4 Speaker Systems

double click to enlarge pix
I have 4 sound systems in use- that probably sounds a bit excessive but as a passionate audio enthusiast/manufacturer that specialises in preamps and power amps it is necessary to have a range of speaker/systems  that generally covers the type of sounds my customers listen to.

No speaker is perfect, but you can find a type that suits you more than another- most people are happy with a well sorted out multi driver moving coil speaker (eg Sonus Faber) Some like to take this a little further with an active version (eg ATC)

Others like the sound of planar drivers either electrostatics (eg Martin Logan)or ribbons (eg Magneplanar)

Although one of the most ancient designs, and very tricky to get right, the sound of horns is very appealing to those dedicated to the principle. (eg Avantegard)

Somewhat related to the horn , the coaxial , or full range  driver is also hard to get right despite the fact it is a beautifully simple concept (eg Tannoy)

There are other exotic types of speaker drivers but these 4 concepts are the most widely used.

I’ve spent close to 40 years building, owning , listening to and enjoying many different versions  of these 4 types of speakers and after all this time I still don’t have a “favourite” type of speaker- there are aspects of all of them I like and dislike. I gave up trying to find the perfect speaker many years ago and have been happy to enjoy the strengths and ignore the weaknesses of all types, and have had a lot of satisfaction getting the best out of each of my many speaker systems over the years.

I’ve found that in my role of amplifier designer and builder that it can be quite perilous to commit to one type of speaker as the designs tend to then hone in on overcoming the weakness of that particular speaker (none are perfect)

You might build a preamp or power amp that is fabulous with a horn speaker but less than that with a moving coil driver, and vice versa.

Eventually the only really successful amplification is that which is accurate under all conditions, into any load- if it appears to be “coloured” or any deviation from flat response it is then the speaker responsible, and recognised as such.

Having an arsenal of 3 or 4 different types of speakers allows one to really come to grips with the strengths and weaknesses of each speaker concept and is invaluable in designing and building amplification that can either specifically suit each type of speaker or find the best design compromises that enable each type of speaker to sound as satisfying as is possible.

In no particular order of merit my speaker/systems are:

ELECTROSTATIC/RIBBON HYBRID

(1) Modified Martin Logan Request electrostatics.  I listen to these the most as they have the best acoustical position in room and are closest to bridging the gap between technical accuracy and subjective listening pleasure. They have a big open sound that is reminiscent of what you would hear if you were in a jazz club or even in a symphonic concert- the sound is very wide with good depth presentation.
 The very high level of accuracy is due to the implementation of these speakers- they have had the passive crossovers removed , and are used as actives with seperate amplifiers directly driving the step-up electrostatic transformers.
They also have multi  line-array ribbon drivers (7 per side) covering the response from 200Hz-500Hz. I use these as I prefer the sound of them to the electrostatics in the range of upper bass- lower midrange. And they help with the integration of the bass- which is always difficult with hybrid ‘stats. If a big bass driver is used it will have good lower bass but struggle to blend seamlessly in around the crossover point, and if a small bass driver is used  it will struggle to do deep tight bass. This has always been a problem with hybrid electrostatics and the ML Requests I use have a 12” bass driver in a sealed cabinet which gives great bass, but has a real problem at the higher frequencies at the crossover point.

The ribbon drivers cover this upper bass to lower midrange region beautifully and seam effortlessly with both the bass driver and the ‘stats. 
A DEQX digital crossover is used with filters of 24dB, amplitude and filter type is easily adjusted at the listening position with Windows7.  The DEQX has the optional Jensen output transformers which  sound very good.  There is a bit of bass eq to compensate for the bass roll off of sealed design, bass is very tight and deep enough for no need of subwoofers.

Power amplification is all solid state- 6 channels of serious power but also load invariant, which means the amps can drive the speakers accurately through the very big impedance swings of the electrostatics.  The amps output 200W/8 ohms and 1000W/1 ohm. Tube amps can drive ML’s but not accurately.

