Product: Harefield active High Definition True Stereo Monitor
Manufacturer: Digital Audio Systems formerly Dyer Audio Systems
Telephone: (+44)1992 468674 Fax: (+44)1992 467581
Price: Harefield £5000 7400€ (exchange rate October 2006)
Reviewer: Mark Wheeler - TNT UK
Reviewed: October, 2006
I was tipped off about Bill Dyer's latest enterprise by someone who heard his prototypes at the Heathrow HiFi News show in September 2005. From what I was told I guessed that these might be products that should get reviewed for TNT. I have a long relationship with active speakers going back to my formative audio days in the 1970s when I first encountered the principle. As I explained in my two-part treatise on them, I would expect to hear better performance from 3x €1500 stereo power amps driving a €3000 3-way speaker (from a €500 acive x-over) than a €5000 power amp driving the same €3000 3-way via their built in passive x-overs. Same money, same speaker, but I'd expect obvious improvement with the active set-up.
"He's biased" shout plebs, stage left,
"He's bound to give active speakers a better review"
"If they're better I will say so", replieth ye humble scribe, somewhat lamely
The pro-audio-sector jury remains slightly ambivalent on this subject. Most engineers wouldn't subject a critical piece of work to a passive monitor, but about half of studios, globally, are reported to use passive nearfield monitors at some stage in producton. Bill Dyer paid his dues in the 1960s, moving rapidly on from managing a Tottenham Court Road audio emporium (TCR remains the London street of electronics traders to this day, tho rather downmarket now) to Chief Technical Engineer at Mayfair studios. Then a period distributing, installing (& repairing) glamourous imported gear like JBL (and he probably stuffed the JBL pro-driver leaflet into an envelope to this schoolboy back then), MCI consoles and tape-machines, EELA miixing-desks, for Feldon Audio. Bill's installed systems in such studios as EMI; Advision; BBC; Scottish TV; London Weekend TV; MGM films etc and heard the British invasion stars of the 60s & 70s era performing live through those systems.
Bill Dyer also had the temerity to modify gear for better sound, phase problems in mixer line-level transformers being a personal bugbear. Having been one the first to bring the Moog synthesiser to these shores, Bill became very aware of the subtle nuances of waveforms, and their envelopes, in the context of instrument emulations, that all had to be programmed manually in those days. Readers my age will remember prog-rock keyboardists, surrounded by banks of manually operated VCOs, filters & shapers that needed constant adjustement, even mid-note.
So Bill's pedigree invites respect, and implies that he'd know what works. While other men might be sitting back to write their memoirs, or watch cricket on the TV, Bill Dyer has chosen to build speakers as his retirement entertainment. The logo for DAS looks like the symbol for a microphone, connected via amplifying A to the symbol for a speaker driver: cute.
There are 3 basic loudspeaker models in the DAS range, all active, plus a subwoofer. Two models are fairly conventional stand-mount boxes, with forward facing drivers, while the third is a curious floorstander with omnidirectional bass-mid aimed upwards at a diffuser and a forward facing tweeter. This review is of the middle "Harefield" model.
The DAS Harefield achieves a flat frequency response despite Bill stating that this parameter is less important than phase accuracy, low distortion and even dispersion. Dyer knows that when all the other factors are correct the frequency response will follow. 40Hz-20kHz -3dB, and beyond. The low end is not a conventional biggest-bang-for-a-buck alignment, but is an alignment chosen for better phase performancve.
The distortion is where the surprises come:
There are two 50W power-amplifiers in each speaker, made by a Dutch manufacturer. I know nothing more about them, whether they are single ended, push-pull or class D; whether they have oodles of global feedback or hand rolled wires made from mermaids' hair. In my humble opinion amplifier nuances become far less important if they are not fighting to control driver movement through a reactive network. Global feedback, in particular, seems especially sensitive to speaker design in its audible effects, hence the divergence of opinion on the subject.
