High Efficiency Speaker Asylum

Posted by Christopher Witmer (A) on March 07, 2007 at 19:07:12

I recently revisited Sound Mates, the wonderful new high-end audio shop that is preparing to open soon in Tokyo's Akihabara district, where for the first time in over a year, I was able to do comparative listening between Feastrex's Dimension 5nf and Dimension 5 loudspeakers. The former uses the D5nf driver with the spherical "Naturflux" magnetic circuit, while the latter uses the D5 driver with a traditional cylindrical magnetic circuit. Both drivers have 16-ohm voice coils; the D5nf's magnetic circuit has an alnico 5-7 magnet that weighs about 1 lb. (about 450g) while the D5's magnetic circuit has an alnico 5-7 magnet that weighs about 5.5 lb. (about 2500g). The former has a Qts of about 0.6 and the latter a Qts of about 0.2. There is about a 2dB difference in sensitivity as well, 95dB vs. 97dB. The D5nf units have been running in for close to one month, while the D5 units are just three days old, and the difference in aging surely had a big effect on the sound. The glues used in these drivers take a very long time to fully cure because Feastrex foregoes the use of catalysts that would harden the glues very rapidly. This lack of total hardness in the beginning allows the voice coil and subcone to move slightly out of phase with each other which produces a loss of output in the highest frequencies which gradually improves as the glue hardens up. I couldn't notice this at all; whatever phase cancellation is taking place is probably outside the range of my ears' sensitivity. What I did notice, however, is that in a few passages of deep electric bass and rock drum passages where I did a direct comparison between the two drivers, the less expensive D5nf driver had slightly better definition. Part of this slight superiority may have been due to the D5nf's Naturflux circuit, but more likely it was due to the D5s' lack of sufficient run-in. (They still had very little run time on them when I started listening.)

I expect that if the drivers had been of similar ages the D5 would have been clearly superior even in that respect, because except for a slight disadvantage in the definition of the deep bass, it had a clear performance edge over the D5nf in all other areas. I think the difference between the respective sounds of the two drivers can be compared to the difference between high-end analog and digital. With high-end analog one generally has a feeling of greater warmth in the presentation of the music, and a sensation that there is a greater amount of detail in the signal. On the other hand, digital seems to be quieter, although that can also be negatively described as lacking in warmth or lacking in detail vis-a-vis analog. This is just a caricature of analog and digital, of course, but I want to say something similar about these two units in describing their relative sound differences. The D5nf seems "quieter" than the D5, and the D5 seems "warmer" than the D5nf, but it is probably related to their respective abilities to resolve detail. On the other hand, I distinctly remember the field coil version of the driver (which I last heard about one year ago), despite its even higher sensitivity and ability to resolve fine detail than the D5, nevertheless sounds "quieter" (and more "analytical") than the D5 -- so this difference in sound may be, in part, a difference in characteristic sound signatures of the materials used in the motors. Different metals and different magnets impart different flavors to the sound, it seems. This makes sense to my mind because the behavior of the cone is the result of the electromagnetic field's interaction with the coil, and different materials used in different ways will produce different electromagnetic fields for the coil to move in. (I'm sure there is even more to it than that, because even though in an ideal world the frame and motor would not be acting as a percussion instrument adding a characteristic sound to the mix, surely every sound transducer is doing that to a small degree.)

Both drivers will pull you right into the music and are very involving. I spent five hours listening mostly to music that I normally dislike and at volume levels normally guaranteed to give you a headache, and when it was all over I felt like I could have gone on for another five hours. The other people who were listening with me also completely lost track of the time. Warning: don't start listening to these speakers if you have work to get done, because any listening session has a good chance of running into overtime. I spent most of my time listening to the D5nf, with the D5 used primarily for purposes of comparison. I had already spent a lot of time listening to the D5 a few weeks ago.

I also decided to forego the opportunity to listen to the Stephens Trusonic 15-inch coaxial and the three-way Western Electric field coil system that I had listened to alongside the D5 last time. There were three reasons for this. 1) Lack of time -- it would have turned into an all-night affair, but I had to work the next morning; 2) the problems with the bass that I had reported with all three speakers last time were now greatly improved, thanks to the excellent room treatments that Sound Mates' Mr. Asayama had subsequently implemented, and therefore I felt no need to compare the bass of the tiny Feastrex units to that of the larger speakers; and 3) the three-way WE field coil system will be undergoing a major upgrade in a few weeks time, and by then the D5 will be better broken in, so a more meaningful rematch can be enjoyed in a few more weeks.

The five major aims of this listening session were:
1) To get a feel for the relative differences between the D5 and the D5nf (summarized above)
2) To see how well these speakers perform in the bass region
3) To see how well these speakers perform when run by an R120 SE amplifier with about 1 watt per channel output, in a room measuring 2.4 meters high by 6 meters wide by 9.5 meters deep, with the speakers about one meter from the rear wall and about three meters from the listener
4) To see how well these speakers perform when powered to near their breaking point by a 200WPC digital amplifier
5) To see how well these speakers can play popular music, especially rock music.

First, I needed to get my bearings, so I played some pieces with which I am extremely familiar. I listen mostly to music from the classical period and earlier, and to very little music from the mid-19th century and later. I am most familiar with the music of Bach Collegium Japan, led by Masaaki Suzuki, and of Orchestra Libera Classica (OLC), led by Masaaki's brother Hidemi. I have heard all these pieces performed by them in concert, and in the case of OLC I was actually present when the CDs were recorded, so they are very useful for my purposes. Here are the CDs that I used to orient myself:

Bach Collegium Japan: J.S. Bach, Ouvertures (Orchestral Suites) BWV 1066-1069
Hidemi Suzuki et al.: Domenico Gabrielli, L'opera completa per violoncello
Hidemi Suzuki: J.S. Bach, 6 Suites a Violoncello Solo senza Basso
Hidemi Suzuki et al.: Luigi Boccherini, Quintets for Strings
Hidemi Suzuki & Orch. Libera Classica: Haydn Cello Concertos 1 & 2 / Symphony 13

The above recordings were listened to with both the D5 and the D5nf powered by the tiny R120 SE amplifier, and were adequate to confirm that both speakers can play the above music flawlessly and with more than enough volume to fill the room. The bass was extremely precise and seemed to my ears to have a flat in-room response at the listening position, although we would soon be moving into more challenging territory as regards the bass . . .

