Beyond the Ariel

There. Four listening techniques that will reveal a great deal of the gross and obvious defects of speakers from $100 to $100,000.


Couldn't we just edit an audio cookbook or something with this thread? I think this is one of the best threads ever. Thank you Lynn Olson, Mr. Geddes, JohnK, DDF, soongsc and many others on this thread for the quality of your posts. You are doing tremendous good for the ones which are just diy-ing, and not making a living from audio.

I wish I could have the time to have a more insight look of the informations posted here, but the faculty I am following is allready asking for it's tribute in time.

@angelo, I have been following the development of your project, at least the fragmented info you posted on the net. I've seen in a pic from your gallery a Radian neodynium compression driver (950 PB?) in what looked an un-finished tractix or LeCleach horn. Did you had the chance to listen to it? And how do you compare it to your coaxial compression driver.

I don't know if Mr. Lynn Olson is still considering a compression driver for the 500-10.000Hz band in a non-traditional horn/waveguide but it's still intersting though.
 
I like the wide-open spatial presentation very much, but my enthusiasm stops about there. Very low efficiency and requires 1 kW amplifiers which pretty much rules out delicacy and beauty. It sounds like the low-efficiency speaker it is - stressed and working very hard at most dynamic levels, with a pretty noticeable look-at-me "hifi" quality to the presentation.

That also describes the overall sound of nearly all low-efficiency audiophile speakers. Take away the spacious MBL imaging, replace it with paper-thin cookie-cutter image quality, and you've got your famous-name Brand W with their prismatic-shaped "giant robot" cabinets right there. Fake images, fake dynamics, plenty of cone coloration, and big big prices. Good reviews, though.

The true charm of high-efficiency - if it isn't grossly colored like a vintage PA system - is the relaxed and effortless dynamic quality. This is such a wonderful change from 84~87 dB/metre systems, which sound much more stressed at all levels, and definitely unhappy at transient-peak levels more than 95 dB.

As for the horn questions, well, as mentioned before, the LeCleac'h is a seriously good horn profile, and available in a wide variety of sizes from Azurahorn in Australia or knowledgeable local builders. Our man in Boston recommended a highly tweaked unobtanium Vitavox S2 compression driver, with all others being only suitable for "primitive" tastes. So all the rest of you cave-men here at diyAudio will just have to make do with TAD, JBL, Yamaha, Radian, et al. Maybe the Olympian deities over at Cogent will favor us with a suitable large-format 500 Hz ~ 5 kHz field-coil driver.

Or you could do the sensible thing and participate in the Geddes group buy for the Summa/ESP15 waveguide+foam, and use the recommended B&C 1"-exit mylar-diaphragm driver at a far lower price than the exotica mentioned above.
 
radian etc...

hello SunRa, hello Lynn

i am waiting for crossover parts from germany, autotransformer, etc. to start to play and tweak with the radian driver. the first pair of speakers i did not have to do any tweak and crossover built, as the hornsystem comes ready to play from Bert of BD-design. With the Radian i will start my journey with crossovers, measurements ( i will get also microfone etc.... ) I guess in a month ore so i can give some info's and comparisons . The most compromised part of the Orphean's is in my opinion the treble. It does not sound right to me, it is beamy, and tone is better with a separate tweeter. I had a Coral H100 before, and i liked it more. Next project, i am waiting my wood worker to finish a big wood lathe. With it i can turn mdf horns up to 1m diameter. Idea is to make a 3 way satellite : 88cm diameter tractrix horn from 150hz to 1000hz, with a 8" Fane Studio8M, 49cm diameter horn ( pic in my blog ) with radian from 1000hz to 10khz, and jbl 2405 tweeter from 10khz up. Might i try as well the RAAL's, many people over the net say it's very good. . to adress directivity of horns, and get more spaciousness, i was thinking to use a radian coax the way i showed in the draw a post before.

