are far enough from the speakers and thus late enough arriving at the listening position to tend to add spaciousness and presentation size without annoying brightness or grit.

I also believe room treatment of the type which minimizes reflections from the listening room walls through absorption can make concert hall recordings sound much larger in presentation, even with small two-way speakers like my Watkins Generation 4. Absorbing the wall reflections from a small second-venue listening room unmasks the recorded acoustics of the vast concert hall, especially with speakers like the Watkins which have time coherent crossovers and angle the sound up toward the ceiling just a bit.

Finally, and perhaps seeming to contradict the prior paragraph, speakers which can intentionally take advantage of undamped reflections from some room surfaces can create a larger presentation. Maggies and other planars can do this, as do cone and dome speakers such as the Linkwitz dipoles. One can quibble about whether such presentations tend to create more of a "they are here" feel than a "you are there" impression, but there is no doubt that the wall reflections can make the presentation seem larger.

Finally finally, keep in mind that some attempts to enlarge to apparent size of the presentation will trade the larger stage/presentation for imaging specificity and focus. You have to decide whether this trade off is worth it for you. To me, it usually is since live unamplified music generally does not produce the kind of focused images possible with many two-channel stereo set-ups, and since, on the other hand, live music heard from my preferred listening positions of fairly close up usually sounds huge and enveloping in terms of staging and overall presentation.
 
Last edited:
  • Like
Reactions: QuadDiffuser
I said in post #1 above:

"The [AR-303a] speakers sit on lower 18-inch-high stands (these stools, actually), are toed in to face their respective ears, and I listen in the near field, about 50 inches from my ears to the drivers. As toed in in my set up, the center fronts of the baffles are about 62 1/4 inches apart. Thus, the subtended angle between the speakers from the listening position may be viewed as somewhat greater than the "standard" 60 degrees. However, since I have the speakers set up with the offset midrange and tweeter on the inside, the subjective separation may be somewhat reduced."

Actually, I have adjusted my listening position to produce what I subjectively find to be the best combination of imaging and staging without moviing the speakers. This involved moving the listening position yet a few inches closer to the speakers. The speaker boxes now appear to be aimed at my shoulders.

However, when I remove the grills so that I can see the drivers, I see that the inboard midrange and tweeter of each speaker are actually aimed directly at their respective ears--the left speaker's tweeter/midrange aimed directly at my left ear and the right speaker's tweeter/midrange aimed directly at my right ear. This is the orientation I have generally found to sound best to me for cone/dome and panel speakers used in my audio rooms and listened to in the near-ish field over the past couple of decades.
 
Grills or no Grills? That Is the Question

Some of the speakers I've used over the years have sounded better to me (especially in terms of high frequency balance and the spatial presentation) when used with the factory grills in place. Harbeths definitely fall into this category (the manufacturer specifically recommends listening and measuring with the grills ON), as well as the Stirling Broadcast LS3/6 and Graham Audio LS8/1. Other speakers have sounded better without grills; examples include the Sanders 10e and my Watkins Generation 4.

In my opinion, in my near-field set-up, the Acoustic Research AR-303a speakers definitely sound better with the grills ON. I have experimented both ways numerous times, from several listening positions, both in daylight where I can see the drivers, and at night when, without the room lights on and the blinds closed, the drivers are very difficult to discern even with the grills off. I have also tried listening with eyes closed and eyes open under both circumstances.

The result is always the same. It is not so much a matter of perceived frequency response. That seems about the same regardless of whether the grills are on or off. But the sound clings to the speaker positions a bit more and there is less perception of depth of field with the grills off. Since, as other reviewers have also noted, the speakers already are a bit depth-of-field challenged, grills ON is definitely the way to listen to my ears in my system.
 
I have doubled down on the AR-303a speakers. I just bought another very nice looking rosewood pair that was for offer on eBay. I will replace the Watkins Generation 4 speakers in my other room with the "new" 303a's for a while. I REALLY like the sound of the pair I have in my Blue Room, especially now that I have applied Room Perfect to them. The changes were more subtle than with the Watkins (see this post), but, again, all to the good, further extending and clarifying the already well-extended low bass, taming the bit of excess midbass, flattening the midrange forwardness, generally firming and stabilizing the imaging, and improving the depth of field. Again, as with using Room Perfect with the Watkins, I did not move the microphone in running Room Perfect.
 
