Harbeth vs. 'accepted wisdom'

[jair44]

Hi Alan,

After reading the article (Hi-Fi Choice 01/08) 'The World According to Harbeth' I was intrigued enough to find out more about your products. Strangely enough what I discovered was that Harbeth seem to go against quite a bit of the received wisdom in some audio circles. For example there has been a trend in the industry to go for heavier cabinets, minimalist crossovers, easier amplifier loads, narrower thicker baffles etc. There are some who also believe in the superiority of paper (pref undoped) to plastic for cone materials. On personal experience I find that paper sounds more lifelike than polypropylene which is why car stereos and portable radios often sound more natural in tone than full blown hi-fi systems. Plastic seems to give an increase in smoothness at the expense of tonal colour in sound. Paper is prone to breakup / environmental conditions but polypropylene is boring once you become accustomed to the blandness in the sound.

Anyway Alan I would love to hear your views on the differences between Radial and paper as cone materials before I can get to a demo of the SLH5s. I have a suspicion that in the last 30 years or so we have generally traded in warmth and colour for detail and speed which is why a lot of audio equipment is impressive but not satisfying.

[A.S.]

... Strangely enough what I discovered was that Harbeth seem to go against quite a bit of the received wisdom in some audio circles. For example there has been a trend in the industry to go for heavier cabinets, minimalist crossovers, easier amplifier loads, narrower thicker baffles etc. ....

We use technology and techniques that have been proven over the past 50 years by two generations of engineers (at the BBC) to make satisfying music at home and in the studio. It is of no concern or interest to me at all what other manufacturers do, nor how they think or go about solving engineering problems, nor what the current fashions trends are, nor what hi-fi journalist write based on a brief encounter with us. We live it, hone it, nurture it every waking moment and we can draw on a deep pool of accumulated knowledge that is right for us and our customers, but not necessarily right for other markets and other manufacturers. All our design work is minutely documented, step by step in daily Log Books for the next generation of designers to further improve what we've done after I retire. This insistence on record keeping by BBC engineering staff has left a legacy of Research Papers, many in the public domain, and I encourage you to scrutinise the BBC archives yourself and judge their merit. Personally, I think they are logically faultless and still a great inspiration. There is still much to do to move the frontiers forward.

The issues you raise would need many hours of discussion to chew over the pros and cons, since it's important to appreciate that there is no ultimate technical solution to speaker design. It is vital to understand what the designer's brief was, and what tools and tricks he had available to him. And crucially, what constraints he was under - who drove the design process? Engineering or marketing? How much time was available? What target costs did he have to meet? How did he have to differentiate his product from others? Did he design by ear, or measurement or both? Did he use speech as the ultimate arbiter of design integrity? Was the design by committee or was one individual solely responsible from start to finish without any outside influence? We don't have any of these those worries: I work alone and make all these decisions myself.

So, briefly:

- heavy rigid cabinets store energy and release it where its audible and/or sound dry and cold

- minimalist crossovers utterly miss the point that every drive unit only works optimally in a certain range and a simple crossover lets too much unwanted out-of-band signal bleed through into the wanted band. In my opinion, minimalist crossovers are totally wrong. I have never encountered a drive unit that could really be at its best with a simple crossover. Also interesting to note that those minimalist crossovers are often used to drive cheap, mass produced woofers and tweeters, further compounding the logical error of using simple networks.

- Easier amplifiers load. Guess that you have not read our comments here that Harbeths are intentionally an easy amplifier load? We fully agree that an easy - we say Universal - load is a good design target and always has been for us.

- Baffle width etc. etc. - has no direct bearing on good/bad sound quality at all. Just styling.

- Cone material. It defines the sound of the speaker more than any other factor. Paper can be a good material, but its production is a mysterious black art. As you may have read, we are not fans of polypropylene, even though it was discovered and patented for speaker cones by our founder, Dudley Harwood. Our RADIAL material has been the cornerstone of the Harbeth 'sound' for nearly 15 years. It just works.

