Frequency response of classic AR's, especially the 4x

[Guest postjob62]

I have never seen published anywhere FR numbers for any of the old single-digit AR speakers. I assume AR must have had a policy of not quoting such things in advertising, brochures, etc. I could have missed it, but I've not seen figures in the library either.

I wonder if anyone has these figures or any idea what some of them may be (or may have been when new). I am personally interested in the 4x, 2ax and 3a.

I'm most interested in the 4x. This is a speaker, perhaps because of it's diminutive size, that just never fails to astonish me with it's sound every time I listen. I just wonder what numbers are attached to the sounds I so enjoy.

Regards,

Ed

[dynaco_dan]

Hi Ed

Each speaker usually had it's own pamphlet with photo, physical description, graph or graphs showing frequency response of each driver.

Some had even red colour as part of the graph.

There was also full product catalogs, not large in physical size but maybe 20 pages in depth, which had a photo of each product and a brief description. There was several versions as I remember.

I don't remember if the library has any, but do please check it out.

In my museum mess, I do have brochures of some speakers, if I feel that I can dig them up, I will try.

They would be all pre 1974, the ones I do have.

You started a good topic, Ed, thank you.

Have a great day.

Knowledge is Neverending.

I have never seen published anywhere FR numbers for any of

>the old single-digit AR speakers. I assume AR must have had a

>policy of not quoting such things in advertising, brochures,

>etc. I could have missed it, but I've not seen figures in the

>library either.

>

>I wonder if anyone has these figures or any idea what some of

>them may be (or may have been when new). I am personally

>interested in the 4x, 2ax and 3a.

>

>I'm most interested in the 4x. This is a speaker, perhaps

>because of it's diminutive size, that just never fails to

>astonish me with it's sound every time I listen. I just wonder

>what numbers are attached to the sounds I so enjoy.

>

>Regards,

>

>Ed

[soundminded]

AR published on axis frequency response curves for every one of their speakers. They were all virtually ruler flat except in the bass where the frequency the speaker started rolling off at depended on the size of the woofer. The more money you spent and the larger the woofer, the more bass you got.

If I had to say what three factors I think are the most important in the subjective perception of a high fidelity sound reproduction system, in order of importance I'd say they are frequency response, frequency response, and frequency response (paraphrasing my real estate friends.) Having said that, I consider the issue of frequency response to be far more complicated than the simple notion we are usually accostomed to reading about. If it were simple, all speakers having the nearly the same frequency response would sound nearly the same. Clearly this is not the case.

[tysontom]

>I have never seen published anywhere FR numbers for any of

>the old single-digit AR speakers. I assume AR must have had a

>policy of not quoting such things in advertising, brochures,

>etc. I could have missed it, but I've not seen figures in the

>library either.

>

>I'm most interested in the 4x. This is a speaker, perhaps

>because of it's diminutive size, that just never fails to

>astonish me with it's sound every time I listen. I just wonder

>what numbers are attached to the sounds I so enjoy.

Ed, it was AR policy to publish a full family of response curves and performance data for each and every speaker they made up through at least the mid-1970s. This information was available to anyone, simply "for the asking." AR did not just throw out numbers; they published curves with measurements made on instrument calibration traceable to the US Bureau of Standards. For example, AR published both on- and off-axis anechoic frequency-response curves for the AR-4x as well as harmonic-distortion data. Later data included impedance and acoustic-power-response data as well. To my knowledge, AR was the *only* "mainstream" loudspeaker manufacturer to publish this level of information for a loudspeaker during that time period.

The AR-4x has a very flat measured response, both on- and off-axis. It also has smooth, uniform response off axis, and AR considered dispersion to be extremely important. Flat response on-axis is one thing; uniform response off-axis is quite another.

http://www.classicspeakerpages.net/dc/user_files/579.jpg

AR-4x Tweeter anechoic response curves -- 1965

http://www.classicspeakerpages.net/dc/user_files/581.jpg

AR-4x Woofer free-field response curves -- 1965

Hope this information helps.

--Tom Tyson

[Steve F]

AR did not publish system frequency response numbers, probably because they felt that they could be misleading or confusing, depending on how they were derived.

AR did specify in good detail how they measured their individual drivers; they then "presumed on good faith" that the consumer could extrapolate the system response based on the individual responses of the drivers in that system.

A good-faith assumtion on AR's part or a cynical, deceptive approach designed to avoid the nasty, messy imperfections of a system FR, complete with edge diffraction, crossover-induced anomolies, etc.? My gut feeling is a little of both, but heavily-weighted to the "good-faith" side of the ledger. I don't think that AR was a deceptive company and if their preferred approach of just publishing drivers' responses looked a little cleaner than full system measurements, then that was just icing on the cake.

