the home of the turntable

Wideband Speakers

amplifiers, receivers and loudspeakers

Wideband Speakers

Postby desktop » 20 Jul 2011 17:48

Avole asked me to describe the loudspeaker systems I listen to. For many years I designed loudspeaker lines based on market research, until I worked for WED (Walter Elias Disney) Enterprises. There I designed loudspeaker systems based on either the required size like the Tokyo Disneyland Tiki Birds, or based on maximum performance like for the French Pavilion Theater at EPCOT in Florida. That was where the money was available to research the hard and fast rules of design for natural, realistic, wideband sound reproduced by loudspeakers.

I believe that we have all participated in the "blind listening test vs listening knowing the identity of speaker models" threads here on VE. Most of my personal opinions about how to take sound from a vinyl disc and turn it into either very realistic sound, or sound that most closely mirrors the musician's/producer's concept of what the music should sound like, come from blind listening tests done on more than 10,000 listeners during 1977 and 1978. Those results stayed consistent although we used 5 locations at Los Angeles, David, California, Seattle, Madison, Wisconsin and at Georgia Tech. Using various "Golden Ears" for their opinions in their own listening environments gave response results all over the place, and so were practically useless since they were so biased.

I also liked the results of BBC's testing done to produce the LS3/5a loudspeaker system. One of the best tests ever is to have a person speaking a few sentences sitting in a chair very near where the test loudspeaker will be set up. Record these sentences with a B&K4130 microphone (I believe this is the 1/2 inch capsule, condenser microphone from B&K, although I haven't used my pair in years now) and then play the recording back through the loudspeaker you want to test. The loudspeaker must sound like the voice. The only way to really achieve this is to have a small loudspeaker enclosure producing the sounds from 200Hz and up because the vocal sound waves will need to reflect off the same surfaces whether they are coming from the test subject's head, or from the speaker enclosure. The LS3/5a can sound very much like a real person. I have a pair of these loudspeakers, but I don't really have a place for them now. I'm sure that a computer-speaker version of the LS3/5a would astound people, and would only need to be shielded for people using cathode computer monitors. I intend to use mine with an LED computer monitor I'm installing.

I accumulated boxes and boxes of component speakers and used some of them to build my own loudspeakers as well as using Martin Logan and Quad electrostatic loudspeakers. In order to avoid the "laser beam" effect that is normal for most loudspeakers in the higher frequencies I use multiples and arrays. Also for loudspeaker components producing sound above 5KHz (assuming small size for the component loudspeaker) designing series and parallel xover circuits is not acoustically problematic too much, because one does not consider damping to be an important factor.

So for the extreme high frequencies above 12KHz-15KHz depending on the design I only use two types of "super tweeters": titanium domes and my best choice are tiny thin film tweeters. There is a third choice though you would need to be handy with circuit design, but surprisingly enough, slot drive piezo electric tweeters can be made to perform with low distortion and high loudness if the piezo device is designed into a resonant circuit (as the capacitor) with a resonant peak at 26KHz-27KHz so its effective range is 20KHz-30KHz + or - 5db. There are a few tricks to doing this so let me know if you are going to try.

I like using the Tannoy Studio Monitor in my home with a home built subwoofer, because the Tannoy monitor is relatively compact and their tweeter is only down -3db at 40KHz in its "laser beam" range. This points out where most tweeters are at. A standard 2 way loudspeaker with a 1 inch (25.4 mm) tweeter (domed with a 1 inch voice coil) only extends properly out to about 12KHz and then as the phase shifts and the cone breaks-up and decouples from the surround, a smaller and smaller surface area produces sound, and so in order to keep the upper frequencies "flat" past about 14KHz the dispersion pattern gets narrower and narrower. If we ever discuss studio monitors used in the 60s, 70s, 80s and 90s, we'd see that they suffered from beaming as well, so that isn't an extremely negative characteristic.

But to produce sound that I prefer, I use an array of 16 tiny (3/4 inch driven surface) thin film tweeters. Tiny film tweeters of almost any kind can produce high frequency extended output past 30KHz. The resulting upper frequency output is "quasi-beamy" but does not have the odd off-axis lobes of horns or tweeters mounted in baffles, so an array can be built easily on a damped open frame. Individually the thin film tweeters are not playing loud, and they do not operate down to frequencies that would cause them to couple. So in the 18KHz to 28KHz range this kind of tweeter is pretty effective if you make the array produce a 270 degree L/R and Up/Down pattern. It looks a bit like a small oval grid shaped like an oversized American football.

I get slightly better imaging using either the Tannoy monitors or electrostatics, but less beaming with my "maximum" system using separate enclosures of inert materials for each component loudspeaker. This makes my "maximum" system the most realistic playing back vocals I know personally. The system using the tweeter array has more good listening positions, whereas the Tannoys and electrostatics have only one good listening position. The electrostatics all have non-flat frequency response curves. The Tannoys have a disappointingly characteristic 10% distortion in the vocal bandpass because of the compression driver + horn, plus beaming in various frequency ranges due to a changing dispersion pattern shape. The Tannoys also have a conical pattern in many frequencies which creates reflective issues. My large home built loudspeaker systems have minor variations in response in various listening positions, but the directivity remains pretty constant above 200Hz (and below 200Hz it is 180 degrees down to 16 Hz).