Our friend wants to monopolize the way to built a good sounding hornsystem, and the drivers to be used. you said right, Lynn, he behaves like a dictator. So the rest of us all, are morons, primitives, and bla bla bla. A view email exchanges with John Hasquin, who made mdf wood tractrix horns in the past, where very enlightening :

Due to the very high effeciency of horn loaded speakers, sound levels become too loud before damage can occur to the drivers from over excursion. You can run compression drivers without a crossover if you want. Romy does use a crossover, but it is a line level crossover that is built into his amplifier. I would recommed a crossover point of 1KHz between the midbass and midrange horns. 600Hz is too low for the midrange horn, it will sound congested if crossed this low. The 88cm horn loaded with Fane Studio 8M is good out to almost 5KHz. The component values for a 1KHz crossover can be calculated with any of the on-line Web calculators. I would suggest a simple first order Butterworth. You can also try a second order Linkwitz-Riley. Use good quality capacitors and low DCR iron core inductors.

The Radian compression drivers work just fine. However, it’s got more to do with the content of the music and not with the design of the driver. See the attached graphic to see the relative acoustic power levels of the typical orchestra verses frequency. The highest acoustical energy is centered around 330Hz. You do not want to put a crossover anywhere near this! You should not put a crossover anywhere between 175Hz and 1KHz. The non-linear reactance of a crossover (active or passive, it doesn’t matter) in this region will messes up the time, phase, and impedance constants. In addition, you need to stay at least one octave away from the resonance frequency of the driver when designing a crossover. The resonance of the Radian 950PB is around 550Hz, a 1KHz crossover barely fits this guideline.

Do not over think things too much. Just biuld it and make adjustments until it sounds the way you want it. There is no single answer, only various degrees of rightness.

I have worked with both the Radian drivers and the Vitavox S2. The thing you must remember is Romy is looking for a very, very, very, very specific sound and the S2 does it for him. I personally didn’t think the S2 was a superior driver. It sounded good, but wasn’t the Holy Grail of compression drivers for me. I like the sound of a lot of drivers, it just depends on what you are trying to accomplish. I like the sound of the JBL 2440, Altec 290-16L, Emilar EC-320, Gauss HF-4000 and a few others.
 

Attachments

  • 1.jpg
    1.jpg
    29 KB · Views: 1,025
mbl's

hello Lynn

i heard the MBL big system at CES this year. I did not have the impression that it sounded stressed. Actually, intereseting, that it was very dynamic, almost like a hornspeaker. However, what disturbed me most, was too much sound radiating backwards. At a concert, most sound radiates to the public, and only a small part is reflected thrue the walls etc. Would it not be better, if only 70% would be direct sound, and 30% reflected ? If this is adressed with horns, it would be perfect. interesting what Avantgarde writes in this regard in their white papers. might you comment it.

Just think of a full orchestra recorded via just two microfones. If you play such a recording with a loudspeaker system, the recorded sound reproduction will include the original imaging of the room situation. If you add to this acoustical information of your own room you will automatically receive a vague, diffused picture. The greater impact you own room has on the acoustics the more original performance will be blurred. This effect is very tipical for loudspeaker systems with a wide dispertion characteristic. A listener with less experienced ears will immediately have the impression to heaer real room staging. The sound seems to come from everywhere, demonstrating depth, width and height. But these artificial reflections have nothing in common with the original sound.

there is also a very informative site from Jean Michel LeCleac'h adressing this matter :

http://freerider.dyndns.org/anlage/LeCleach.htm

download theory of operation

Angelo
 
there is also a very informative site from Jean Michel LeCleac'h adressing this matter :

I have a problem opening the link. I get something like this: Die Seite kann nicht angezeigt werden... Page can not be desplayed

Regarding MBL, I've read so many coments that I'll probably just have to wait and get a chance to listen them. Someone was saying he heard them drived with some 18-25W amp and sounded great.. I'm somehow curious about new tehnologies. Most of the comment where more similar what Angelo experienced.