Naturally a Supratek preamp is used for controlling all the inputs and presenting them to the DEQX at the correct level and no negative impedance effects.  A Grange preamp with 6H8C/101D DHT triode linestage and E180F, 7308(6922) and 6E6P phono  stage .
Every system should have some tubes in it and the best way to use them is in a preamp that has low output impedance – with a really good design this results in the all the magic and positive attributes of  tubes  without any negatives (the slow, warm sound of traditional old style tube design)

Sources  are phono from a Luxman PD-444 direct drive turntable , with two arms , a Dynavector DV505 arm and a Ortofon SPU cartridge modified by Expert Stylus in London (fantastico!)and a FR-64 with DL-103 cartridge, again modified by Expert Stylus . They replace the cantilevers and diamonds with the very best available and I cant speak highly enough of their work.

Other source is computer audio based – Windows 7 , JRiver Media Centre , exaU2I usb-I2S convertor, Sabre 32 dac, capable of 32bit/384Khz. Files stored on external hard drive- see other blogs for details.

The sound of this system is very dynamic, it can play at very loud levels and stay clean and distortion free, whilst retaining its ability to resolve micro-dynamics.  It loves concert and live recordings as it expresses the sense of aliveness and “being there” in spades.

It is very capable both as an analytical tool and as a source of listening enjoyment. It surprises most people when they hear it as it doesn’t sound like what they are used to,  and it does take some acclimatising to, but I always find my other speakers lacking it’s sense of life and realism.

And with computer control of just about every parameter it can be changed in a instant to suit any mood or recording.

It’s only fault is a very tight sweet spot that makes it a one person speaker.

MOVING COIL  LOUDSPEAKER-PASSIVE CROSSOVER

I have a friend who has a mastering studio- it’s a relatively small room but built solely to sound good(accurate). It has had a lot of acoustical engineering and he’s a clever guy who can understand the considerable physics involved with acoustical engineering. And he’s pretty good with a computer and programs which is essential these days.

The first time I heard this room I was blown away by the sound- I had never heard such good imaging and clarity of tone. I realized most of the magic was coming from the room, but I also realized how good his studio monitors were and I had to get a pair.

The Lipinski L505 is a curious beast- it is a designed to be a full blown studio monitor, which are usually ruthlessly accurate, but this monitor has none of the aggressiveness usually experienced with studio monitors. It’s fantastic imaging properties really shine through and even though it remains very accurate, it has a 3D imaging ability that simply gives a sense of realism that is quite breathtaking, especially in a near perfect listening room. 

This  quote from the Stereophile review:
"While the Lipinski L-707s displayed excellent imaging, extended dynamic range, and translucent mids and highs, their strongest characteristic was their wide, deep soundstage, with an unusual level of spatial resolution for individual orchestral instruments and choral voices." 

Taken from the Lipinski specs we can look  for clues as to why they sound so good.

L-505 Monitor

1. A sealed enclosure tuned for the best impulse response rather than a low-end extension. We strongly believe that various kinds of vented enclosures always cause coloration that does not deliver reference accuracy.

2. Sturdy enclosure made of 1" (25mm) thick MDF with internal bracing to bring unwanted resonance and box coloration to an almost nonexistent level.

3. Stiff Glass Fiber Cone mid-woofers with low dampening rubber surround, diecast chassis, and a low distortion magnet.

4. Neodymium Ring Radiator - super low distortion tweeters with frequency response up to 40kHz and extra wide dispersion.

5. The absence of a grill in front of the tweeter. Our laboratory and listening tests prove that even the most transparent grill fabric causes high frequency comb filtering.

6. An acoustic, rather than an electronic, time-coherence of tweeters.

7. Unique tweeter surrounding. Perfecting the tweeter environment appeared quite possibly to be the most challenging aspect of the design. The goal was to eliminate the edging effect by carefully shaping the tweeter surrounding with specially designed Belgian foam.

Patent pending.

8. A low-order crossover for the best phase response. We also selected the lowest possible crossover point to deliver the best transient response.

9. Premium quality crossover parts. Foil Inductors wound on a wooden core - with their skin effect conductivity - provide performance unmatched by less expensive wire wound coils, and provide much lower coloration than iron core inductors. We also use non-inductive resistors, premium audiophile grade capacitors and special flat, Super OFC internal wires.

10. Matching internal dampening design. Internal dampening is outsourced and precut to absolutely identical shape and weight for the best speaker-to-speaker consistency.