The balanced XLR connector is terminated by an RF suppression circuit, which precedes the electronically balanced input. There is a level control too. The crossover is a second order filter network, including time-aligment for the drivers. The electronic filters sum with the usual 2nd order natural driver roll-off to produce a 4th order combined response. Fourth-order filters (I'd guess they're probably Linkwitz-Riley) are naturally in-phase at the crossover frequency, and have less overlap, thus improving vertical dispersion, and hence in-room power response regularity. The internal 2 channel amplifier is mounted on aluminium panels that function as a heat sink, and is claimed to reproduce high level transients 'faithfully'.
Bill describes fidelity as meaning "does a kick drum sound like a kick drum...does a human voice sound human?", so the real proof of this pudding will be hearing. Bill notes that he has chosen a bass alignment for accuracy rather than to extend the low end beyond its capabilities, setting a 40Hz box-tuning in an oversize box. With a driver free-air resonance of 39Hz, this would be close to a classic B4 alignment in a smaller cabinet, but the larger box and rectangular 15mm x 250mm port keeps port air velocity low and power-handling high. Combined with the lack of resistive and inductive losses, enabled by the lack of passive-crossovers, this should lead to tighter, better defined bass, without the boom and overhang of alignments of equivalent Q greater-than 0.707.
There is no EQ fix in the crossover, as is so tempting with an active filter. X-over eq-tweaking often damages phase integrity. Bill is determined to make speakers that "reproduce the whole wave-front from bottom to top, with coherent timing, from the moment the orchstra begins to that last dying nuances of ambience". This ties in with his long intersest in Ambisonics, and my task post-review will be to deliver the DAS loudspeakers to a University of Derby researcher who is to use them in place of PMC monitors in Ambisonic experiments.
Such a demand for accuracy is a grandiose task to set. Although this pair is finished in charcoal, silver is available, and being sprayed finishes, anything should be possible. Some real wood veneer would increase DAF (Domestic Acceptance Factor).
Positioning is easy. It does not seem to matter whether they are near-spaced (2m apart) or far-spaced (3.5m apart); toed-in slightly (10°) or toed-out (firing straight down-room); tweeter at ear-height or bass-mid at ear height. The Harefields are much better at least 1 metre from room boundaries. The insensitivity to orientation implies good crossover integration and consistent polar-response, both horizontal and vertical. Vertical polar-response will be helped by the acoustical 4th-order x-over slopes (24dB per octave) because there will be less overlap in driver output each side of the x-over frequency, and hence less lobing due to cancellation & reinforcement. Horizontal dispersion will be dictated by driver choice and cabinet shape and x-over frequency.
A walkaround test establishes that the DAS Harefield also has textbook horizontal dispersion. If dispersion is shown on a 3d plot, x & z axes represent amplitude over a 360° horizontal plane, and y axis represents rising frequency. The plot should look like a cyclinder from Fs to baffle-step frequency (baffle width = half-wavelength), then gradually becomes a cone shape, lobing in front of the driver. A single small-diameter moving-coil driver usually achieves this and some well-designed multi-way (moving-coil) systems also manage it. It makes speaker placement easier and allows for more versatile seating with a much broader sweet-spot in the room.
With no obvious quirks around the x-over point, the in room power response resembles that of the on axis frequency response. The LEDR test confirms this with a good symetrical result. I find the LEDR test, available on various set-up cds, including HiFi News Test Disc III, a very useful set-up check. I apply this test, after trial and error with familiar music, to confirm speaker positioning.
I assumed the Harefields would excel on my heavy stands. These are usually used as inert platforms for testing prototypes. These heavyweight monsters were originally built by a friend to support some speakers I built for him in about '86. They comprise a mixture of welded and epoxy-glued scaffolding legs filled with kiln-dried sand, between a pair of 6mm mild-steel plates, through-bolted, and glued, to 25mm birch-ply. The top and bottom plates have 4 adjustable steel spikes, fitted before I learned that 3 usually sound better. I am wrong. After several listening sessions, I realise that the Harefileds work better on my old open-frame Origin Live tripods, despite them not being sized correctly for the Harefields.