About the cross-over point, I know I have thought again and again about it... how to avoid a cross point in the 300-3000 band, the speech band, while the music content is so massive exactly in this 330Hz region, the power band or the so called music band. This calls again for compromises as a single driver working in the 300-3000 area will have to be accurate with voices and in the same time would have to be capable sustaining the huge amount of musical information, dynamic range, and so on... Probably the 900-1000hz cross-over is a good compromise afterall considering Mr.'s Geddes observation regarding the changes in our hearing mechanism with frequency. From this point of view I think Mr.'s Olson cross-over typology is adressing the problem exposed by Angelo.
 
Lynn

I too appluad your "tests" for audibility as attempts to determine what truely matters in sound quality. However, I would caution about the potential pitfalls with almost every one of them. The cause and effect relationships are never really well established. Like listening to the cabinet very close. There is no guarantee that the sound you hear in this test is actually what is heard in the far-field.

Your tests are useful when one lacks the sophistication of doing things more scientifically, but I, for one, would not bet my designs on tests like this. They lack the quantification and proven correlation that I look for in my subjective testing.

Almost any data is better than no data, but we must always keep in mind exactly how much validity we can put in this data. Not all data has equal validity.
 
Re: mbl's

angeloitacare said:
hello Lynn

i heard the MBL big system at CES this year. I did not have the impression that it sounded stressed. Actually, intereseting, that it was very dynamic, almost like a hornspeaker. However, what disturbed me most, was too much sound radiating backwards. At a concert, most sound radiates to the public, and only a small part is reflected thrue the walls etc. Would it not be better, if only 70% would be direct sound, and 30% reflected ? If this is adressed with horns, it would be perfect. interesting what Avantgarde writes in this regard in their white papers. might you comment it.

Just think of a full orchestra recorded via just two microfones. If you play such a recording with a loudspeaker system, the recorded sound reproduction will include the original imaging of the room situation. If you add to this acoustical information of your own room you will automatically receive a vague, diffused picture. The greater impact you own room has on the acoustics the more original performance will be blurred. This effect is very tipical for loudspeaker systems with a wide dispertion characteristic. A listener with less experienced ears will immediately have the impression to heaer real room staging. The sound seems to come from everywhere, demonstrating depth, width and height. But these artificial reflections have nothing in common with the original sound.

there is also a very informative site from Jean Michel LeCleac'h adressing this matter :

http://freerider.dyndns.org/anlage/LeCleach.htm

download theory of operation

Angelo
I have experienced the same.

Additionally, stored energy in the drivers also blurr the original recording. Teaking some drivers I have found the the faster the initial part of the CSD drops, the more of the recording you hear. A favorite test of mine is the absolute polarity test. The more one polarity is distinctly identified as having better image depth, focus, and realistic transients, the better this speaker will perform in general, and that a polarity switch in the system will help select the right polarity to maximize listening experience.

I have listened to the large MBLs a few times, and my guess is the CSD probably is not so good. I doubt anyone would let me do testing on these and get away with the data.
 
Lynn,

I'm not sure what your sketch (Post #2120) shows... if it is your design, then the following comments apply:

While the shape and look of the cabinet is very "sexy" and "space age" I have concerns about both the dimensions vs wavelength for the baffle, and edge diffraction and also the issue of shadowing/reflection(s) for the tweeter.

Unfortunately my limited experience has shown that what looks good or even sometimes theoretically would appear to be good frequently causes anomalies that are undesireable - the baffles that seem to work best are often not terribly desireable visually or aesthetically.

Ymmv.

_-_-bear
 
Late to the party and my experience with 18sound

Awhile back I decided to take a stab at making my own speakers. Not having any knowledge, and looking to get by without having to learn, dictated the approach I took (stupid I know and yes I now know I am in over my head and have to learn). I decided to go with a CD waveguide, open baffle mid, and dipole bass (but am still using BR for now). I used active crossover/solid state to bass and passive between mid/high.