AR did publish full systems' energy response curves, which they felt were more indicative of sound quality and reproducing accuracy in the home than raw FR curves.

As for the 4x, it would be instructive to seek out Julian Hirsch's review in Stereo Review, around 1965 or 66. (I lost that issue due to some water in my basement about 12 years ago.) One of his comments--and I remember it almost verbatim-- was "This {frequency response} would be remarkable for any speaker but is almost unheard of for a speaker of this size and price...We know of no competitively-priced speaker that can compare with it."

I remember that the response was something like +/- 2 dB from about 300Hz to 15,000 Hz (which was the upper limit of his measuring microphone).

Interesting: This is also virtually the same FR as he measured from the AR-7 some seven years later--although the 7's much wider dispersion and therefore more uniform energy response gave it a far "crisper, snapier" high end than the 4x.

I guess AR's presumed suspicion that system FR measurements could be mis-interpreted was well-founded!

Steve F.

[tysontom]

>AR did not publish system frequency response numbers,

>probably because they felt that they could be misleading or

>confusing, depending on how they were derived.

>

>AR did specify in good detail how they measured their

>individual drivers; they then "presumed on good faith" that

>the consumer could extrapolate the system response based on

>the individual responses of the drivers in that system.

>

>A good-faith assumtion on AR's part or a cynical, deceptive

>approach designed to avoid the nasty, messy imperfections of a

>system FR, complete with edge diffraction, crossover-induced

>anomolies, etc.? My gut feeling is a little of both, but

>heavily-weighted to the "good-faith" side of the ledger. I

>don't think that AR was a deceptive company and if their

>preferred approach of just publishing drivers' responses

>looked a little cleaner than full system measurements, then

>that was just icing on the cake.

>

>AR did publish full systems' energy response curves, which

>they felt were more indicative of sound quality and

>reproducing accuracy in the home than raw FR curves.

Steve F is right that AR did not generally publish “system” frequency-response numbers for their speakers, simply because end-users could incorrectly interpret these numbers as the actual “in-room” performance of those speakers. The vagaries of frequency-response numbers have always been questioned, of course, unless the testing was carefully documented, controlled and performed according to standards in a manner that could be duplicated by others trained in this sort of measurement. However, I don’t believe that AR was ever intentionally deceptive in this methodology, or accused of being cynical or deceptive by any of the reputable reviewers and critics of the era. Others with “an ax to grind,” might have felt differently.

AR did, however, publish response curves and associated numbers for the AR-1 with the first brochure and also showed a system “on-axis” response (but no numbers) of the AR-3a with the 1967 brochure. The AR-3a on-axis curves, measured anechoically, were performed with the speaker grill panel and the grill decorative-edge molding removed. The reason for this was that standard measurement techniques dictated that the measurement microphone be positioned approximately 1-2 meters in front of the speaker to be measured; as such, it would be difficult *not* to record diffraction from the edge molding and grill. However, by moving the microphone position just a few inches in any direction would cause a change in the frequency of each dip and peak associated with the diffraction, such that once one was back in the reverberant field, those effects were largely swamped by the reflected sound from all angles.

AR felt that the individual anechoic response of each driver was vital to understand each driver’s on- and off-axis response and, especially, uniformity. This would isolate such things as transient distortion, “ringing,” and other aberrations in each driver. The acoustic-power response of the “system” therefore best described what the speaker was doing in a normal listening environment, and was measured by taking a average of its response at all angles, and this measurement technique was greatly enhanced with the introduction of AR’s Reverberant Test Chamber, and later the semi-reverberant chambers.

--Tom Tyson

[soundminded]

"Steve F is right that AR did not generally publish “system” frequency-response numbers for their speakers, simply because end-users could incorrectly interpret these numbers as the actual “in-room” performance of those speakers."

Then why publish the response of the individual drivers at all? Their only value even today is for those who want to design their own speaker systems and since AR did not market these drivers individually to experimenters, that was not their purpose. If the implication was that you could infer the system response by splicing the response of the indivudual drivers together, that would be very misleading. The system response would depend on the crossover network design, the physical displacement of the drivers from each other, and the differences in propagation time delay between the electrical input and mechanical output from one driver to the next. These would determine the exact nature and locaton of the reinforcements and cancellations cause by the interference patterns of the drivers operating together in the crossover regions. This shows the bogus claims of "coherency" in so called time aligned speakers. Ironically, while manufacturers in the 1970s pretended to worry over the difference in propagation time delays between drivers, they gave up completely on the concerns of manufacturers in the 1950s who addressed space displacement differences by building coaxial and triaxial speakers. It is only when you have both simultaneously that you have any chance of reproducing a mechanical wavefront analagous to the electrical input and then only on axis.