The Harefields are distinctly affected by their support. They are fitted with proprietry instrument feet; 4 stick-on chunks of plastic. These look out-of-place on a high-end audio product.
The 4 nasty plastic feet also sound out-of-place on a high-end audio product. I try the Harefields on 3 and 4 points of contact and they are marginally better all-round on 3, as usual. The Origin Live stands facilitate this experimentation by insertion of machined-steel discs, on top of any combination of the 5 upward pointing welded spikes. I had this option incorporated for experimental reasons. Welded spikes are also a better energy path than adjustable screwed spikes.
Further subtle improvements are wrought when 2 PolyCrystal Point-Discs are substitued between the spikes and the front of the Harefields, in place of the Origin Live plain machined-steel cups, and a Polychrystal footer at the centre-rear. The difference between this set-up and the standard plastic instrument feet is as big as a tweeter change.
The Harefield manage a clean 97dB at the listening seat in my big 5.5m x5.5m x3m (average lengths) room. The bass drivers go beyond compression into gap-jumping overload distortion by 104dB at the listening position. This equates to 102dB clean, and 110dB distressingly distorted, at 1m with my hand held sound level meter set to unweighted frequency response. This is VERY LOUD.
The low distortion of these speakers permits much higher levels than typical domestic loudspeakers whose iron-dust inductors reach saturation, often with a strange cracking sound.
Balanced feed is supplied in my system by an IAG built multi-tapped variable Sowter transformer, operated by a Shallco switch. This is heresy to Bill Dyer, who regards line-level transformers as a bad thing, causing unecessary hf phase innacuracies. My experience of my own system is that this is the most transparent single-ended to balanced adaptation I have tried, with far smaller consequences than phase-splitting by electronic means.
The systems comprises:
Bill notes that most so-called stereo is just two channel panned mono. A single microphone feeds a
single mixer channel where the engineer pan-pots the signal to two channels, the difference in amplitude
selected on the pan-pot supposedly positioning a virtual position between the speakers.
"That is not stereo", says Bill, "It is biphonic mono".
"Stereo is greek for solid", reminds Bill, "not for 2 channel".
A stereo signal should be created with two microphones preserving the phase information. If this phase information maintains its integrity through the recording and playback chain, the retrieval of ambience will be much better. He demonstrates this to me during his visit with some recordings he has brought along, made by Mike Skeet of 'garage door' fame (the famous test track on the first Hi-Fi News test cd, confusingly anotated cd0003) using a Soundfield microphone.
The ambience clues are as good as I have heard, not just the discrete reflections of room boundaries that a high-end system can reproduce (very noticable in Chesky jazz recordings), but a sense of the scale of the space. Each instrumental or vocal source creating a vague impression of its dispersion into the soundscape. This is not the daft delusion of 'stereo image', which I have heard from expensive gramophones, but never in an unamplified concert-space. I recall listenning to the Delta Saxophone Quartet in the Djanogly Room at Notingham University Arts Centre and being aware that the audience-in-the-round would each hear a very different performance if audiophilia is to be believed, but that when one's eyes close the sound hovers around and above each performer more like a translucent hologram than a stereo image. Playing their cd through these monitors does not create the same impression, but probably because of the way it was recorded.
The distortions in passive crossovers have the irritating quality of being non-linear and higher-order than those of linear electronics. Active loudspeakers don't sound as loud as passive speakers at the same measured spl because of this. You'll only realise just how loud these speakers are playing when you try to hold a conversation. It took me a long time to re-adapt to passive speakers after 5 years of living with my own home-made passive speakers some years ago. I kept wanting to turn the volume down with passive systems, while I turn active systems up until that sweet volume every recording has that sounds just right.