Last edited:
I have received my second set of AR-303a speakers. These are both in excellent physical and functional shape. I have begun installing them in my audio room, replacing for now my Watkins Generation 4 speakers.

In the meantime, here are a couple of pictures of the AR-303a speakers in my Blue Room AV system.

I now have installed my new Penny Mustard barrel chairs. These sit a bit higher, bringing my ears right up to the tweeter axis of 39 inches above the floor. The tweeter axis is definitely the best sounding listening height for these speakers.

IMG_0428.jpgIMG_0426.jpg
 
Last edited:
I have installed my second set of Acoustic Research AR-303a speakers in my Audio Room, replacing for now the Watkins Generation 4 and Graham Audio LS8/1 speakers I previously used in that room. I am keeping both of those sets of speakers in case I decide later to return to using them in my Audio Room or elsewhere. For example, I may decide to use the Watkins in my desktop computer audio system as replacements for the speakers built into my Apple Studio Display monitor.

My Audio Room set up of the AR-303a's has not revealed any surprises as to the 303a's sound. Everything I said above about their sound in my Blue Room A/V system has been confirmed in my Audio Room.

However, my extensive use of acoustic foam room dampening in my Audio Room (perhaps assisted by other set-up tweaks) has further ameliorated the AR-303a's shortcomings in portraying depth of field. The stereo is quite entertaining/engrossing and the changes from one recording to the next are as well revealed by these speakers as with any others I've had in this room.

No, the degree of envelopment is not quite up to the standards of the Watkins or Graham speakers, but it's not far behind. And perhaps there is a slight loss of certain types of small-sound detail (e.g., fingernails on piano keys). But these shortcomings are counterbalanced by the overall advantages the ARs bring to the table:

1. Once Room Perfect is applied, the 303a's are truly full-range in the bass and have great bass power, punch, and detail.

2. The AR power range (100 to 300 Hz) and the lower midrange warmth region above that are full-throated before any equalization, in contrast to many speakers, including the Watkins and even the Grahams if used on less-than ideal stands.

3. The ARs have superior ability to sound both "generous" or "big hearted" in the lower frequencies and unstrained at higher volumes even on big orchestral music and bass-demanding pop and electronic music.

4. They have a natural warmth to their presentation without residual midbass bloat and no midrange or high frequency anomalies. I really appreciate their "natural sounding" presentation and the degree of the high frequency roll off evidenced by the graph of their frequency response shown in post #3. I am admittedly a sucker for the vintage AR speakers sound, My first speakers were AR-4x's and I've owned several of the other classic models (AR-2ax, AR-5, AR-3a) as vintage speakers over more recent years. The 303a updates the sound of the classic models enough to meet the demands of modern digitally recorded music while adding just the right amount of treble to sound more "open" without ever moving into the too bright/too edgy camp. Instead of the "comfortable old shoe" sound of the classic ARs, the 303a's to me are analogous to "optimally broken in shoes"--comfortable but still at the peak of their performance.

5. Factors 1. - 4. makes these speakers the most enjoyable I've owned for home listening to my very eclectic tastes in reproduced music and makes less necessary any Lyngdorf Voicing equalization from recording to recording.

6. They sound majestic on large-scale classical works and can boogie with other types of music. I often find myself involuntarily moving to the rhythm of the music, both in response to their truly excellent frequency balance and their wonderfully nimble portrayal of recorded dynamic shifts from moment to moment.

7. Compared to the Graham LS8/1, the AR low end to me seems even more natural sounding and definitely more extended, at least once Room Perfect is applied. The AR mids are not quite as detailed, but are detailed enough and at least equally natural sounding. The AR highs are a bit more mellow, which is very nice.

I estimate that the AR's are about 5 dB less sensitive than the Watkins and about 3 dB less sensitive than the Grahams. That's the difference in the Lyngdorf volume control settings necessary to achieve similar subjective playback SPL in this room with equivalent speaker and listening positions. The Lyngdorf still has plenty of power to drive the ARs as loud as I care to listen, however. The nominal 6-ohm impedance of the ARs means that the Lyngdorf is able to deliver about 300 watts per channel into their load, as opposed to 200 watts into 8-ohm loads.