Any of the BBC designs precede todays Harbeths - going right back to the war - are still exceptional in their own way. That proves the integrity of the original design concepts. Physics doesn't change in two generations - if ever.

[shseto]

the most famous accepted wisdom that I believe in is that don't belive in what others say, go for an audition/ demo and let you ears/ brain do the decision.

[Shutterbox]

the most famous accepted wisdom that I believe in is that don't belive in what others say, go for an audition/ demo and let you ears/ brain do the decision.

This, I gotta agree with 2 arms up, only decisions are not based solely on initial impressions.

[sglim]

In hi-fi there is just too many teaching and school-of-thought, we are overloaded with information, the moment we are about to made up our mind someone suggested there is something better out there it is a tiring exercise just to buy a pair of speakers!!!

I bought a pair of Compact7-ES3 last week, I can?t be happier. This is the best upgrade ever. Yes, go for an audition; don?t just test on good recording try some ?ordinary? recording as well.

[A.S.]

...only decisions are not based solely on initial impressions.

I think that this means be cautious about initial impressions. I agree with this. The way our ear/brain works (and our eyes too) is that just like the magpies in my garden, evolution has programmed us to be attracted to bright, shiny new things. I can guess that many inexperienced listeners, sitting for the first time in a hi-fi demonstration room (an intimidating experience) will be thoroughly confused by the loudspeakers they are presented with. Thinking back to my experiences in this situation, I can recall that I had these feelings ....

1. Shock that no speaker exactly recreates my experience in the concert hall or at home in familiar surroundings with real musical instruments ... which leads to ....

2. Semi-panic that there is something wrong with me or my hearing ... then to ...

3. Needing some guidance from the demonstrator who has seen this reaction many times before even though I really wanted to resolve this selection process entirely on my own without outside help or influence ...

4. {{CRITICAL POINT IN THE DEMO}} The demonstrator now has the power over me to kindly and objectively explain by example why speakers sound different to each other and to draw-out their strengths and weaknesses in a constructive way (there is no such thing as a perfect speaker) according to my expectations of how music should sound at home - or he can steer me towards the speakers that he wants to sell to me.

5. Under these conditions, the soft, sweet, warm Harbeth sound probably will not jump out and say 'buy me'. We are the pastel shades of Turner in contrast to the vivid, exciting colours of Warhol. But - living with a Harbeth is easy, day in day out, year in year out. It draws you in. It sounds more and more 'right' as you listen just as you accept the sound of a musical instrument for what it it.

A good dealer knows this, as he makes the time for you to reach the right conclusion as and when you are ready. You may need several listening sessions. Maybe you long for a Warhol - that's OK! - there are innumerable suppliers for you to chose from, but just look at the wonderful fine brush-stroke detail and colour that a Harbeth offers if you are willing to sit and really listen. Less initially impressive and ear-catching but a much more satisfying long term investment* that you want to return to again and again.

*Harbeth speakers are not consumer durables with a short life span. They are, in fact, capital equipment, earning their keep over 20+ years disguised as consumer goods! That's why we sell the very same speakers to professionals (esp. the BBC) who write-down their investment in their accounts over the very long term. Proof that their Harbeths are assets not costs.

[A.S.]

We all know that the Harbeth sound is soft, warm, extremely detailed, easy on the ears, natural etc. and sound like real music, not hi-fi. But how unique is it? Can it be bought for less from another brand? How close do other brands come to the Harbeth sound?

I think that the answer is truthfully this: to the casual observer not interested in audio, probably all loudspeakers sound the same. To them, all TV's look the same, all cars drive the same and all cameras take the same pictures. But we know that once you own Harbeth's that's the end of the hunger for transparency and low-level detail. Now, what would another designer or brand have to do if they wanted to copy the Harbeth performance? Is there a single magic button they could press to transform the ordinary speaker into the extraordinary one? The answer is no - because if it was that simple, they would have surely done it 13 years ago when we launched the Harbeth RADIAL cone.