11. Dual, all-brass, gold-plated posts, which accept banana plugs or up to 2 AWG wires.

12. Magnetic Shielding. All our speakers are magnetically shielded for demanding pro video applications.

13. All parameters perfected in anechoic chamber.

I believe the important points are 1,2,6,7 and 8. My own conclusions about loudspeakers is that phase performance is the key to good imaging, soundstaging and that 3D ability to convey width and depth, that delivers the vividness and coherence of a good recording.

Combined with the Lipinski’s inherent accuracy this totals up to a very good moving coil loudspeaker. Somewhat in the mould of the Dunlavey range of monitors, still very highly regarded- anyone got a Sovereign they want to sell?

Unfortunately the L150 subwoofers I bought with the L505 aren’t up to the same level of performance and they have been replaced with a set of JBL active subs.

Pre-amplification is a Supratek Cabernet DHT preamp with 6H8C and 71A dht triode.

The speakers are bi-amped with either a 4 channel load invariant solid state amp or 4 monoblock KT88 100 watters with adjustable positive feedback.

The solid state amps are used for critical/analytical listening and the tube amps rarely when a more romantic , perhaps “musical" sound is fancied.

Input is a MacMini – either PureMusic,Decibel or Audivarna file players, Audiophilleo1 usb-spdif convertor into a Lavrey DA 10 dac. Files are stored in external hard drive.

MOVING COIL LOUDSPEAKER-FULLY ACTIVE

While the Lipinski could be considered to be state of the art for the early 21^st century the AR-LST was considered state of the art back in the late 20^th century.(Laboratory Standard Transducer)

A very unusual speaker, although designed primarily as a studio monitor it was also used as a very high end audiophile speaker .

With 9 drivers per speaker -4 dome midrange, 4 dome tweeters, and 1 10” bass driver  it is capable of a very dynamic performance with a very wide soundstage , at the cost of very high power requirements.

 I was fortunate in that I was able to purchase 2 pairs and stacked as a D'Appolito configuration the sound is very similar to a big stat or ribbon transducer, with very serious bass. The 10" AR bass driver is a beauty, done properly it is very hard to beat the bass of a sealed box.

The only downfall is that one pair of LST is a serious load for any power amp and two pairs require an amplifier that is unique and not so easily available, and you can forget about tube amps completely.

Most of the problem comes from the  complex passive crossovers in the LST, so to bring these speakers into the 21^st century I decided to remove the passives completely and go all active.

This is no small job as the crossovers are quite large and bulky and there was a lot of wiring , however with the crossovers out and a digital active crossover inserted between the preamp and power amps I was keen to see if  I see if I had enhanced or destroyed two pairs of perfectly good speakers.

The first thing I noticed was that the drive requirements were greatly eased- 6 channels of 200 watt each was plenty to drive the two pairs of speakers with the bass, mid and tweeter sections all paralleled . I suspect that 6 channels of strong tube amplification could also work reasonably well.

However replicating the original crossover frequencies and slopes wasn’t entirely satisfactory- 3 way active speakers can take a very long time to dial in as the respective amplitudes, crossover frequencies and filter slopes present a multitude of choices – it can take years  to get it right, and I’m still on the trail to success. 

However it is very apparent that dynamics are much improved, these are explosive speakers  that can output a lot of clean , undistorted power- great for big symphonic recitals.

COAXIAL OR FULL RANGE SPEAKER

Always coloured, only marginally accurate, coaxials tantalise and seduce, but always leave you wondering if there is “more”.  I’ve owned Goodmans Axiom 80’s and Triaxioms, Altec 604, Corals, Lowthers, Fostex, in fact my hi fi enthusiasm started with a pair of Altec 604B and Trimax EL34 UL amps.

I’ve also owned one of the small 10” Tannoy coaxial monitors, but again, like all full range drivers the lack of consistency in the frequency response had me move on to something else.

Recently I heard the 15” Tannoy Gold Monitor drivers and the story of how I finally found a speaker that is happy with tube amps AND is reasonably accurate is here:
http://supratekaudio.blogspot.com/2011/07/3500-for-pair-of-speaker-drivers.html 

HORNS

While I'm here, lets mention horns- like coaxials capable of very good sound, but so damn hard to get right- getting them to seam properly with good integration is very difficult.
This system cost in excess of half a million dollars (and powered by Supratek amps) and is the most impressive system I've heard, not in terms of accuracy, but incredibly dynamic. The bass horn went out the wall and 30 foot long!