As with my Blue Room use of the ARs, the application of Room Perfect to my Audio Room set-up produced relatively small sonic changes compared to its application to the Watkins speakers. The sonic difference is even smaller in my Audio Room than in my Blue Room when quickly comparing the Focus and Bypass settings of Room Perfect. In my Audio Room, on much material I have to listen carefully to hear the changes Room Perfect makes in such quick A/B comparisons. Still, these changes are significant with longer-term listening and all to the good. As in my Blue Room set-up, bass is yet further extended, residual mid-bass boom is eliminated, the midrange is no longer a bit forward, and the imaging and staging are both more focused and expansive.

A few set-up notes:

1. As in my Audio Room set-up, while I have the AR speakers toed in 30 degrees, I have moved my listening chair forward so that the inboard tweeter/midrange drivers point directly at their respective ears.

2. To both ease the movement of the speaker stands on the carpet (the stool legs tend to sink deeply into carpet) and to level the speakers in this room (old house, slightly un-level floor), I use one to three 2-inch diameter circular thick felt pads under each of the four legs of the speaker stand stools.

3. This increases the tweeter height of the speakers by about half an inch compared to the same stands used in my Blue Room without such careful leveling. The tweeters in this Audio Room set-up are 39.5 inches above the carpet. To get my ears up to that height, I'm using a combination of a Target equipment stand wood shelf and a thick gel cushion under the regular seat cushion of my Drexel chair.

4. The center of each speaker's woofer is 24.5" above the carpet, 36 7/16 inches from the near side wall, and 59 inches from the wall behind the speakers.

If your tastes in speaker tonal balance are similar to mine, I highly recommend these speakers to you. For such listeners, I think it would be well worth your effort to hunt for a pair of the AR-303a speakers on the used market. There are usually pairs on offer somewhere in the world for asking prices of $2,000 a pair and sometimes much less. These are a true bargain for the price. The Watkins, when available new, are about $3,500 a pair and the Grahams are $10,000 or more a pair.

A caution: Ideally you want to pick these up rather than have them shipped since the original shipping materials are rarely kept for 30-year-old speakers. While only two cubic feet each, at 54 pounds each, they are quite heavy for their size and thus can be easily damaged if the shipper does not take adequate care in padding and packaging these for shipping. The fact that the speakers have an exposed tweeter dome which sticks out in front of the front baffle makes the tweeter especially vulnerable to shipping damage. That said, I've managed to have two pairs of the 303a speakers shipped to me in non-OEM packaging and still arrive in fully functional and cosmetically acceptable condition. My most recently purchased pair arrived looking even better than the seller's pictures indicated.
 
Last edited:
Both the Stereophile and The Absolute Sound reviews are of the original AR-303 speaker. That is the model whose frequency response graph run by REG I have reproduced in post #3 above.

As far as I know, the only differences between the 303 and the 303a are the use of different binding posts and the ability of the 303a to be biwired or biamped since it has two sets of posts. As I understand it from online comments, the 303a quickly replaced the 303 and sold more than the 303.

I suppose only Ken Kantor, the designer, would know for sure if there are any other differences. I've not seen this issue discussed online. If anyone has any information on this, please chime in. Thanks.
 
Another review of the AR-303 was published starting at page 33 of the June 1995 issue of Stereo Review. The review was by Julian Hirsch and compared the sound of the 303 to the vintage AR-3a. See the following link for the full text of the review:

chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://www.worldradiohistory.com/Archive-All-Audio/Archive-HiFI-Stereo/90s/Stereo-Review-1995-06.pdf

If that direct link doesn't work for you, go to this link, pick the June 1995 issue of Stereo Review and then scroll down to page 33.
 
Last edited:
The issue of proper measured bass level for speakers is a matter of some controversy. It does not help that many reviewers who publish speaker frequency response measurements do not include frequencies below about 200 Hz since this area is greatly affected by the listening room and speaker placement within that room. Most subjective reviews, including those in The Absolute Sound, almost never include frequency response measurements.

That leaves Stereophile as one of the very few, if not the only, historical record of USA speaker frequency response measurements going back decades. Bless John Atkinson for having used largely consistent measurement methods over all that time. His measurements of speakers over the decades are thus directly comparable.

Still, a misunderstanding of JA's bass measurements persisted for many years despite JA's repeated explanation that his near-field (as in an inch or less from the woofer cone) bass measurements result in an apparent bass rise and the need to artfully "splice in" the bass measurements with those taken from less near field at higher frequencies. JA measures the bass in the very near field to help eliminate the room effects from his bass measurements, at least part of the theory behind this being that the direct sound from the cone will be much louder compared to the room modes, reducing the apparent contribution of the room modes to the measured result.