So, a Harbeth is a holistic solution and is the sum-total of these (and more) key ingredients. They all have to dovetail together harmoniously.

- Advanced cone technology in Harbeth designed and made woofer and midrange drivers
- Thin wall, critically damped, rectangular traditional-BBC cabinet, removable front / back
- Harbeth SuperGrille
- Complex multi-function crossover
- Designed for optimum performance at moderate listening levels (this is a crucial design factor)
- Designed by extensive computer simulation plus exhaustive listening tests on real music
- Designed as professional instrument that happens to find a home market
- Voiced and balanced by one person who hands the prototype to production only when he is totally satisfied that the design is market-worthy
- Optimally damped throughout the speaker's signal path to eliminate ringing and coloration
- Constant awareness during design that a Harbeth must be sweet and detailed together
- Rejection of current audio fashions in any part of the design as they introduce unwanted harshness
- 50+ year heritage of carefully documented BBC speaker research to draw on constantly throughout the design
- Daily, hourly, continuous use of the most sophisticated audio analysis equipment to augment (and explain) sonic characteristics and to keep design 'grounded' in science
- Recognition that the truly great audio researchers have passed on - we need only continue their pioneering work by ever more careful study of their legacy to make great loudspeakers
- Commitment to a business, audio, ethical and social responsibility unchanged by time

Pick only one or two features from the list and you will not have a Harbeth - a Harbeth is a combination of all of these. As for the future - is there still more to discover from our BBC legacy? Absolutely. The more I study of the breadcrumbs left by our predecessors the more clues I see for the way forward but, the relatively easy solutions (like RADIAL) we've already implemented. The challenge is the next generation of technical progress which must not in any way disturb the famous Harbeth sound. That's why I am so cautious. It must truly be progress not change for the sake of change.

[A.S.]

It is with some interest that I note that Harbeth is now seemingly the only UK speaker manufacturer (and probably the only speaker manufacturer is the world) still using the original BBC-researched thin-wall cabinet system. We do so - regardless of cost - because thin-wall produces a beautiful sound as it allows us to tune the cabinet panels just as we want. This is the same technique that demands that musical instruments are made from thin, tunable panels not CNC cut MDF. Thick, glued MDF or plywood cannot be tuned. But - a big but - thin-wall is very expensive to produce as it demands a very high standard of skill from the cabinet maker. The cabinet parts are fiddly and easily damaged; panels must be veneered both inside and out to prevent warping, the fronts and backs must be made and fitted as separate processes - it all adds to the cost.

Our founder, Dudley Harwood as head of acoustics at the BBC thoroughly researched the subject of what makes the best cabinet system and concluded that, perhaps surprisingly, when sound quality is put first, the thin wall can not be bettered. You can read the 29 page original BBC research department report from 1977, the same year that he retired from the BBC and formed the Harbeth company here with the revolutionary polypropylene-coned HL Monitor.

We will continue with thin-wall because it provides our critical customers with a natural, easy to live with warm sound. We believe that our customers put sound quality above cost reduction.

HOW TO IDENTIFY A THIN-WALL SPEAKER CABINET

(A) Measure the thickness of the side/top panels

(B) Look inside for the damping counterlayer - see attached - usually bitumen or similar grey/black rubbery sheet glued and/or stapled to the side, top, bottom and back. It's very expensive (oil based) but soaks up resonances as micro-heat (refer to Harwood's paper). There is no known substitute for this damping layer. Thicker (MDF) panels are not a substitute - thicker MDF increases not decrease audible resonance. Thicker panels just can not be tuned.

(C) Rap your knuckles on the cabinet. If thin-wall will have a very dull low frequency sound. Refer to Harwood's Fig. 12 which clearly shows the unique properties of thin wall as a solution to cabinet resonance.

(D) Thin-wall is usually associated with removable front and/or back

(E) To avoid confusion best to ignore sales literature and check yourself

>

[A.S.]