Something a bit more affordable for most of us, the best violin reproduction I've heard came from Martin Seddons Azurra horn set up. It seems his horns are just about perfect at the response needed for violins, and you would swear that Nigel Kennedy was in the room. Martin has a very good sounding room, and I think that contributes quite heavily to the sound.
At the end of the day though, horns are just too uneven and disjointed for me, however people with less technical requirements could be very happy with them.

Art Pepper 7 LP Vinyl set

This recording was made using multi-microphones in a straight stereo mix without any noise reduction system, limiting, compression or equalization on either the location or on the transfer of the original master tapes to the lacquers from which the final pressings are made.
Great recording, fantastic conversion to vinyl.
Definitely recommended for both the sound and music performances.
http://www.purepleasurerecords.com

Also enjoying Junior Wells Hoodo Man on Delmark Stereo LP, the sound on this record is extraordinary.

More on Accuracy

So far the response to this blog has been very positive- thanks to those who have appreciated my ramblings thus far.
As I predicted the only negative feedback has been from those who don't value accuracy as the highest aim of "hi-fi", and don't agree with my philosophy on this subject.
The following excerpt is from a review that Jonathon Valin recently did on an expensive Magico speaker.
He is a much more eloquent writer than I am and his piece on accuracy which sets up the review is very interesting. From the Absolute Sound July2011.

Jonathan Valin

If it does nothing else (and it does plenty else), the

Magico Q5—the current top-line, full-range, fourway

dynamic loudspeaker from the Berkeley-based

company that has, over the last four years, shaken up the

status quo in the ultra-high end—cuts straight to the core

of what we mean when we say something is a “highfidelity”

component.

This is the very issue that led to the foundation of this

magazine, and the position that Harry Pearson staked out

almost forty years ago has been a beacon and a bone of

contention ever since. Should “high fidelity” components,

as HP argued, aim to reproduce the sound of acoustic

(i.e., unamplified) instruments as they are heard in life in

a concert or recital hall? Or, in a significant variant of the

absolute sound approach, should they reproduce precisely

what was recorded on the disc, whether that sounds

like the absolute sound (as it ideally should) or not? Or

should they aim at something else again, something far

less prescriptive and more personal? Should they simply

(or perhaps not so simply) consistently please whoever

listens to them?

Although these views aren’t mutually exclusive, over

the years they have typically been cast as if they were,

as if they represented opposing sides in a never-ending

battle between the forces of “realism,” “accuracy,”

and “musicality.” All three positions are rife with

contradictions, all three share certain patches of common

ground, and all three have been “shaped,” like battlefields,

to reflect the prejudices of individual reviewers and

listeners. The absolute sound school, for example, has

trouble dealing with amplified music, such as rock ’n’

roll, which in today’s world makes its proponents seem

old-fogeyish. After all, what is the “absolute sound” of

a Fender Stratocaster or Telecaster? By the same token,

will a speaker that delivers the whomp of a Fender

Precision bass guitar as it sounds at a rock concert via

a Marshall stack also do justice to the pitches, timbres,

and dynamics of an unamplified cello or doublebass? For

that matter, will an “accurate” system tend to make both

Fender bass and cello sound a bit too cold and analytical,

like an unretouched glamour shot?

There is no single answer to these (and a zillion other)

questions that will satisfy all music lovers, which is

precisely why I try to take the biases of different kinds

of listeners into account whenever I write a review. The

way I see it most of us fall into one of three basic groups:

what I call the “absolute sound” listeners (who prefer

music played by acoustical instruments recorded in a real

space, and gear that makes those instruments—no matter

how well or poorly they were recorded—sound more like

“the real thing”); the “fidelity to mastertapes” listeners

(who want their music, acoustical or electronic, to sound

exactly as good or as bad, as lifelike or as phony as the

recording, engineering, and mastering allow); and the “as

you like it” listeners (who care less about the absolute

sound of acoustical instruments in a real space or about

fidelity to mastertapes and simply want their music to

sound some form of “good,” which is to say exciting,

beautiful, forgiving, non-fatiguing). Though I think these

groupings are valid, I also think that no listener is purely

one type or another, i.e., the fidelity to mastertapes listener

also wants his music to sound like the real thing, when the

recording allows; the absolute sound listener wants his music

to sound beautiful, when the music or orchestration allows; the

“as you like it” listener puts excitement and beauty ahead

of fidelity to sources, but is not at all unhappy when those

sources also sound like the real thing as he defines it. What

I haven’t been as clear about, perhaps, is where I stand in

this triumvirate—and why.