Remember that JA started measuring speakers decades before the Klippel technology existed. The Klippel allows measurements to be taken in an ordinary reverberant room and still closely approximate the measurements which would result inside a very large anechoic chamber. Such large chambers are very hard to come by. Certainly most audio magazines have no access to such. Soundstage is an exception and some of their speaker reviews are supported by measurements made inside the NRC's anechoic chamber. However, there is some doubt as to whether even that chamber is large enough to properly measure the 20 to 40 Hz output of speakers.

In the past few years, a few online audio sites, notably Erin's Audio Corner, have used Klippel measurements to show anechoic-like bass measurements which seem to match well with real-room results in the bass for properly placed speakers. Erin has thus far reviewed well over 200 speakers, but most of these are newer models and most of those are lower-priced and smaller than the type of speakers audiophiles aspire to own. Also, seemingly he doesn't use much classical music when subjectively evaluating speakers. Still, Erin's subjective comments are a breath of fresh air in that he expresses a preference for warm low end over thin low end and clearly dislikes speakers with "hot" high ends. The industry needs more such reviewing voices to stop the insanity of speakers seemingly designed with thin low end and hot highs.

Here is a link to a fairly lengthy thread from some 20 years back in the Classic Speaker Pages forum arguing about the worth of the AR-303 speakers in view of the way their bass and overall response measured in JA's Stereophile review. It is evident that the reality of the bass bump caused by JA's bass measurement method is largely misunderstood by the posters.

Those of us who, like me, want our home systems to emulate the weight and warmth of the low end and the overall tonal balance we hear from large orchestras from good seats in good concert halls have realized that if you are looking for speakers which will do this with a lot of commercially recorded classical music without adding a lot of equalization, you want speakers which Stereophile's JA measurements show to have a midbass and lower midrange excess and which roll off gently in the top octaves.

If you agree with my tastes in speaker tonal balance, flat is not the way you want the bass or top treble to look in JA's measurements. Worse is speakers where the mid and low bass seem to roll off in level. Still worse are such speakers where the bass slopes down in level and the highs above 2 to 5 kHz are higher in level (peaked up, in other words) than the rest of the range. If you have my sensibilities, you run away from such speakers and don't even remotely consider acquiring them.

It's odd that JA himself seems not to have a preference for speakers with warm, full low end. He tends to criticize speakers which measure the way the AR-303 does. He also criticized the Stirling Broadcast LS3/6 for its low-end rise. But most classical music fans who hear that speaker (and the
Graham LS8/1 and the AR-303/303a) love that balance. All I can say is that I apparently differ from JA when it comes to my preferred tonal balance in speakers, especially for classical music.

Sure, you can try using electronic equalization to conform the response of a speaker you otherwise like to your preferences. But pushing the low end response of a speaker up in level can be tricky. You risk running out of both amplifier power and low frequency excursion capability of the woofer cone. At a minimum you will raise the distortion level of the boosted bass frequencies. Equalization is probably safe enough with most speakers if you listen at low to quite moderate levels--say, never exceeding steady state levels of 75 to 80 dB. But if you want louder, beware.

In defense of the modern trend in speaker design which leans toward thin bass and hot highs, I will say that apparently this formula sells. Many audiophiles and music lovers simply hate the resonant midbass boom which can occur all too easily in room if your speakers have a generous low end. They also seem to like the extra "detail" seemingly added by boosting the top octave or two.

To this I respond that, yes, midbass bloat sounds awful, but it is the easiest thing to negate via electronic equalization. There is virtually no downside to cutting off the midbass peak (usually between 60 and 80 Hz and due to the floor/ceiling room mode( with electronic EQ: the sound improves subectively not only because you get rid of the boom, but also because you gain amp headroom and reduce speaker distortion by reducing bass power demands on the amp and reducing woofer excursions for any given bass level and overall SPL. Cutting off bass peaks is where electronic equalization can be used all to the good. You don't want to throw the baby out with the bathwater by lowering the entire bass range. What you want is a fairly smooth and generous bass and lower midrange, minus the usual midbass peak. You can also safely use electronic EQ to tame a top end rise by cutting the level of the highs, but that is trickier because EQ can't attack the subjective effect of excessive mid- and high-frequency reflections off the listening room walls.