The subject of 'thin-wall' cabinets is of great importance as it is the essence of the 'Harbeth sound'. I've made this little screen-video by explaining some key graphs in Dudley Harwood's 1977 BBC report (formerly head of BBC audio section, and Harbeth's founder). This video runs for 5 minutes, and will stream as it plays on the PC or MAC using the Adobe Flash format - if you need to update your free Flash player the appropriate link will open for you.

Link to Harbeth talk-through on 'thin-wall cabinets' (Alan Shaw talks you through BBC research)

If you find this interesting or informative please tell us so we can consider making more. There's 50+ years of pragmatic BBC audio research to illustrate and talk through! This was made in one take, unrehearsed and it's easy to make more of these if you wish.

P.S. I do hope that this is not too 'rough and ready'.

[Clark Williams]

Alan,

This was very informative. It helps owners and potential owners understand another reason why their Harbeths are so unique. Another suggestion for a talkthru would be the origins of RADIAL perhaps?

This method of education is great and adheres to the Harbeth ethos of lo-cost grass roots, but highly informative marketing. I do not know of another speaker manufacturer doing such; another Harbeth first perhaps? People like storytime. We like it here at Absolute Sounds and we refuse to begin work for the day until we get storytime from our fearless leader.

It's not too "ruff'n'ready"....you can't dig up Evelyn Waugh to narrate so...

Keep up the great work. Go make some Monitor 40.1s please.

Clark

[Don Leman]

Very interesting and informative Alan. Would have been nice to see some graphs of the Super HL5 to see the relationship to the LS 3/6.

Those wishing to read the whole research paper may do so by downloading the pdf file at...

http://downloads.bbc.co.uk/rd/pubs/reports/1977-03.pdf

Don
West Coast Audio
Surrey BC

[Gan CK]

The 29 page original BBC research department report from 1977 indeed makes for a great read. I continue to be fascinated by the BBC Thin Wall cabinet & totally agree that it cannot be bettered. By the way Alan, are the current Harbeth cabinets still using birch plywood or some other material?

[A.S.]

Yes, the report is very typical of Harwood's methodical and pragmatic approach. You will notice that the test equipment is humble bits-and-bobs and nothing remotely high-tech. No doubt they raided the department's petty cash tin for the necessary equipment funds! What excited me as a boy about the BBC's approach was its extreme pragmatism - one knew that if pocket money permitted it would be possible to emulate their results on the kitchen table. And I did - and still do, on a slightly glorified kitchen table! And you could too. And guess what - you'd reach exactly the same conclusions because real engineering based on science is timeless.

Thin-wall is one solution for one specific market and user. It is not really the best solution for for heavy-duty rock and roll music. But what is? Thin-wall is extremely expensive to implement because it requires a precision in cutting and assembly far greater than conventional thick-wall with glued-in front and backs and frankly, cabinet makers hate making thin-wall because it taxes their skills to the limit. And, the bitumen (or rubber equivalent) counterlayer, being oil based has rocketed up in price - and will continue to do so.

Thick-wall has two 'advantages' that some consumer will be attracted to - lower cost, and the absence of perimeter screws to hold in the front and/or back. In a conventional thick-walled cabinet, the front and back drop into a groove routed into the side/top/side/bottom; a couple of slaps of PVA glue and the cabinet is assembled and is guaranteed to be square even in the hands of the most junior worker. But a thin-wall cabinet, where the back and front are made at a different stage of the process, maybe days later, has no mechanism to pull itself square whilst the glue dries. The wet PVA in the 45 degree corner cuts lets the cabinet slip all over the place. It takes skill and enormous rubber bands, the whole length of the cabinet to gently tease the cabinet square. And lots of time to nudge and tap here and there. Misjudge the cabinet squareness and the front and/or back (which are cut at exactly 90 degrees) will not fit. Result: scrap cabinet.

No, we have not used birch ply for about 15 years except in the LS3/5a. Harbeth LS3/5a's were always birch ply but the other manufacturers changed to MDF as far as I know.

[s.a.b.]