I liked what Dr. Geddes was saying but at that time he wouldn't sell his horn so I looked for something close and discovered 18sound. Not many others had tried them (at that time) so it was a gamble. As luck would have it the USA distributor was near by and I was able to work out a way to buy direct. IMHO the products are top quality and the data on the website isn't cooked.

I wanted the XT1086 horn but it was back ordered so I got the little XT120 and what I thought was thier best compression driver, the ND1090. A few weeks later I was able to get the XT1086. The little horn isn't bad at all but I like the bigger one much better. There is no problem at all getting 120dB cleanly.

All that have heard it feels it offers outstanding clarity, no shout/honk, and a very wide sweet spot. It really doesn't sound like a traditional horn in these respects and when I played with adding a rear firing tweeter I begain to apprecitate the directivity and felt the lack of early reflections was big advantage (my room is very live). However I did find I liked the added soundstage of the mid range in open baffle. Outside of wondering if an alminum diaphragm might have been a better choice I am very happy. I high pass around 1K7 FWIW.

18sound has some suggested kits and one uses the xt1086/nd1090 and one can cobble up a passive crossover as a starting point from that info.

Now as to one of their's for a wide range....well most of the woofers look to be made with horn loading in mind. Why I like a larger mid, the 6nd430 looks like it could run from near 80 hertz to upwards of 3khz. For my mid, I started with an old standby, the audax pro170mo but have recently picked up some 12" alnico (made by RSC) to try.

Anyhow, I know this is late coming but I saw some posts asking for experience with the 18sound stuff. Based on my experience I didn't care for a rear firing tweeter and liked the lack of early reflections a lot more in the treble range. I also liked a paper cone open baffle mid for warmth and that big spacious soundstage (even if it is an illusion/trick:)

I am recovering from a hernia operation but have hopes to get started on the dipole bass part in the near future. I'll post my findings. For food for thought, it has been suggested that the dipole bass would be optimum when combined with multiple subs asymmertrically located.


Russ
 
8NMB420 massive basket

Lynn Olson said:
..., internal reflections in the driver, unwanted radiation from the spider, dustcap resonances, etc. etc. etc. [/B]

Don't you think Eighteen Sound's 8NMB420 is a questionnable choice regarding these kinds of problems ? Its basket looks more of a bell with few holes than a modern areated basket (the kind most of the rest of their line has).
 
BD15 bass from BD-design

hello Lynn

you might also consider BD15 Bass from BD-design :

http://www.bd-design.nl

he uses it also in a open baffle model. His homepage is under construction, so the picture might not be available of the finished speaker. He uses 2 of his woofers in a open baffle, and a lowther driver for mids / treble. i asked in his forum about OB, here his answer:

All depends on the design of the box and the panel.... you can create a worse sounding open baffle system compared to a good designed and well build "boxed" system.

You'll have a bigger change to get "cleaner" bass from an open panel but in return you will get more distortion from the driver as it needs to move a lot to compensate for the loss in bass (cancellation) and more power to handle.

Everything is a compromise in speakerland...


rds Angelo
 
mige0 said:
Hi


JohnL, What bass driver is this ?
Are you satisfied with it ?



Greetings
Michael

Michael...

it's a 10" reclaimed driver from an Infinity RSII from 25 years ago...

measures ~36Hz Fs, ~.68 Qts

VISCLON3.jpg


bass is reasonable, given the baffle size (18" at top, 32" at bottom). These two OB's are augmented by a sonotube sub, crossed @ 80Hz, so low bass is not expected.

John L.
 
Lynn Olson said:
This one is so simple it looks really stupid. That doesn't stop me when nobody is looking.

4) Cabinets have really weird-sounding internal standing wave colorations, and especially cabinets where multiple drivers share a common chamber (line arrays, I'm looking at you). If you visit the proud audiophile or manufacturer when the drivers are pulled out, if the baffle opening is big enough, put your head inside (don't get stuck!!!), or if you want to play it safe, put your ear right at the aperture.