So nothing is perfect. Each system sounds a bit different from the others (electrostatics being the most unique). But each has its uses. These are the systems I use to try to listen back to the phono cartridge's audio characteristic "sound". I do this when I am trying to get the most out of any vinyl I've cleaned and listen to, to hear what it is possible to hear.

To audition new records I often use a pair of B&O broadcast monitors. This speaker is so old the little labels have fallen off and the connectors are on the end of a heavy wire/cable, but the components are visible. There is an 8 inch woofer with an 8 inch passive radiator that acts like a highly tuned port. There is a 2 1/2 inch domed midrange which I believe was made by SEAS. Then there is a very tiny 1/2 dome tweeter. These 2 loudspeakers are a mirror-image pair. Every component loudspeaker is heavily sealed into the baffle board. The box is finished with teak, and is completely sealed.

I use Yamaha NS 10 loudspeakers for some new record auditions plus a small KEF loudspeaker (I believe it is the C5 with a 5 inch bextrene woofer and a small titanium dome tweeter) plus a home built low frequency enclosure with xover so I can do passive network bi-amping. I also use a pair of JBL 4311 monitors, and sometimes a pair of Dynaco A-25s.

I am always experimenting with new loudspeaker designs. I don't get free samples any more but I have purchased many new loudspeakers using woven carbon fiber cloth molded into a perfect circular dome (concave) for woofers and midranges, and large domed (convex) mid-tweeters. This material is quite inert and has little sound of its own. While bextrene also has little sound of its own, it is not comparable in stiffness to the newest heat-molded carbon fiber cloths. Sadly I no longer have the time to go around town to a bunch of audio retailers to ask if I can audition all their newest loudspeakers. So any of the readers here at VE may have a great loudspeaker system that I haven't heard, especially if it was made during the last 10 years (post 2000).

JBL has stamped out a number of titanium cones with applied foam damping on the back and these work very well in ultra-small enclosures (5 inch woofers), but mine are mounted on a steel frame bolted to a rear wall and I use this model (which I will have to supply at a later date when I access them again) as the rear channels of my CD-4 test system.

So this is what I use to listen to vinyl with.
senior member
senior member
Posts: 368
Joined: 10 May 2007 00:42

Postby Officerpup » 20 Jul 2011 18:13


Fascinating in all sorts of ways and inspires me to experiment more.

Any chance of some photo's or illustrations - a picture's worth a thousand words, 'n'all that there'n't'other :)
Posts: 91
Joined: 17 Jun 2011 14:45
Location: Bath

Postby Rap » 21 Jul 2011 00:42

Interesting read and approach to different mediums. Thank you for taking the time to write it desktop. :brit:
User avatar
senior member
senior member
Posts: 459
Joined: 07 Jul 2008 22:47
Location: Brussels


Postby hulbar » 21 Jul 2011 01:27

I am restoring a pair of Jbl Lancing L101 speakers right now. I have just reconed the le-14a woofers and I am now looking for good diaphragms for the 175 dlh:s. When working like the originals such speakers can be amazing.

junior member
junior member
Posts: 5
Joined: 20 Jul 2011 16:21
Location: sweden

Postby RED_One » 21 Jul 2011 22:12

Thanks for the comprehensive post.
very interesting especially the elliptical array - can't find anything on thin film tweeters though have you got some manuf's names to hone down the search? preferably with minimum front to back depth or do they need an open baffle?

How would these be wired up in an array?

How is amplifier load on these blighters?

Apart from the above the Yamaha Ear shaped drivers have intrigued me - have you heard these or have any info on them??
senior member
senior member
Posts: 588
Joined: 17 Dec 2006 22:54
Location: UK

Postby Alec124c41 » 22 Jul 2011 05:52

I have Yamaha elephant ears, waiting for future experimentation. They are quite efficient, and have little excursion.
The large radiating surface gives a very open sound, rather like planar speakers. Very pleasant.
The high-frequency beaming, and roll-off particularly with the larger ones, necessitated a separate tweeter. The plastic foam is easily damaged, and the vinyl coating on some of them detaches after 30 years.

Keep them spinning.
User avatar
vinyl addict
vinyl addict
Posts: 14936
Images: 56
Joined: 28 Oct 2002 04:24
Location: Toronto, Canada


Postby XenMaster » 22 Jul 2011 16:41

Alec124c41 wrote:I have Yamaha elephant ears, waiting for future experimentation. They are quite efficient, and have little excursion.
The large radiating surface gives a very open sound, rather like planar speakers. Very pleasant.
The high-frequency beaming, and roll-off particularly with the larger ones, necessitated a separate tweeter. The plastic foam is easily damaged, and the vinyl coating on some of them detaches after 30 years.


I got to the bit about the coating detaching before I worked out you weren't talking about your own analogue aural inputs!

Posts: 73
Joined: 28 May 2011 23:38
Location: UK

Postby desktop » 22 Jul 2011 21:39

To reply to Red_One I am surprised that there are no thin film tweeters being sold as parts. I checked eBay, but perhaps there are other sources. A thin film tweeter is simply a plastic film or sheet with circuit conductors applied to it, This encompasses everything from a Magnaplanar with wires applied to a big plastic sheet, the ESS Heil AMT tweeter where the plastic film is folded up, or the more standard thin film tweeters used on some infinity and Cerwin Vega loudspeakers from the 70s and 80s.