I stated my opinion on this crucial topic about twenty

years ago when I wrote a book about RCA recordings,

and in spite of occasional forays into other kinds of

listening I haven’t really changed my mind. Since The

RCA Bible has been out of print for a very long time, let

me quote what I had to say way back when:

“How much of the ‘absolute sound’ of an orchestra

does a microphone really capture? Well, it’s a fact that

microphones differ significantly from the response

of the human ear. Throughout the fifties and into the

sixties Mercury Records, for instance, used German

microphones (Telefunken 201’s and Neumann M 50’s)

with a rising high end. Are Mercury’s ‘living presence’

recordings [from Watford Town Hall] actual transcriptions

of the sound of the LSO with Dorati at the helm, or

are they the products of hot mikes—ones that added a

little upper-midrange sheen and bite to the LSO strings,

winds, and brass—or are they some incalculable blend

of both?

 “Well, you’d have to have been at the Watford Town Hall to

know for sure. And even then, you’d have to have been sitting

where the microphones were placed. And since you don’t hear

in three channels mixed down to two and your chair’s not tall

enough to put you where the mike heads were located and your

ears have a different frequency balance and directional pattern

than mikes, you’d be hearing sounds that were different from

those which the microphones recorded. How different? The

question is unanswerable. On the basis of a recording we can

never know what the LSO ‘really’ sounded like on a particular

afternoon, on a particular piece of music. All we can know is

what the tape heads recorded.”

Twenty years on, I stand by what I wrote. For me high fidelity

means fidelity not to the absolute sound and not to some idealized

sound but to the sound of the mastertapes, which still seems to

me to be the one and only “truth” we’ve got. That this truth is

inevitably a compromise that will be further compromised in

playback is simply the way the recording/playback process works.

To achieve high fidelity as I define it means that the

loudspeakers and everything else in the playback chain need to

“disappear” as sound sources. To accomplish this, they must

be neutral, transparent, high in resolution, seamless in top-tobottom

coherence, low in distortion, and capable of a high degree

of realism rather than romance. As beguiling as such things can

sometimes sound, pieces of gear that impose a beauteous or

exciting or forgiving sonic template on the presentation—and,

thus, don’t disappear—are, in spite of any other virtues, finally

not for me. This doesn’t mean that they aren’t or shouldn’t be for

you. I have no argument with friends and colleagues who prefer

a less “neutral” component, either because they think a more

bespoke presentation makes music more like the real thing (as,

for example, those “absolute sound” types who eq their systems

to roll off the treble and/or boost the bass—or who prefer

equipment that effectively does the same thing because of builtin

dips and boosts in frequency response) or because they think

a romantic presentation makes recorded music more attractive

and, well, “musical.”

What I do have an argument with is calling such presentations

“high fidelity.” By my lights anything that makes you more aware

of the way sources are being colored and distorted by your system

is, ipso facto, less of a true high-fidelity component and more of

a tone control. I don’t want to hear my equipment automatically

adding virtues or subtracting flaws from every record (even

from records that benefit by such additions and subtractions); I

want to hear what is on the recording, good, bad, or indifferent,

because, as I just argued, the recording is the one indisputable

truth that stereo systems can be faithful to. The way I see it,

if you’re unhappy with the sound of the LPs and CDs you’re

playing back, then don’t try to correct the problems with your

stereo system. Instead, go out and buy better records.

My position has had certain undeniable consequences when

it comes to the kind of playback gear I prefer and how I set

it up. While as a reviewer I’ve recommended any number of

different kinds of loudspeakers for different kinds of listeners

(and was sincere in these recommendations), as a civilian I’ve

always owned electrostats, planars, and (occasionally) twoways.

Why? Because they were (and in many respects still are)

the lowest-distortion, lowest-coloration, highest-resolution,

most transparent-to-sources, least-present-in-their-own-right

transducers—the “highest-fidelity” speakers, if you will, by my

standard of high fidelity.