If you agree with me about your tonal balance preferences in speakers and are looking for something more pleasing than what you currently own, it's simple enough to look at JA's Stereophile measurements to see the shape of the frequency response measurements. As I said, look for a rise in the lower midrange and midbass (even the low bass if the speaker is capable down there) and look for a gentle roll off from 5 kHz upward.

I'll give you a hint: you won't find many such measured responses in all the annals of Stereophile's measurements. Yes, they do exist, but probably no more than a very few percent of all the speakers Stereophile has measured over the decades meet or even approximate these criteria and even fewer are so balanced among current offerings. Thus, if you want to start with speakers whose inherent response is "in the ballpark" for natural-sounding reproduction of classical music, you really don't have that many to choose from.

A couple other examples of spakers which meet or approximate these criteria include the Dutch & Dutch 8c, and the Alon IV,. I also believe that some Harbeth models like the M40.2, the M30, and M30.2, older Spendor models like the SP1/2 and SP 100, and some other speakers manufactured in the BBC tradition will either meet or approach these criteria.
 
Last edited:
I'm often asked why I change speakers so often. I suppose the reasons lie in both the speakers and in me.

No man-made objects are perfect. That includes the speakers used in an audio reproduction system. All speakers have flaws in terms of how they reproduce sound if one is comparing the reproduction to the absolute sound of live acoustic instruments in a favorable concert hall space. And of all the equipment making up an audio reproduction system, the flaws in the speakers are the most obvious and most differentiated from each other. That is why no two speakers ever sound really close to the same when substituted into the same system when nothing else is changed.

Some listeners can ignore speaker flaws and live happily and contentedly for many years with the same speakers. That is not me. Sooner or later, and often sooner, I begin to fixate on the flaws in any particular speaker which are revealed and seemingly more and more emphasized in long-term listening and begin searching for another speaker which will have less of that flaw while not having other obviously worse flaws.

One would hope that this search would lead ever upward/forward, ever more closely approximating perfection. But I've been in this hobby long enough now to know that the reality is that the path is not straight, but meandering and even sometimes circular. Often I just end up trading one set of flaws for another. Sometimes I get hung up on pet audio theories, pursuing what I think I have finally identified as "the" most important aspect(s) of speaker design.

Thus, for example, with the Watkins Generation 4, I convinced myself that the simplicity of the single-capacitor first-order crossover, time alignment, tweeter damping, and special bass extension in a small box just had to be the ultimate approach. And the sonic results really are truly excellent as speakers go. But then I encounter a speaker like the AR-303a which has none of these technical elements and yet quickly convinces me that, in ways currently important to my ear/brain/emotions, something superior to the Watkins is here in this 30-year-old design.

But the quest is not as hopeless as those last paragraphs may make it seem. The quest can be redefined. With more decades of experience, comes both greater self knowledge and the ability to make audio choices which better meet my changing goals for my home audio systems. I no longer seek the ultimate loudspeaker. I seek what satisfies for now. Certainly there is a hope and expectation that experience brings a maturity which transforms my goals in ways which seek authentic and enjoyable reproduction as opposed to eye candy, current fads, and unneeded conspicuous consumption. But there is also the knowledge that the maturation process continues and that my goals can and probably will continue to be refined and changed.

Right now, in the AR-303a I have no trouble sacrificing the ultimate level of detail, clarity, and 3-D immersiveness which seems to come from a design like the Watkins if what I gain is truly full-range bass extension in a reasonably sized box without the complexities of subwoofers added, coupled with a very program-friendly overall tonal balance which, once Room Perfect does its thing, seems to make program-to-program equalization mostly unnecessary. Together with the tonal balance, the dynamics, both micro and macro, and perhaps other factors, get to me at an emotional level, causing me to involuntarily move and sway with the music. The ARs are enjoyable and not thin sounding at low levels and respond gracefully and big-heartedly at higher levels. The sound is generous and warm yet still very open and immersive, just as in a real-life acoustic concert. The stereo is focused more than usual between the speakers, but on simply miked recordings and ones which intentionally manipulate phase, bursts forth with great expansiveness and envelopment. Like the Watkins, the highs are never abrasive, but in the AR seem in better balance with the lower frequencies. Instruments and voices sound as much like the real thing in terms of overall tonal balance as any speakers I've ever used. And these ARs seem to sound quite similar in two different rooms, one of which is extensively treated acoustically and fronted with a variety of electrical tweaks and the other of which has no acoustic treatment and uses no special electrical tweaks.

