109 replies to this topic

[speaker dave]

The BA SubSat 6 and 7 were a challenge because BA actually cared about maintaining a smooth, continuous FR from the bass module to the satellites, without having the 'acoustic hole' at 140-200 Hz and without having the bass module go up too high in frequency to make it too localizable. The SS6 and SS7 sats were 4" 2-way designs that responded down cleanly to 140 Hz and they sounded great. The 6 used dual 6.5" woofers and the 7 used dual 7" woofers. It could really woof! They didn't sell at Bose-like levels, but were very successful for BA and were probably the best-engineered 3-piece systems around in the early '90's. I know they absolutely blew away the Cambridge Soundworks systems for FR smoothness, bass extension and power-handling. Blew them away. The AM-5 was not real 'hi fi' compared to them.


Steve F.

So the AM-5 was the worst of the three and the best seller by far. This just shows the strong imbalance between marketing/advertising vs. product quality. (something to think about as we enter the Presidential Campaign).

At Snell we tried to stay away from the loudest theater demos and play HT selections with more of a music focus. My favorites were Kundun, James Tailor Live at the Beacon, Black and White Nights (Roy Orbison), Sessions on West 54th St. and a few others.

Regards,
David S.

[Gerry S]

So the AM-5 was the worst of the three and the best seller by far. This just shows the strong imbalance between marketing/advertising vs. product quality. (something to think about as we enter the Presidential Campaign).

At Snell we tried to stay away from the loudest theater demos and play HT selections with more of a music focus. My favorites were Kundun, James Tailor Live at the Beacon, Black and White Nights (Roy Orbison), Sessions on West 54th St. and a few others.

Regards,
David S.

I think BA tried for the "middle ground" / "upwardly mobile" music lovers with SOME "disposable income". A "budget" Snell can exceed the total system cost of a typical BA target customer.

As for Bose, I remember having a "heated argument" with some of their engineers when BA gave a "factory tour" when they entered my lab. Later, Andy Kotsatos said to me: "Heard you had a "pissing contest" (or something to that effect). I don't know whether that was his "admonishment" to me for "non-professional behavior". Just goes to show how passionate we can get regarding "design philosophy".

In hindsight, the AM-5 "set the bar", so to speak. As an "audiophile company " they are IMO certainly NOT. But, they DID inspire OTHER companies to produce "better" products. In that regard, I must begrudgingly "tip my hat" to them.

Regarding demos: As a designer with an admittedly "eclectic" musical taste, I "forced myself" to listen to stuff I normally would not. For the HD9 submitted for this "shootout" (ah; back to topic), I used Run Dmc’s "rap hit" "Walk This Way" (later made "famous" by Aerosmith)! I normally don’t like "rap". But, one can never guess who does when customers go into the showroom. So I used Run Dmc’s version to "torture test" the HD 9 in the bass frequencies to insure the HD 9 could "do it justice" and SURVIVE while doing it .

As "crude" and/or "vulgar" as this song can be to people with more "refined taste", I admit to liking it.

[Gerry S]

Preliminarty BA HD-9 FR measurement is in !

Thought I'd post it before I accidentially deleted it....senior moments ! I'm sure final results for ALL speakers in this "shoot-out" will be published on the Boston Audio Society at some point on the "BAS Speaker"

Attached Files

•  gerry_sheeto.pdf   1.08MB   71 downloads

[Pete B]

Too bad an EPI 100 or 110 was not in the shoot out:
http://www.classicsp...t=0

I'd rank it as one of the best economy classics of reasonable size.
I prefer the Genesis II but it is quite large.

[Gerry S]

FR data: interpreting the graphs is tricky. I don’t know the specifics of how these graphs were obtained. Nor, do I have the original data I made myself. But here’s what I CAN tell you (from looking at them). The actual crossover point is about 2.8 kHz (that sharp notch in the FR curve). That notched curve was probably made at a point ½ way along the vertical dimension of the baffle. The curve with the flattest FR should be on the tweeter axis (or above).

This speaker was "optimized" based on the following assumptions: The user would place this speaker on the floor, near room boundaries. For those who wanted a "sweet spot" for near-field listening (like me), I placed them on a 15" stand so that tweeter positioning was at ear level for a seated listener. I listened to them with the speakers slightly pulled away from room boundaries for best "imaging".

The curves shown were made with the grill off. With the grill on, there would be a very slight attenuation of the higher frequencies (barely audible). I THINK my original curves were made with a General Radio sine wave sweep generator (forgot the model #), captured with an Ivie IE30 analyzer, using a B & K mike. All data was written on a GR "strip chart recorder". If the original data I obtained were indeed using the "old-fashioned" method, I’m amazed at how "flat" the "new curves" made with the Keith & Larson test equipment is.

But the proof is in the listening. So far, they still sound pretty good to me. And these only retailed for $340/pair when new.

[ironlake]

regarding the latter cu tests did the same people do the tests and did their ears still hear the way they did on the early tests. I know I do not hear like i did when I was 25 years old. I could hear up to 13,000 hz then, now I am lucky to hear 8000 hz.

[Carlspeak]

FR data: interpreting the graphs is tricky. I don’t know the specifics of how these graphs were obtained. Nor, do I have the original data I made myself. But here’s what I CAN tell you (from looking at them). The actual crossover point is about 2.8 kHz (that sharp notch in the FR curve). That notched curve was probably made at a point ½ way along the vertical dimension of the baffle. The curve with the flattest FR should be on the tweeter axis (or above).

This speaker was "optimized" based on the following assumptions: The user would place this speaker on the floor, near room boundaries. For those who wanted a "sweet spot" for near-field listening (like me), I placed them on a 15" stand so that tweeter positioning was at ear level for a seated listener. I listened to them with the speakers slightly pulled away from room boundaries for best "imaging".

The curves shown were made with the grill off. With the grill on, there would be a very slight attenuation of the higher frequencies (barely audible). I THINK my original curves were made with a General Radio sine wave sweep generator (forgot the model #), captured with an Ivie IE30 analyzer, using a B & K mike. All data was written on a GR "strip chart recorder". If the original data I obtained were indeed using the "old-fashioned" method, I’m amazed at how "flat" the "new curves" made with the Keith & Larson test equipment is.

But the proof is in the listening. So far, they still sound pretty good to me. And these only retailed for $340/pair when new.

Yes, the proof was indeed in the listening to the EPI 110's. They sounded as good as the measurements said they would.

For the measurements, I used a calibrated ECM 8000 mic and HolmImpulse software. The mic was 1 m from the speaker which was raised off the floor about the height of an AR3a (which it was sitting on ), Obviously the far field (1 M) measurements were gated so the data is only valid down to about 300 hz. IIRC, my measurements were done will the grille off.

[Gerry S]

Yes, the proof was indeed in the listening to the EPI 110's. They sounded as good as the measurements said they would.

For the measurements, I used a calibrated ECM 8000 mic and HolmImpulse software. The mic was 1 m from the speaker which was raised off the floor about the height of an AR3a (which it was sitting on ), Obviously the far field (1 M) measurements were gated so the data is only valid down to about 300 hz. IIRC, my measurements were done will the grille off.

regarding the latter cu tests did the same people do the tests and did their ears still hear the way they did on the early tests. I know I do not hear like i did when I was 25 years old. I could hear up to 13,000 hz then, now I am lucky to hear 8000 hz.

I have for the last 12 years (since leaving BA) STRONGLY resisted the urge to buy test equipment so I can "improve" the speakers I developed so they can "measure" as best as possible. I resist for several reasons. ALL my speakers are "good enough" when I voiced them originally... exhausting amount of data generated for even a "low priced" pair like the HD9's I'm listening to now. In the end, my choice for the final x-o was always based on LISTENING to as many CD's as possible and from as many genres as possible. While doing so, I always wondered about the age, sex and hearing cababilities of the various mix engineers who "voiced" the masters in THEIR rooms/studios.

In a perfect world, everyone would hear identically regardless. In the real world, the differences are staggering to the point where I don't want to think about it. I don't know what my hearing is like today from a test meaurement standpoint; undoubtedly a HF roll-off. When I do have "guest" over to listen, most are "impressed" so I guess my hearing is still "on track". When listening alone, it's rare I encounter a recording that sounds "terrible". If I can't "correct it" with tone controls (which I rarely do), then I consider it a "bad" recording.

Within reason, most speakers made by reputable companys used by "normal people" sound "good enough". For just any ONE pair among the dozens of models I have, I can significantly change their "sonic signature" just by varying speaker/ listener placement. The speaker/ room/ listener/ mix engineer permutations are infinite. If I change the x-o design so it's "correct" for a given situation, then it will be "wrong" for all others.

And that's why I don't bother anymore. If there is a "holy grail" of loudspeakers, there are countless numbers of them.. an oxymoron IMO.

[Gerry S]

Brand forums are added in response to interest from members. There are plenty of other manufacturers of the "Classic New England" speaker genre that don't have a forum here. All that means is that there hasn't been much (or any) interest expressed in having one.

I'm surprised DBX wasn't on this site (correct me if I'm wrong). Specifically, the DBX Soundfield line.

[Gerry S]

Probably because it's just as easy for his competitors to notice that they're losing sales to him and release "me-too" products. it's done all the time. Being regarded as an industry leader and trendsetter may be of as much or more value in sales as having a 3 month jump on your competitors in finding out what sells.

There's no reason why a listening panel A/B'ing a series of reproduction products against a live source could not be a usable measure of "realism" or "accuracy" from listener POV.

Impractical (impossible?) to consistently/reliably to achieve this. At the most basic level, try doing this for ONE listening session, using ONE "live source", through a SINGLE loudspeaker (i.e mono, not stereo). Assume sensitivity of the two competing speaker brands are "normalized" in the midrange when "A/B' ing" (we all know the louder of two speakers will sound "better"), done in a "typical" (but pretty large) listening/living room (yours, mine, whoever).

Assume listeners/judges in the panel are BLINDFOLDED. Why blindfolds (not a "acoustically transparent curtain")? VISUAL considerations of the "listening room" can skew the panel's perception of realism or accuracy of what's being "heard". If skeptical about this skewing effect, try eating quality/expensive "banquet food" at a posh/huge restaurant setting. That SAME food "tasted" at home won't provide the same total sensory experience.

Despite listening to a "live source" as the reference to which speaker A or B is being compared too, "replicating" that same source is impossible because it IS "live". Add another speaker channel (stereo), or another live peformer, or change the enviroment; things will get out of hand very quickly. For any test to be scientifically valid, it must be repeatable.