I wish I knew more about the 40 thin film tweeters that sat in an unmarked carton in my garage for 15-20 years. They were about the size of a very large coin and about 1/3 inch (8mm) thick. Individually they were useless with about 3 watts max power handling and with moving diaphragm surfaces only about 1 inch in diameter. They wouldn't play loud. Thus the need for so many to cover the range from 18KHz to ~30KHz. The upside is that these tweeters had flattish frequency response in the upper regions I wanted, but the downside is that if thin film tweeters have too much energy sent to them, they heat up the plastic which melts and stays deformed. They also beam allot so there is only one good listening position unless they are the Heil design or are curved. or if you use many of them.

So for extremely extended frequency response these tweeters with next to no mass can reproduce the highest frequencies pretty well, but they are fragile and with low excursion potential, they won't play loud. The Infinity and CV models were the size of a small woman's hand. The ones I have are much smaller and I believe their original purpose was to be used in an audiophile boom box. I am not sure where they came from.

The following link explains the idea ... technology

The folded version of the thin film tweeter seems to be available in that form from 3 companies. See below ... r-speakers

The difficulty with these devices is to make sure the beaming patterns align properly when using multiple units. There is also the power handling issue. Ribbon tweeters are much larger in general and they have the same problems so they are usually horn mounted for directivity (the Decca ribbon tweeters were often used with the first Quad electrostatic loudspeaker as a tweeter). But of the 3 methods described by me 1) tiny domes of light material like titanium, 2) piezos with special circuits to create a transducer that really only works between 15KHz and 30KHz, and 3) thin film tweeters, I prefer thin film tweeters in a small size and then just use many of them.
senior member
senior member
Posts: 368
Joined: 10 May 2007 00:42

Postby desktop » 22 Jul 2011 23:15

I am replying to comments by Avole from the Grado Carts thread in the Cartridges section.

Avole said (Start)

"My CDs go to 22khz.

Also, the CD standard was defined first, and manufacturers use this to ensure their players meet the requirement. The CD is not tailored to meet the requirements of the player, it is the other way round.

Their are physical limitations to the human ear's ability to hear above and below certain levels. Do you dispute this?

No-one has golden ears, but everyone can learn the skills required, since the term describes the ability to critically evaluate recordings. If you're interested in doing a course, Google is your friend."


I am unaware of any CD with actual 22KHz material on it, but list one or two of the titles and I will ask someone to do a narrowband test to see if there is any program material in the 21KHz to 22 KHz range on the disc. I've been selling off my spectrum analyzers and other devices I could use and I don't have an audio CD player that will produce anything in the 21.5KHz region. I could burn a CD with a 21.5 KHz tone on it but I have no audio CD player that works in those upper frequencies. I am, of course, interested in the titles of discs with such extreme high frequencies on them.

I agree that the CD is now the prisoner of specs developed in the 70s and 80s. Even with the best digital filtering there are still limits to the capabilities of CD to portray a realistic audio image if the program material has lots of energy above 17KHz (IMO). This is why the SACD was developed. Real SACD output only needs to go to ~35KHz to produce a very realistic image of reality. Many times it is the high frequency energy in musical program that defines the precise location of instruments and voices. I've never heard a CD player capable of clear representations of high frequency material and I didn't hear one at this Jan's CES. It might be helpful if someone who is in the business of making CD players was to comment on the ability of CD players to play back 22KHz tones.

While the ear hears the lowest frequencies in combination with our bodies, the limitations in the highest frequencies are poorly defined. Turning one's head with one ear towards the sound source is common all the way back to our homo erectus days or perhaps beyond to the days when stereo hearing won the day with animals. The limitations in the highest frequencies seem to be due to the flexibility of the ear drum diaphragm, and having a clear uninfected path from the ear drum to the outer ear. The AES found numerous test subjects able to clearly detect sound waves above 50KHz. I can't speak for them because my hearing isn't that good.

I disagree completely about real "Golden Ears" which I list akin to perfect pitch. Many people can be trained in some form of audio knowledge, and they may understand the complete physics of acoustics (Al Bodine of Bodine Soundrive is one of the best examples I know of), but that doesn't mean they have Golden Ears. Most real golden-earred people I've met don't even know they have them and most of those people are not acousticians or audio engineers. I've met many people who can perfectly locate sound source directions, because they were born wired to be able to do it. Many performers are fussy about their equipment because they really do have golden ears, but they've never trained as audio engineers.

This is why I was always so frustrated when some great engineers of audio electronics tried to design loudspeakers and the results were poor. These engineers could do the math but they didn't have great ears for hearing detail, and connecting sounds, or distortions to problems in component loudspeaker matching or crossover circuit design. Having extended range hearing is one of those characteristics I notice is common in real golden earred people. These people and I consider myself one of them, are extremely fussy about their ears and their hearing. But that's because they can hear, AND recognize any tiny change in their hearing, or in what they are hearing.

I find that most audiophile writers may be decent writers, but few have true golden ears. Most would scoff at any hearing test results that didn't show a perfectly normal and healthy result from 50Hz to 21KHz. But they won't be tested because they don't want to have any real results that might conflict with their mental views of themselves.