Drums thwack convincingly with the Harefields. 'Tangible' describes the soundfield they project. These speakers portray what seems like a big, firm, soundstage. No delusions of pinpoint images, just the sense of where the performer and their instrument are positioned.
True mono, recorded in mono, not blended stereo, is a tough test for any speaker. I always voice my own projects in mono with a single speaker. Mono recordings through a pair of speakers should produce the illusion of a central source midway between the speakers. The impression should not change in size at different frequecies, nor wander about with different freqencies. Image wanderings from side to side indicate unmatched speakers, and from front to back indicate poor phase performance. A late 50s Westminster recording of Dvorak's piano trios illustrates these strengths superbly via the Aqvos 2CI where the central mono source also portrays the ambience of the recording space so well that for a moment this listener does not miss stereo at all.
The impression of ambience, of the concert space, or the recording space is very fine with the DAS Harefield monitors.
The 2001 Grateful Dead, Birth of the Dead, double HDCD of 1965-6 recordings or their earliest incarnation as the Warlocks (although the band were so new that a studio tape box names them as "The emergency Crew") were probably laid down using similar JBL monitors to those Bill Dyer was importing into the UK in the mid 60s. The transparency of the DAS Harefield provides an open window into this mixed bag of recordings, some mono & some stereo. There is a simplicity to these arangements of traditional bluegrass tunes and their newly penned Dead originals that the Harefields effectively communicate. And the HDCD qualities are there too, obvious when switching.
The small long-throw bass-mid driver is really working hard to shift so much air at high spls (sound pressure levels). A big room like mine would benefit from a 200mm driver, but then the crossover point would have to be lower to maintain good dispersion from bass-mid driver to tweeter. The soft-dome tweeter has the usual hallmarks of the breed, higher frequencies are maintained by a series of controlled resonances, although much better than plastic soft-domes. I'd love to hear what Bill could do with a larger lf driver and a high-tech tweeter, as this is the best sound I've heard from any system using drivers like these.
The domestic market probably demands a less industrial finish too. An extra-cost option of real-wood veneer would help find homes for many more pairs of Harefields than the durable charcoal spray finish in which my samples are clothed.
The time Bill spent analysing wavefronts of sounds so that they could be more accurately mimicked by the Moog synthesisers he was selling impressed upon him what a complex wavefront looks like on an oscilloscope. Bill has persued the reproduction of these blocks of acoustic energy, a simple kick-drum, as simple as a single-skin 18" Ludwig, produces a complex envelope of sound that looks more like inhospitable mountainous terrain than something apparently so straightforward as a single blow to a single membrane. Bill has succeded in making a loudspeaker system (including inbuilt amplifiers) that mimic those waveforms as well as anything I know of at £5000 (####€). If I didn't build my own loudspeakers I would reach for my chequebook.
Having a fairly even off-axis frequency balance they were far less sensitive to position and toe-in than most domestic loudspeakers. This scores plus-points for domestic acceptibility and neatly balances the minus-points earned by the industrial finish. The wide listening area renders them more sociable than the high-end sweet-spot designs, but the high-end sweet-spot magic is also irrelevant here. These are accurate, low distortion, no-nonsense compact 2-way nearfield monitors that portray instrumental timbre and recorded ambience accurately.
The DAS Harefields do exactly what they promise. They go loud. They do it clean. You get exactly what the engineer put into the recording, within the physical limits of the box size and drivers.
At the 2005 Heathrow HiFi Show, a studio engineer, who wandered into the DAS room, was heard to remark
Although not quite as elegant or subtle as Proust's Madeleines' olfactory stimulation of reminiscence, this sums up the DAS Harefields better than any of my remarks.
In Part 2 I review the smaller D.A.S Hailey active loudspeakers.
© Copyright 2006 Mark Wheeler - www.tnt-audio.com