Next we decided to start listening to music that I almost never listen to, and much of which I had never heard before, and at volume levels much higher than I ever use. Listed in alphabetical order, we listened to:

Beatles: Let It Be
Brigitte Fontaine, Areski, Art Ensemble of Chicago: Comme a la radio
Bryan Ferry: Taxi
Capercaillie: Capercaillie
Clementine: a suivre/The Very Best of Clementine
David Bowie: Ziggie Stardust
Edgar Winter's White Trash
Jacques Loussier Trio: high-quality bootleg recording of a concert in Tokyo
John Martyn: No Little Boy
k. d. lang: Ingenue
Led Zepplin: Untitled

At less than 1 watt of output, the R120 SE amplifier was able to play these pieces much louder than I would listen to anything in my own home, although not necessarily as loud as some people would like to hear them. You could still talk with the person sitting next to you if you raised your voice considerably and maintained eye contact while talking. We never maxed out the amplifier's volume . . . I think we went to the 2 o'clock position on the volume.

But it was clear we needed a more powerful amplifier if we really wanted to crank things up, to push the drivers to their limits. (Actually, it was only the D5nf that we pushed to its limits. The D5 could play just as loud without as much stress, due to its higher sensitivity, and we really did not want to play any louder, as we were already giving ourselves more than enough of a beating with the D5nf. At the point where the D5nf's knees start to buckle, I think one is already past the point where most people who like to listen loud are sated. If someone is not satisfied with that kind of volume in their regular listening, they have already probably lost a lot of hearing ability.)

We hooked the D5nf up to the 200W/8 ohm digital amplifier and had the volume turned up to about 11 o'clock while listening to John Martyn's No Little Boy when the combination of the powerful bass line and the kick drum started to cause the spider to bottom out on the back stroke. No harm was done to the speaker but we quickly lowered the volume just a tad to prevent that. There is no *audible* distortion whatsoever up to the point where suddenly that starts to happen. I'm a poor judge of the frequency content of a kick drum, but it must have been somewhere between 80Hz and 40Hz. It was mighty loud, and definitely "visceral" bass but it was not what one might call "dorm party" "disco" or "rock concert" loudness -- the kind of volume where one needs to almost shout into the other person's ear to be heard. I think the D5nf (and of course the D5) are entirely capable of such levels of loudness but only if the bass -- say below 100Hz -- is given over to a subwoofer.

In a typical listening room, I would say there is no need to use a subwoofer on 99% of classical music and acoustic jazz. A subwoofer that cuts in at around 50Hz might be good for pipe organ music. A subwoofer that cuts in at around 100Hz is definitely needed to achieve extremely high volume levels with rock and other popular music that has a lot of electronically enhanced deep bass content.

There is no doubt in my mind that both the D5 and D5nf can do a decent job with rock and other electronically manipulated electronic music, with the caveat that some people might feel the need for augmentation of the bass on occasion. For classical and acoustic jazz, it his hard for me to imagine speakers that would be overall more satisfying than these.

For me, by far, the most interesting musical find of the evening was the last piece we listened to -- Ravi Shankar, Andre Previn & London Philharmonic Orchestra: Concerto for Sitar & Orchestra, recorded live. I knew Ravi Shankar had done some of these kinds of collaborations but I had never heard any before this. It certainly is an unusual piece but it works very well and is great fun, especially if you love challenging rhythms and unusual percussion. The LP recording that we listened to was outstanding. It is obviously a multi-mike affair and probably mixed down in a studio from multiple tracks, but to me that is hardly objectionable at all because there is very little use of harmony and counterpoint; rather, the melody travels back and forth among the various individual instruments or groups of instruments in succession, and one rarely has the whole orchestra playing tutti. So an approach to recording that would offend many one-point microphone purists seems to work perfectly well here. The only thing I can't say for sure is whether the volume of the sitar has been boosted so that it sounds better on the recording than when listened to in live performance, or whether the sitar was perhaps even amplified for the benefit of the listening audience. But in any case, the record makes a pretty good audio test record while also being real music (really good music), which is more than can be said for many audio test records. The fast transients of the unusual percussion instruments, the many different types of instruments, wide dynamic range, large scale of the orchestra and ambience of the hall all came across really well, and even after having been impressed all night long I was still amazed that such little speakers could produce such a huge sense of scale. Regrettably for this final recording we were hooked up only to the Dimension 5 speakers and I did not have a chance to hear how the D5nf units sounded with it. (Somebody finally noticed that we were past 2:00 A.M. so we packed up and got out of there fast.) While not wishing to knock them in the least, I doubt the D5nf would have been able to fully convey the huge sense of scale that came across through the Dimension 5 loudspeakers.

Before I close, I would be remiss if I did not remark once again about the amazing things Mr. Asayama of Sound Mates is doing with D/A converters. He really seems to be at the cutting edge, and the results are impressive indeed. After it opens, Sound Mates should definitely be a destination to visit for any audiophile in the Tokyo area. His tremendous knowledge, huge collection of music, excellent room acoustics, fine selection of equipment and friendly demeanor make it a great pleasure to visit.

• My earlier review   (Open in New Window)