I've heard numerous reviewers discuss loudspeakers, when I knew by what their listening rooms looked like, that they couldn't hear the differences in the terms they used. Sometimes these golden ears hit a bulls eye and gained great status, but then failed completely on the next product they tested.

In the highly competitive 1970s and 1980s the chief executives of loudspeaker manufacturers often chose 4-6 employees and one or two retailers to audition potential new models. For these execs, those people had golden ears. I was surprised when I was part of these groups, that opinions were often similar except in the range above 12KHz. If there was a high frequency problem in the product, those with the true golden ears tried to explain what they heard (the bad thing) so a test tech could try to use test equipment to sort it out.

Only at Disney did we develop systems with equipment first and then listening. There were just too many uses to fulfill. But we all listened as we went along and real stinkers were eliminated. The designs I did for EPCOT seem to have held up for nearly 30 years, especially the theater systems for the French Pavillion.

I never claimed to have golden ears during my audio biz years. But many manufacturing execs put that into my job description. We were also the people sent out to various booths at CES to hear if something really interesting was being made by a competitor.

Some golden earred people might be piano teachers who can "hear" the slightest hesitation in playing a certain note at a certain time. But if you are defining "Golden Ears" as people who can listen to loudspeakers and define all of the product's characteristics properly, then I would say there were very few I met in 25 years working full time in the audio business. I met quite a few normal-people listeners with golden ears who had no training at all, but could ask a question about a real "sound" they heard when a certain loudspeaker was being played. Whatever the reason, like placement, or improper manufacturing, or damage, etc. these "civilians" could shock you with their ability to perceive the tiniest details in the sound, because they were listening with golden albeit untrained ears.

One of the factors that influenced retail sales floor staff turnover (IMO) was how golden were a salesperson's ears. You can teach pleasant behavior, and sale techniques. You can educate sales people to know store policies and product specifications. The only thing you can't do is "golden-up" their ears. Sales people can learn about speakers that sound good with bassy material, or voices etc. But many fell by the wayside because their ears could not pick out the subtle differences in sound.

Nothing kills credibility for a sales person in retail audio more than a customer who doesn't know anything about audio, but who is better able to better distinguish between loudspeakers, because the customer is really sensitive to tiny differences in sound. I loved retail selling although the hours were lousy. When I worked for loudspeaker manufacturers I often did sales training inside retail locations in the USA and Canada. I noted which sales people were there year after year and which ones were seen failing with customers too often. If the failings were due to ears that weren't good enough, then there was no hope. I've seen many professional salespeople who were very successful selling other products, but were unable to make it in the audio business.

Now I have friends in their 50s who were listening to audio during its golden age during the late 70s and 80s. They mention they really looked forward to going to a 70s or 80s audio retailer for any product and then revel in a demonstration of some really good audio systems in store after store. That's all past. Many large retailers of audio equipment don't really want to display and demo most loudspeakers they sell, and the few that do don't want "lookers" listening to loudspeaker systems the retail owner knows they can't afford. Most retailers wanted to eliminate golden earred people from having so much influence on product selection and store set-up, so they eliminated emotional-level demos.

But great musical sound is like a drug for people with golden ears. They almost can't tear themselves away from that kind of sound because it seems to connect more directly to the brain of someone with golden ears than most people. We are a visual society now. Sound is much less important now. But in an era when music conveyed multiple messages to new adults, everyone listened to audio somewhere, somehow, and it was possible to see who had golden ears and who didn't.

When I worked retail in Canada, I met Ed Meitner. I think Ed has golden ears. CDs frustrated him so much it motivated him to sweat blood to make SACD systems, and now he develops most of the best ones. Ed was working as a service technician in Montreal for a distributor of UK products called Norm Yeager and Assoc. They distributed Leak, Decca, Celestion and others.

Ed used to design amplifiers in his spare time. The method he used involved listening to how certain amplifiers made a reasonably good speaker sound, and then modify his amp designs to make the loudspeaker sound better. Ed wasn't a loudspeaker or acoustics guy, but he had golden ears and could hear what amplifier circuits did to the sound of loudspeakers. I think that when Ed and I talked about loudspeakers, Ed could describe sounds in the high frequencies which made me certain he had golden ears.

It does make it easy for a good loudspeaker designer or amplifier designer to make a career in the audio business if they have golden ears, but others I've known made good audio biz careers with damaged hearing. So I think it is a gift, that if well preserved can help the G-E'd person to find great audio and enjoy it to its fullest (or spend way too much on it). But I don't think any amount of training will make a golden earred person out of someone who can't hear tiny subtleties really well, especially in the higher frequencies.
senior member
senior member
Posts: 368
Joined: 10 May 2007 00:42

Postby JoeE SP9 » 23 Jul 2011 17:22

Parts-Express has a selection of "thin film" and true ribbon drivers.
ARC SP9, HW19, RB300, Sumiko Blackbird front: Acoustat Spectra 22, 2 x 12" TL subs 2 bridged Crown XLS 402, 2 modified Dyna MK-III's. Behringer CX2310, DSP1124P, rear: Acoustat Model 1/SPW-1, Paradigm X-30, 2 Adcom GFA-545
User avatar
JoeE SP9
long player
long player
Posts: 2210
Images: 2
Joined: 23 Feb 2009 20:20
Location: Phildelphia, Pennsylvania

United States of America

Postby Whitneyville » 24 Jul 2011 05:13

When I was a music major and professional musician (winds) I may have had "golden ears" because I could play an Alprecht system English Horn in tune with itself. No adult human being can hear frequencies above 30Khz. They may be able to sense them, but they can't hear them. It's physiologically impossible for an adult's cochlear to respond to those frequencies. This has been proven by Mueller (1939 Nazi Germany) Bell Laboratories (1949), RCA Victor Laboratories (1950) The BBC (1950) W.M.Armstrong (1948) and over a hundred more citations on I can still hear the whistle of a US TV conventional CRT flyback transformer at 15,600Hz clearly and I'm 57, 8 months, from 11 feet 7 inches across my living room over my respiratory equipment (40-43 dB). My A-flat sopranino clarinet can reach 14,886.3 Hz (yes, higher than a piccolo). My Guinea Pig "churls" loudly at the top-octave on the "baby" clarinet, as these are frequencies piggies communicate at. Strangely, no one ever commented on hearing the supersonic tones on Quad 4 or CD-4 LP's (I still own a few dozen), and those people claiming to "hear" ultra-sonic frequencies never have audiologists tests to back them up. The hearing in my right ear is from 22 to 17,250 Hz and my left ear, 19 to 18,875 Hz +/- 10dB from Dr. Thomas Dotson MD. last Wednesday. My left ear is 3.7 dB more sensitive than my right, probably due to shooting, even with ear protection. For having taking steroids all my life for asthma and being my age, my hearing is "exceptional". My hearing threshold is 3.3 dB on "pink noise" in my left ear and 7dB in my right ear, again, at my age, "exceptional". My threshold graph at high frequencies and minimum perception level is very jagged, as any person's over 35 will be.
long player
long player
Posts: 2278
Joined: 13 Feb 2009 20:08
Location: Tulsa, OK

United States of America

Postby desktop » 24 Jul 2011 06:43

JoeE SP9,
thank you for the resource recommendation. The tweeter I use is most similar to the BestonRT003C @ this website, with a few tiny differences. Mine have no mounting flange (I glued mine into place), no phasing plug, and thus is not loaded to go as low as 4KHz. The ones I have will only go as low as 7KHz with an 18db/oct xover. The magnet is also larger on the Beston RT003C compared to mine.

Minus the flange on this unit, my tweeters are about 1.75 inches (44mm) x 3/4+ inches thick (20mm, I was able to measure one with a micrometer).

There are 2 great ways to use tweeters like the Beston RT003C by making a stalk, or hanging them from a ceiling or a metal bracket frame. The "holder" of this tweeter is a small cube about 3 inches square on a side. The circuit design would be to put 3 in series and then parallel those 2 sets of three. The passive crossover could easily be made on a circuit card outside the tweeter "box" itself. Every surface of the tweeter box not directly in front of the diaphragm areas needs to be covered with sorbathane or at least neoprene foam. Getting a scoring tool with a number of ridges allows you to make V-shaped grooves all over the sorbathane/neoprene. This will drastically reduce the secondary reflections of high frequencies off the (previously) hard flat surfaces like the flange.

Once the box is built and the tweeters mounted into it, you need to align the box so one of the tweeters faces the listener, pointed at about a 30 degree angle slightly off axis away from the box (up-down direction), and 90 degrees turned so that one of the corners points up and down. A wooden stalk can be made from a 1x2 piece of wood covered with scored sorbathane/neoprene. I find it best to mount the tweeter slightly further away from this super tweeter compared to the other components in the users speaker box. This stalk will be mounted to one of the bottom corners.

A second choice is a sturdy metal L-bracket or two covered with scored sorbathane/neoprene. Screwed to the back of your existing cabinet, this(these) bracket(s) allows the tweeter to be hung with fishing twine so that when gravity allows the tweeter to hang, in the same orientation as above. The fishing twine won't re-radiate edge-transmitted sound, and they won't reflect sound either.

I would need to find my papers on crossovers, but it you have a 2-way loudspeaker with a 1 inch tweeter, I'd recommend a crossover frequency down 3db at about 12KHz to the new tweeter array. The imaging doesn't suffer because only one of the tweeters contributes to the direct sound. But the sound power of the system will increase dramatically. The Beston RT003C is actually a very small tweeter. I believe the specs are a bit optimistic. But you would only need a resistor pad (T) in front of the crossover, to balance this super tweeter with the rest of the loudspeaker components in the original box.

The component speakers at Parts-Express seem very expensive if you are building your own loudspeaker systems from scratch. The easiest way to make a major system is to do without a single hollow box except for the speaker producing the lowest frequencies. A very good loudspeaker can be made from normal components if each component loudspeaker only handles 2-2.5 octaves each. The large heavy bass unit should sit right on the floor and act as a base for the entire system. The rest of the component loudspeakers can be arranged in the smallest possible boxes, hanging off of a solid steel bar of a reasonable height. By having a crossbar holding each component loudspeaker in place, you can even positionally phase align each component.

Sadly with skin effects, reflections, resonances, internal standing waves and re-radiation, boxes make a huge amount of spurious noise. Once you make a box that is inert (like making it out of 3 inch thick - 75mm, concrete) it is usually too heavy to be movable. In point of fact, in most cases the box interferes with the sound from the components, almost as much as it helps the bass to have that one component in a box.

If you choose (up to 7KHz) component loudspeakers that produce frequencies that are not higher up than the frequency whose length is equivalent to 4x the actual cone diameter (not counting the surround or frame). So for example, a 14 inch speaker with a cone exactly 12 inches in diameter, could be producing a 360 degree radiation pattern that would fill all the space from the floor into the room, up to about 300 Hz. This rule becomes more difficult to implement as you go up in frequency, which is why I use arrays for the ultra-high frequencies to retain the 270-360 degree radiation pattern. 16 inch or even 13 inch cones on 18 inch and 15 inch loudspeakers are also useful, although their "full loading" frequency is lower.

6 component loudspeakers, with 2 inch domes can overcome the "best" -3db crossover point of one 2 inch dome component, by coupling. With 6 x 2 inch domes mounted in a box similar to the one mentioned above, the xover frequency could rise to -3db at 6KHz. Although not absolutely optimum, the crossover point on the 6x ribbon tweeters listed above could be lowered to 6KHz as long as pure tones in that frequency range are not played at full power. The tweeters that would cross over @ 6KHz, could have a sensitivity of 92/93 db with 2.83 volts applied, and a real power handling of about 30w RMS. The dome upper mids also mentioned above would also be range of 1500-6KHz, and their sensitivity could matched to the tweeters.

I would add an extremely durable 6 Box of mid-bases/midranges would utilize a very inexpensive 5 inch speaker with a 4 inch, polypropylene cone (or carbon fiber as I mentioned in a previous post) which as an individual component would have a high frequency cutoff of 900-1KHz, would not be useful to 1500Hz @-3db on the top, and they mate well with the 14/12 woofer at 300 Hz.

The SIX BOX as I call them have many advantages like, huge power handling, coupling drivers for 360 degree radiation (although one driver must point at the listener so there can be phase alignment). This way you only need 3 xover points, 300Hz, 1500 Hz, and 6KHz, but unlike some companies like JBL, you would have a system that fully couples to the air in your room and produces a dispersion pattern similar to reality. Reciprocal opposing forces when the component loudspeakers "play" are all balanced and thus there is no micro-movement of components that badly affects phase alignment.

There is no perfect loudspeaker system. There are good loudspeakers and bad loudspeakers, but few try to achieve the sound of reality. I use 15 inch loudspeakers that are in my "floor stand" (2x) and they are tuned to resonate at 20Hz in my room (24x18x9 feet). I'm thinking of cutting a port in a door that goes to a garage/living-area or huge dimensions (33x33x18ft) to reduce the pressure on my woofers, so this is why I need a 5 way when a 4 way would work really well.

I use 3 converted 160w@4/8/16 ohm tube guitar amplifiers, with gentle parametric equalization for the midrange domes because the human voice range of 400-2KHz is the most critical frequnecy range there is for "reality". I just overpower the bottom with 3200w of interchangeable low end power (I live on a ranch with 400 amp incoming service to satisfy my needs). Normally ranches need electric wood chippers and such like and with barns etc. so the provisions for power require a 1 gauge incoming twisted pair from the street transformer, and a service box nearly 4 feet high. I bought it, I didn't make it, but it suited my interests.)

The midbass is the region I consider more important than the low bass. So in this range of frequencies I use a Phase Linear 400, one of the best sounding of the monster amps. In another similar system I use a Yamaha Class A/AB 150w/ch amp which is smoother but less powerful on the crescendos of full orchestras. But now-a-days amp power for the lowest low end is cheap. But once you approach the vocal range, use good quality amps. You'll hear it once reality becomes really transparent. REMEMBER that each component loudspeaker group will need a passive crossovers to protect the component loudspeakers. This makes amp needs

For the tweeter array I suggest a very high quality full 25w/ch Class A amplifier. High frequencies actually consume more power per given loudness level than extreme low frequencies. The compactness of the 20KHz signal requires more power to reproduce at the same SPL level. Luckily there are very few times when the highs require such power, unless you are enamored by the London Verdi's Requiem of a Thousand. If so then you'll need the power figures I recommend. A 25w/ch full Class A amp requires about 1000w total power to operate properly.

The way I have suggested designing your "reality" loudspeaker system" you will have about 2900w/channel power consumption at MAX. MAX continuous power is really not important at this level. Peak power becomes more important as the system will really be operating at about 8-10w/ch (MAX) for the loudest music normal people play. The system peak power would be about 5000w/ch for short duration signals. I can recommend cheap watts per channel for the extreme lows, but the other amps would be determined by what you can find. Remember that tube amps store about 300-500 volts on their plates for peak power (you do the math). But tube amps don't have great current output for the bass, or high power output for the extreme highs. Here watts per dollar amps work well for the lows, and super clean Class A amps work well for the extreme highs. Your total 17:1 ratio of loud continuous power to peak power gives you a rough 50:1 max power to peak power ratio. This is pretty transparent.

I use other combinations for other systems but this is the best to make a reference system. If you find a pair of amps or receivers with high power or current that you like, then it is possible to configure systems that utilize these kinds of power components. If you MUST have a speaker "box", then I can recommend boxes with all non-parallel interiors. But the box really does "get in the way" of the best sound, component loudspeakers can produce. I guess that sums up my philosophy of sound based on listening test results from 12,000 listening tests on regular music listeners. The smallest radiating enclosure possible produces the best sound. At the extreme low end, in order to match the radiating patterns of the other components, enormous power compensates for the required enormous rooms to reproduce extreme low frequencies like 20Hz.

The recommended system above is a reference system. It will give you an insight into vinyl discs you have never experienced before. Fewer component loudspeakers can be made into systems that reproduce sound well, but usually with changes in radiation pattern that affect the way we perceive sound in enclosed rooms. Remember the voice test. If you know someone who sings and plays something, you can test your system for "reality" by making a recording with reference mics surrounding your musical helper and then when you play the recording back, it should sound like the "musical friend" is right "There". Then you'll know you made the system correctly.
senior member
senior member
Posts: 368
Joined: 10 May 2007 00:42

Postby dlaloum » 24 Jul 2011 07:32

Fantastic posting - very interesting stuff...

I worked for a brief time (about a year) as an Audio salesman.... (I was a better consultant than salesman...)

Got to love the Quad electrostatics - that in the late 80's took up residence with me...

It is only over the last 3 years that I replaced my ESL's with a "magnetic" driven speaker - The Gallo Ref3.1.

I was having "WAF" issues at home due to the size (and room dominance!) of the Quad ESL989's.

The relatively diminutive Gallo's surprised me when I heard them in a terrible listening environment - surround by TV's, a couple of dozen TV's and floor stock all making sounds all around them - but when these babies came on they performed quite a bit like an ESL. Extremely unexpected in a mass market store!!! - Soon thereafter the store decided to cease stocking these high end speakers and associated gear - they dropped the price sufficiently to (in conjunction with the WAF mentioned before) - cause me to sell my 989's and 63's.

Longer listening has convinced me that they are not quite as transparent as the ESL's - they can get confused a little when the music is highly complex and layered (full symphony orchestra going at it).... And when it comes to listening at lower levels - nothing beats the ESL's in my experience.

The Gallo CDT tweeters apparently have useful response out to 35kHz... (although the speaker specs talk about +/-3db out to 20kHz)

Wondering whether you have heard these and/or would care to comment?

bye for now

long player
long player
Posts: 3529
Images: 188
Joined: 29 Dec 2009 06:21
Location: Melbourne, Australia

Postby avole » 24 Jul 2011 09:42

Posts: 1617
Images: 3
Joined: 31 Jul 2008 21:02

Postby desktop » 25 Jul 2011 02:36

to dlaloum,
I have not heard the Gallo 3 loudspeaker but I will try to audition it when I have the chance.

to avole,
I looked at the introduction to the Moulton course materials for Golden Ears and I feel that while there may be good intentions by the developers, they seem to miss the point. Perhaps my definition is different. I think GE'd people can hear things trained non-GE'd people cannot hear. People can be made more aware of various kinds of sounds through training, although so much depends on the loudspeakers being used to hear the course material, and this is not mentioned.

But the people I've met with true Golden Ears IMO, don't need training. I certainly wasn't trained until I worked for Disney. It is somewhat like the difference between a prodigy and a student who is trained for years but can only produce a very boring and uninspired version of a musical piece. Yes, people can be trained to hear some things in a studio environment, and then remember that training. But when you look at course material that focuses allot on signal processing, you realize that Golden Ears aren't what they are training people to have. They are trying to help people who want to be studio engineers. Studio engineers do not require GEs.

Gene Czerwinski had GEs IMO. He could come into a room where a new loudspeaker was playing so a bunch of us could listen, and he sometimes closed the session by saying something like "the positive and negative wires on the midrange on both speakers are reversed". Invariably he was right. There is no way to train someone to hear things like this incorrect phasing. A course that is dominated by learning about level changes, isn't giving the students golden ears.

Whitneyville is likely on track when he says that he may have had GEs which he noticed while playing an instrument professionally. There was quite a bit of training involved for him, but having perfect pitch played a big part in his ability to recognize sounds, levels etc. I've met many totally untrained audio newbies from 1972-1996 who could recognize incongruities in what they were listening to, and they did it as well or better than people who spent every day listening to sound as engineers or designers. You would have thought that the trained professionals should have done at least as well, but the majority of the trained pros didn't as well as the GE'd untrained newbies. It often led to embarrassments for pros. On the other hand, I was often extremely frustrated to demo a very fine product to a studio engineer, and then compare it to the product he/she used, while asking, "Can't you hear how much better the "insert some instrument", only to have them say, "Hmmm, no, not really". Perhaps 30% of the industry pros I met had GEs, and that's being charitable.

This was part of the UREI "mystique". UREI had trouble in the beginning getting people to be able to hear the different between their improved xovers and a normal Altec 645. But a few respected engineers could actually hear the difference and bought the UREIs. Then in a tidal wave, everyone who had that size hole in the wall for monitors bought the UREIs because even if their engineers couldn't hear the difference, "What if our customers can hear the difference in these phase incongruities?"

That's why I had too many experiences where people who designed audio equipment showed me something (usually loudspeakers) they were very proud of, and tried to "sell" me on it, while I often thought their new product (read: baby) was horrible. Usually the marketplace bore out that judgment. Retailers in my era could often tell just by listening (without handy comparison equipment) which audio products had problems and which ones didn't. Those retailers who were more "specs" oriented, or "deal" oriented, often went out of business because the consumers with GEs wouldn't buy the poor products with problems no matter what the offer made by the retailer (who didn't have Golden enough ears). Those consumers often influence their friends as well.