No, the AR-303a's are not perfect speakers. But for my current tastes in music reproduction, they are a wonderful compromise at a bargain price.
 
Last edited:
  • Like
Reactions: hopkins
So, how do my current three speaker systems, the AR-303a, the Watkins Generation 4, and the Graham Audio LS8/1 stack up sonically?

In terms of bass extension, the AR goes the deepest, comparable to wonderful subwoofers, like the Audio Kinesis Swarm which I had in this same room a few years ago in combination with the Stirling Broadcast LS3/6 speakers. The Graham is second most extended, seemingly flat down to almost 30 Hz. The Watkins is flat to 40 Hz and thus misses most of the bottom octave.

The Graham is by far the most particular about its positioning. It MUST be used on low, open frame stands in order to have that 30-ish Hz bass extension and full, warm midbass and lower midrange. And it MUST be listened to no higher than the lower tweeter (about 34 inches above the floor) in order to avoid adding tizz to the highs, which means an extremely low chair, one which is increasingly difficult for me to get in and out of.

The Watkins is the most open and enveloping presentation, together with the most focused images. The Graham is second, and the AR is third. But these three speakers are surprisingly close in these aspects. On average, however, the AR stereo seems more focused/confined to the area between the speakers (REG's The Absolute Sound review specifically mentions this aspect of its reproduction) so with a lot of material the ARs might seem a more distant third. But feed the ARs simply miked classical material or any recording where the engineers have manipulated phase and the AR's imaging/staging "wakes up," shockingly expanding to be a very close third indeed. In this sense the AR's imaging/staging changes more from recording to recording than the other two and might thus be regarded as more accurately reproducing its input.

The Watkins has the cleanest high frequencies of any speaker I've ever owned, but there is just a bit too much HF relative to the rest of the range. The Graham is not quite as clean, but the highs are in better balance and thus never obnoxious as long as you listen on the axis of the lower tweeter. The AR highs are, as REG's review noted, a bit on the "rough" side, but combined with the just-right roll-off of the highs, never sound obnoxious and are in close-to-perfect balance with the rest of the range. When listened to on the optimum axis--vertically aligned with the tweeter axis, with the tweeters aimed at your ears, and from the near-ish field--this roughness is really minimized and not really noticeable unless one compares it directly to something better in this respect, like the Watkins.

In the all-important midrange the Graham often give me the feeling of "yes, that's what that instrument or voice sounds like live and unamplified." Still, the AR gives me close to that same feeling. The Watkins sounds the clearest and cleanest, while sacrificing just a tiny bit of midrange tonal realism compared to the other two.

Once Room Perfect does its magic, from the mid-bass up through the lower midrange warmth region, the AR is superbly full, generous, and big hearted without any excesses. It sounds majestic and powerful on large scale works, and at higher volumes to an extent that the Graham only hints at by comparison. The ARs give up nothing to the Graham or Watkins in terms of bass detail and the ability to clearly differentiate the different tones of a walking bass line. By direct comparison to the others, the Watkins pulls up the rear in this mid-bass through lower midrange area, but is vastly improved in this aspect when Room Perfect does its magic. The Watkins is equal in bass detail to the other two, but falls behind in the sense of weight, majesty, and power.

The AR is by far the most immune to its surroundings. As I stated in my last post, these speakers sound remarkably similar in my two rooms, one of which has extensive acoustical padding and the other of which has no such intentional treatment. The other two speakers, even in my larger audio room, simply MUST have extensive acoustical foam padding to sound their best. Otherwise, side wall reflections can easily add an obnoxious bite/edge to the upper mids and highs. I suspect that the evenness of the AR's off-axis response below 6 kHz or so, as shown in the Stereophile review's measurements, is the reason for this.

In terms of sensitivity and SPL capability, the AR is probably about 5 dB less sensitive than the Watkins, with the Graham being about 3 dB less sensitive than the Watkins. The Watkins thus requires less amplifier power to reach a given SPL in any given situation. While all three speakers will cleanly play as loud as I care to listen in my relatively small audio room, the AR sounds most subjectively comfortable at higher listening levels, followed by the Graham, and then the Watkins.

All three speakers seem unusually micro and macro dynamically responsive compared to other speakers I've owned.

Finally, for some reason or combination of reasons, the ARs seem to induce a stronger emotional connection, causing me to more often bob and sway involuntarily to the music than the other two.
 
Last edited:

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu