Review of AR-303 by Julian Hirsch in June of 1995

[Howard Ferstler]

If one cares to, they can go to the "kitchen" section and check on one of my posts dealing with the "Roy Allison interview in 1992 issue of The Audio Critic, David Ranada was the interviewer." Some here may be interested, because that particular post by me (which shows up on page 3 of the series) has an attached review of the AR-303 loudspeaker done by Julian Hirsch back in 1995 (Stereo Review magazine). What may interest some people here even more is that Hirsch does A/B listening and measuring comparisons between the 303 and the AR-3a. I would post the attachement here, but I have only a limited amount of attachment space left.

Howard Ferstler

[genek]

I would post the attachement here, but I have only a limited amount of attachment space left.

Hirsch_review_of_AR_303__June__1995.pdf

If anyone feels the need to start a food fight over this, go to the Kitchen. I'll open a thread for one there.

[Steve F]

Hard to believe that anybody would want to get into a row about this particular review. In any case, the attached review is already there in the kitchen for all to discuss.

Howard Ferstler

This review has always been somewhat of a mystery to me. JH says some things that just don't correspond with either reality or his own measurements.

First, he says the 303 has 3-5 dB more treble energy than the 3a above 6kHz. That seems about right for a modern ff-cooled tweeter, and that's about the degree of treble boost that most listeners would agree is needed to make the 3a sound "right" by today's standards. A nice, solid 5dB treble boost, and then all the 3a's other nice attributes--the smoothness, the wide dispersion (or cd, whatever you wish to call it), the deep, solid bass, etc.--all come into balance nicely and it's a very satisfying speaker.

But then he says something absolutely incomprehensible--he says the 3a was "distinctly brighter" than the 303, and only when the 3a's controls were turned down to 'Nor' were the two speakers' balances about equal.

How can a speaker with 3-5dB MORE treble energy than the 3a be DULLER at the same time?

We've all been listening to 3a's for 40 years. We know their sound inside and out. We know their strengths, we know their weaknesses, we know it all. But there's one thing we can agree on--the 3a has NEVER been accused of being "brighter" than any other speaker, ever! Smoother, deeper, wider-dispersing, less efficient, less colored, yes. But ''brighter''? Never.

And his measurements indicate exactly the opposite.

In addition, JH got all the original 3a x-o frequencies totally confused and wrong, so, unfortunately, I have to question the competence of this entire review. Just doesn't quite add up. It's bothered me for 15 years. JH's confusing statements relative to his own measurements don't mean that the conclusion that the 303 is an outstanding speaker is wrong, just as it doesn't mean that the 3a before it was a bad speaker. But it does mean that JH was both inconsistent and somewhat factually inaccurate in this review.

Steve F.

[genek]

This review has always been somewhat of a mystery to me. JH says some things that just don't correspond with either reality or his own measurements.

First, he says the 303 has 3-5 dB more treble energy than the 3a above 6kHz. That seems about right for a modern ff-cooled tweeter, and that's about the degree of treble boost that most listeners would agree is needed to make the 3a sound "right" by today's standards. A nice, solid 5dB treble boost, and then all the 3a's other nice attributes--the smoothness, the wide dispersion (or cd, whatever you wish to call it), the deep, solid bass, etc.--all come into balance nicely and it's a very satisfying speaker.

But then he says something absolutely incomprehensible--he says the 3a was "distinctly brighter" than the 303, and only when the 3a's controls were turned down to 'Nor' were the two speakers' balances about equal.

How can a speaker with 3-5dB MORE treble energy than the 3a be DULLER at the same time?

We've all been listening to 3a's for 40 years. We know their sound inside and out. We know their strengths, we know their weaknesses, we know it all. But there's one thing we can agree on--the 3a has NEVER been accused of being "brighter" than any other speaker, ever! Smoother, deeper, wider-dispersing, less efficient, less colored, yes. But ''brighter''? Never.

And his measurements indicate exactly the opposite.

According to the article, Hirsch measured the 3a with the HF and MR pots turned full up, which was not the AR-recommended setting for regular listening when the speakers were made.

I think Hirsch's listening impressions tell us is that the 3a's sound with both pots at max is brighter than the 303's sound, which is not adjustable, because the 303's crossover was designed to "voice" the speaker's sound to be similar to the 3a with more extended HF, but at some 3a pot setting other than both pots full up. But we'll have to let Ken address that.

Have you ever heard any of the big, floor-standing console speakers from the 50's and early 60's that used full-range drivers only or woofers with relatively large diameter cone tweeters? The 3a could definitely produce "brighter" sound than those at its highest pot settings.

[Steve F]

According to the article, Hirsch measured the 3a with the HF and MR pots turned full up, which was not the AR-recommended setting for regular listening when the speakers were made.

I think Hirsch's listening impressions tell us is that the 3a's sound with both pots at max is brighter than the 303's sound, which is not adjustable, because the 303's crossover was designed to "voice" the speaker's sound to be similar to the 3a with more extended HF, but at some 3a pot setting other than both pots full up. But we'll have to let Ken address that.

Have you ever heard any of the big, floor-standing console speakers from the 50's and early 60's that used full-range drivers only or woofers with relatively large diameter cone tweeters? The 3a could definitely produce "brighter" sound than those at its highest pot settings.

My point is that the 3a was never accused of being 'brighter' than any of its later-day late-60's/early '70's competitors from Advent, KLH, EPI, JBL, Dynaco, etc. Never.