When I did the listening test the AES conducted to try to argue intelligently about CD specifications, the lo pass filter inserted at 35KHz seemed to cause the outdoor sound to "lose something". It seemed as if one option sounded natural and one sounded like my ears hadn't adjusted to a change in air pressure. I have developed a bad habit from my audio days. I constantly flex my neck muscles in a way that would equalize the air pressure behind my ear drums, in order to have the cleanest, clearest hearing possible. But now my day is spent moving horses around or picking plums, not designing loudspeakers. Now this "tick" is compulsive or neurotic.

I cannot say how I may have sensed a change in sound bandwidth above 40KHz (especially since the AES was doing this testing 20+ years ago), but I did have a chance to play around with 2 of the Panasonic leaf tweeters at Disney and we were all astonished by what we could hear coming out of it in frequencies Whitneyville says we would never be able to hear. I assume that someday actual wires will be attached along the circuit path from the ear drum to a human brain and only then will we know what the ear can do with an ultrahigh frequency sound signal.

The German 1939 test is suspect because transducers of the day were poor to say the least. Of course, I am not an audiologist but I did go to school with one, so I'll ask him about it. In his opinion there are many, many "non-standard" ears he tests that defy explanation. It is like an argument I was having about roof prism binoculars vs porro prism binoculars. Roof prism binoculars must be flawed by definition because you cannot compensate for an infinite number of phase distortion frequencies. You can compensate for 60 or 100 different visual lightband frequencies, but not an infinite number. Eventually the consensus was that the corrections made by Leica, Zeiss and Swaro were sufficiently close enough to perfection that most people wouldn't care, because the corrections were centered near the 3 peak light wave frequencies that the average person has in their eyes.

It was at that point that I asked, "Well what about the people who have 4 color sensitivity peaks?" You could have heard a pin drop, not because those 7-10% of the population who had completely different peaks from the average matter to the market, but for a much more insidious reason. The reason is that almost no one in the world has completely "average" eyesight peak sensitivity frequencies. Satisfying the "average" color peak frequency ranges with phase filtering in roof prism binoculars really only makes them "good enough" for about 15% of the population because the center frequencies are only an average, not really the way most people see. Porro prism binoculars do not have phase distortions introduced into the colors because they don't use the roof prism that causes the trouble. Both binocular designers walked away muttering after our discussion, certain that through marketing, they could con people into buying roof prism binoculars anyway.

When my hearing was last tested there was a noise floor calibration test done as part of it. But I asked the tech to continue with quieter and quieter signals, until they were much quieter than the noise floor. I could still pick out signals @-10db below the noise floor. Noise floors are complex and audio tones are not, so detecting the tones was easy for me. The tech herself tried to hear these tones that far below the noise floor and wasn't able to do it. I applaud Whitneyville because it sounds like he has preserved his hearing as well as possible (considering he is a shooting enthusiast). But since I have not ever heard that it is physically impossible for anyone in the world to ever use their ears to detect frequencies above 35KHz, I obviously need to go back to school on that subject.

My hearing has deteriorated slightly over the years, but it is rare that I run into cases where I know I used to be able to hear something and now I can't. The 19KHz stereo tuner tone drives me nuts if a tech is doing a test and that signal gets sent to the amplifier. Squeals, and tweets and tiny car, house and power equipment chirps can keep me awake at night or give me a headache until I find the source. Transformer whining can be especially bad. I dislike even going into a machine shop because there will invariably be many noise sources that project ultra-high frequencies.

Whitneyville seems to think that the rush by many manufacturers to make loudspeaker systems and components capable of producing low distortion output at 40KHz is a waste of everyone's time because no one can hear any of the 35KHz-40KHz frequencies anyway. Just sensing them isn't enough to market products like the Tannoy studio monitor with their super tweeter that is capable of producing low distortion sound at 40KHz. But contact Tannoy and see how many people they know of who can actually hear these ultrahigh frequencies, at the company, amongst the users, at the shows, etc. While some people may have a product like this to show off (status), many others can really "hear" the difference that this extended frequency response makes when they are mixing down digital signals that extend above 35KHz all the time for SACDs.

But I'm not saying that Whitneyville is incorrect or that the scientific papers researched over the years are rubbish. Perhaps it is time, using better quality test equipment, to see if the original papers were correct or if the paper writers used their own anecdotal "evidence" to give opinions that no one could refute because those people writing the papers had more data and evidence than everyone else, and these were also the most esteemed audiologists of their day. But that was then.

It's like the CD. If we had had terabyte mass storage and digital-throughput radio broadcasting in 1985, we wouldn't have ended up with such a poor spec for CDs. Perhaps now is the time to replace studies last done in 1950 (19...50???) and get a more precise measurement of what the human ear is capable of. I may be mistaken but I hear that there are efforts to extend the upper frequency range in some new loudspeaker components to 50KHz. Wouldn't want those guys to be wasting their time.
senior member
senior member
Posts: 368
Joined: 10 May 2007 00:42


Return to Amplifiers and Loudspeakers

Design and Content © Vinyl Engine 2002-2014

faq | site policy | advertising