Listen for awhile. Note just how different it sounds than when you're several feet away in free air. The cabinet, even if is filled with Miracle Absorbing Material, has a hollow, muffled, resonant - well, boxlike quality, but with odd colorations from the damping techniques themselves. It's not really a simple "box" sound - there's other stuff, too, and hard to describe.

Speaker diaphragms are acoustically transparent, no matter what material they are made from. These weird boxy qualities are there all the time, overlaying their signature on the honest sound of the driver. This is especially severe in line arrays with a common rear chamber - the best reason for using multiple, small, isolated chambers, preferably of dissimilar dimensions.

There. Four listening techniques that will reveal a great deal of the gross and obvious defects of speakers from $100 to $100,000.

Lynn...

Interesting observations... and mostly what drove me to build these BG75 dipoles 2 1/2 years ago...

An externally hosted image should be here but it was not working when we last tested it.


They sound pretty much the same while circumnavigating, essentially have no cabinets (other than the smallish bass boxes for the peerless 6 1/2" mid bass, have minimal xover components (the planars run from 400 Hz on up with 1st order xover. Low bass is augmented with two large sonotube subwoofers from ~80 hz down to < 20Hz.

Hope the surgery heals well. I have a 6" ss plate in my left arm from 1985 auto accident. Luckily, my left arm was only paralyzed for a year while the radial nerve grew back to my fingers and wrist. The pins and plate cause only minor discomfort from the occasional low pressure system about 2 days before it arrives. Seeing your xray brought back memories of looking at my upper arm and seeing those deck screws holding it together :eek:

I've noticed the tweeter effect b4, I use this often, and have found the source material often is contaminated as well... as in the recording technique.

John L.
 
gedlee said:
Lynn

I too applaud your "tests" for audibility as attempts to determine what truly matters in sound quality. However, I would caution about the potential pitfalls with almost every one of them. The cause and effect relationships are never really well established. Like listening to the cabinet very close. There is no guarantee that the sound you hear in this test is actually what is heard in the far-field.

Your tests are useful when one lacks the sophistication of doing things more scientifically, but I, for one, would not bet my designs on tests like this. They lack the quantification and proven correlation that I look for in my subjective testing.

Almost any data is better than no data, but we must always keep in mind exactly how much validity we can put in this data. Not all data has equal validity.

All four of the "tests" are more ear-training than anything else. They make no pretension to replace any measurement, and are not intended as such. I have gradually realized, with something of a sinking sensation, that most audiophiles are genuinely unaware of what a highly colored loudspeaker sounds like.

I blame the Big Two magazines, which routinely describe expensive loudspeaker systems with measurable, gross, and quite obvious peaks and resonances as "detailed" and "accurate". As James Boyk perceptively remarked, there is no sound too ugly to be described as 'accurate'. This word in particular has come to mean "unlike a live, acoustic instrument", and something closer to a disco PA system with multi-kilowatt amplifiers.

As a result of the gradual debasement of audio-language over time, many people think that brutally harsh sound has something to do with reality, or literally, fidelity to the original, when all they've done is accustom themselves to technically defective reproduction. Most of the folks I meet at trade shows really do think the $100,000 speakers on exhibit represent good engineering, when all they represent is glossy packaging and successful marketing.

This is different than cars, for example, where a $100,000 sports car really is a very fast car representing advanced engineering. In audio, we routinely have Pintos, Yugos and Trabants dressed up as Ferraris, priced accordingly, and worst of all, the market accepts this.

So I gently try an re-acquaint people with the basic elements of a loudspeaker intended for enjoyment of music: freedom from the most obvious resonances, cabinets that work as intended, and freedom for phasey-sounding dispersion and diffraction problems. When I see flagrantly bad-performing loudspeakers get top reviews and command astonishing prices, the problem is deeper than the manufacturers and the reviewers; it's the listeners, who aren't demanding better.