The confusion over Hirsch's comments stem from the 303 having 3-5 dB more measured output above 6k and his seemingly-contradictory comments that the 3a was "distinctly brighter." Again, the notion that the 3a--which wasn't 'brighter' than any speaker in its time frame--could be 'brighter' than the modern ff-cooled 303 (even though the 3a had measurably significantly less HF output) is confusing, that's all.

Toss in Hirsch's inexplicably-inaccurate recollection of the 3a's x-o frequencies (couldn't he have just referred back to his own test report on the 3a, where he made a big deal about the 3-3a changed x-o's, or asked Tom T, who supplied the 3a's, or consulted old AR lit, which he undoubtedly still either has or has access to, etc.), and it adds up to a review that just isn't JH's best effort, not by a long shot.

Steve F.

[genek]

My point is that the 3a was never accused of being 'brighter' than any of its later-day late-60's/early '70's competitors from Advent, KLH, EPI, JBL, Dynaco, etc. Never.

The confusion over Hirsch's comments stem from the 303 having 3-5 dB more measured output above 6k and his seemingly-contradictory comments that the 3a was "distinctly brighter." Again, the notion that the 3a--which wasn't 'brighter' than any speaker in its time frame--could be 'brighter' than the modern ff-cooled 303 (even though the 3a had measurably significantly less HF output) is confusing, that's all.

I don't think anyone would ever have accused the 303 of being "brighter" than any of the speakers you mention, either. It most likely wasn't designed to be. But neither the 3a nor the 303 were being compared to any "competitors," they were only being compared to each other.

The numbers-vs-listening statements suggest that Hirsch may have based his "3a sounds brighter than 303" conclusion on the outputs of the two models' midranges with the 3a on full rather than that of their tweeters.

The big question in my mind is, why Hirsch measured the 3a with both pots at max and didn't recheck it at the settings he concluded were the closest match for the 303 in a review that was intended to make direct comparisons between the two models. Along with his error about the 3a's crossover, it does suggest that he was rushing this session.

[Zilch]

We did the Hirsch 303 review before here. Put "Hrisch" in the Search engine, to find it.

I find no discrepancy between JA and JH regarding the brightness issue....

[genek]

We did the Hirsch 303 review before here. Put "Hrisch" in the Search engine, to find it.

You sure it wasn't on another site? No "Hrisch"'s on CSP that I know of.

[Zilch]

It WILL find that post.

Spelled correctly it finds our discussion over a year ago in at least three threads, including this little gem:

http://www.classicspeakerpages.net/IP.Boar...amp;#entry79784

More misinfo here:

http://db.audioasylum.com/cgi/m.mpl?forum=...ight=lst&r=

[genek]

Ah, I see. I thought you were saying to search on the misspelled name "Hrisch" to find previous discussions.

[genek]

Also, assuming the 3a tweeter disperses better than that of the 303 (I am not sure about this, but just suppose) that additional power out to the sides, if the room were reflective, would make the speaker sound brighter than a speaker with similar power input to the room, but with a more beamy treble.

I don't need to suppose, I know from listening that the 303 was a more "beamy" speaker than a 1975 2ax. It wasn't a huge difference, but enough to notice if I got up out of my chair and walked around the room. Since the 2ax has a cone mid and the 303 a dome mid, I wouldn't expect that it was anything but the difference in the tweeters that accounted for this.

[Steve F]

Attached is a copy of a one-meter curve run on both systems, with them overlapping for clarity. The sitnals were adjusted in terms of level to overlap below 100 Hz. The darker line is the AR-3a; the lighter one the 303. I have no idea where the level controls were set on the 3a. However, if they were set to flat it looks like the user would want to back off the midrange level to get the spectral balance to be similar to that of the 303. If they were all the way up the midrange control would definitely best be backed off. The treble performance of the 3a looks fine, even with that moderate dip above 15 kHz or so.

Note the bass drop. This was, I believe a free-field measurement, so the systems were 2 pi down to about 100 Hz, with the rolloff below that indicative of the transition to a 4-pi space at those lower frequencies. The upward ramping below 40 Hz indicates that maybe a nearby room boundary was regenerating a 2-pi field. I have no idea of the environment here, so any number of factors might be impacting the lower range.

A one-meter curve is not indicative of much if we are talking about the way a speaker will be perceived in a normal room at a normal listening distance, but for comparison purposes, such as we have here with so similar a pair of speakers, it can be helpful.

Howard Ferstler

To me, this isn't a relevant comparison. I've always normalized curves so their midrange levels--the perceived 'loudness' levels--were matched. I'd nomalize to their 500-1000 Hz levels were as close as possible, and then see what you had.

BY JH's own sens measurements, the 303 was less sens than the 3a, so that has to be adjusted for. It looks like this curve is simply showing the 3a's measured 5dB greater sens, but this comparison doesn't display their actual spectral differences as we'd hear them.

Steve F.

[Zilch]

AR3a mid looks to be set at "Max," and there's the 1.8 kHz hole (suckout, notch, dip, swale) I also found....

[genek]

AR3a mid looks to be set at "Max," and there's the 1.8 kHz hole (suckout, notch, dip, swale) I also found....

The only time AR recommended a max setting on the 3a's MR pot was in combination with a max setting on the HF pot and +2dB on amplifier treble for "flat" response. Without the accompanying amplifier treble boost, with a max setting on the tweeter control I use a MR setting about halfway between the white dot and max (roughly on the "S" in "increase," the same knob position for max HF).