The first steps are to break free of the miasma of pseudo-technical "reviewer language", realize that cost, prestige, and famous-name recognition have almost nothing to do with good taste and good engineering, and listen with fresh ears. The four "tests" I described are one way of experiencing loudspeakers without all the baggage of reviewer-speak.
 
Hi

Yes clever marketing and lawyers power overrules basically EVERYTHING.
This also is true in cinema sound widely accepted as kind of audio reproduction reference for the crowd.

There, for decades, Dolby has successfully managed to place the term " Dolby Digital " as synonym for top notch sound.
Almost nobody knows that the AAC3 format used there is nothing better than MP3 just extended for surround sound.

All it was better for compared to the usual 35mm optical sound tracks was noise and bandwidth in addition to the digital inherent advantage of long lasting without loss..

The competing digital formats from Sony and DTS are way better due to less compression rate and the ancient analog 70mm magnet tone was superior to all of them.


Greetings
Michael
 
Lynne

I couldn't agree with you more. People often rejected my loudspeakers because they don't sound like other loudspeakers - hence they must not be any good. "Expectation" is the bottom line - if the speakers sound as people have come to expect, then they are deemed to be high quality even though they offer very poor reproduction of sound.

I am often criticized because I downplay the subjective, but I do this for precisely the reasons that you state. The audio media and its jargon allow people to accept poor reproduction as sound quality because "it sounds good to me" - a position that cannot be argued with. But remove the subjective and its not so easy to support this erroneous point of view - in fact its pretty much impossible. It's the bottom line "I know what I like" that allows for the situation that you describe.

In reality I am a passionate listener, I just will not argue audio from this standpoint since the trump card can then be played at any time. I don't like to play when the cards are stacked against me.

Mige0 - AAC really is a much better codec than MP3. Your comparison is not fair or accurate. Dolby Labs is a very respectable organization who do first rate work.
 
I have gradually realized, with something of a sinking sensation, that most audiophiles are genuinely unaware of what a highly colored loudspeaker sounds like.

I've had the same sinking feeling. It's especially disturbing when parts of the system have potential, are even excellent, only to be demolished by absurd mismatches to the other components.

In fact I sometimes have epiphanies that show that run of the mill solid engineering can produce substantially better results than audiophilia. For instance some ceiling speakers in stores can have quite decent reproduction of the essentials: great, certainly not, but at least, not absurdly distorted.
 
observable pontification

Not sure I totally agree with the Boyk model (or parts therof) wherein ultrasonics out well beyond human audibility have an impact on what we actually hear in the 20 - 20Khz alledged human range (or ~25 - 14K for myself;) ), notwithstanding generation of beat frequencies between ultrasonic components in a live performance influencing the timbre of instruments when heard. If those in-band frequencies are there, they will be captured w/o the need to reproduce the ultrasonic signals per se.

Not all of us "audiofools" are gullible enough to buy into the Julian Hirsch model of review, wherein all speakers are, to paraphrase Garrison Keilor's Lake Wobegon Days' "all the drivers are strong, all the cabinets are good looking, and all the technologies are above average"... often followed on the opposing page with a large ad for the device being reviewed.

Having played several instruments (reed) in orchestral works, listened to many more in person, and owned or listened to hundreds of systems over several decades, I personally have never been under the illusion that what I hear from a system is close to what I experience at a live event. Two different beasts altogether.. always will be. I have managed to build systems that reproduce some of the "you are there" attributes, (see pic previous post) sometimes even sounding subjectively better than a live performance ever will.

Having been involved in electrochemical and aerospace engineering for 35+ years, all I can say with bemusement is the level of "ingineering" exhibited in many (especially loudspeaker) audio components is laughable at best.

It's almost as bad as dealing with the health insurance morass we're all currently living through (my current endeavor :bawling:)
with respect to accountability and accuracy.

John L.