No one reading Alan's posts, however, should fear that Harbeths are artificially "euphonic" or have rolled-off treble. Listen, for example, to well recorded brass- it has that wonderful natural "bite" lacking in most other speakers.

[Gan CK]

No one reading Alan's posts, however, should fear that Harbeths are artificially "euphonic" or have rolled-off treble. Listen, for example, to well recorded brass- it has that wonderful natural "bite" lacking in most other speakers.

Yes can't agree more with you on that. One such very good recording is the Harry James & his big band titled 'the king james version' album. Every time when i play this Sheffield Lab LP, i feel that i've been transported to the Wylie Chapel where the recording took place in 1976. The realism, dynamics & bite is really sensational. In fact, listening to this recording through my SHL-5s never fail to ignite goosebumps & even gets the hair on my arms standing to attention. Superb stuff!

[audisp]

No, we have not used birch ply for about 15 years except in the LS3/5a. Harbeth LS3/5a's were always birch ply but the other manufacturers changed to MDF as far as I know.

Hi Alan,

Just bumped into this post. I'm interested in the reason for the change from birch ply to MDF. What are the pluses/minuses of birch ply and MDF for loudspeaker cabinet?

[mhennessy]

Hi Alan,

Just bumped into this post. I'm interested in the reason for the change from birch ply to MDF. What are the pluses/minuses of birch ply and MDF for loudspeaker cabinet?

Me too. Although I note that Alan's quote doesn't actually say that he uses MDF - he's referring to other manufacturers of one specific model.

If not birch ply, what? I'd like to know because birch ply is very expensive and relatively hard to find in DIY-quantities - alternative suggestions welcome :-)

All the best,

Mark

[A.S.]

We have not used plywood for about (guessing) fifteen or twenty years. MDF is so much easier to work with. And the price is reasonable.

[mhennessy]

Thank you - I also like working with MDF, although I worry about the dust problems. Glad to hear that its use in "thinwall" construction is perfectly viable in your experience... I know that Rogers used it in their "thinwall" domestic models as well (e.g. LS7).

Interestingly, the last sheet of birch ply I bought didn't have the right number of veneers. It was 12mm and had 5 (IIRC) thick layers instead of the expected 9. No idea what this might have done to the acoustic properties in a speaker cabinet, but as my project involved exposed end-grain, I couldn't use it. Nor could I return it because a relative had brought it from a timber merchant on my behalf. A sign of the times, perhaps...

You mention the problems of assembling boxes with mitred joints. Assuming the box is veneered once it's been glued together, why not use a "proper" joint? I've always been taught to avoid mitred joints if possible - I only resort to them when using veneered stock.

[A.S.]

Yes, I'd expect 5 layer ply and 9 layer to have different acoustic properties. How different I couldn't say.

Actually, our cabinets are made by first laminating veneer onto MDF that is just wider than the final front to back depth of the cabinet then V-cutting the corners at exactly 90 degrees. Then the 'wrap' is folded up into the cabinet. It is nothing short of a miracle that a perfect box results from machining the long, flat panel as extreme skill is required.

Cabinet makers hate to work with thin-wall cabinets. They are the most difficult, frustrating and labour intensive box structure with hundreds of interdependent variables all waiting to catch-out the inattentive worker. Just finding a source who is willing to even attempt to make a perfect thin-wall box is a serious undertaking. The reject rate can easily be 30% when you come to close-up the wrap into a box if you are not extremely skilled. And to do it in quantity to our standard is a nightmare for all concerned.

By contrast, modern thick-wall cabinets - even very sexy looking ones - are a pleasure for a cabinet maker to produce with zero rejects. Attached a picture of a stack of C7ES3 'wraps' v-cut ready to be folded up into cabinets and then the bitumenn damping layer applied to tune the structure. You are looking at the inside - the white softwood bearers provide an anchor for the front and back to be screwed in.

As you can see, this 'BBC thin-wall' construction is a completely different animal to any other modern cabinet. Nobody with an eye to cost reduction (and a simple, stress free manufacturing life!) would chose this type of cabinet! This is the really difficult end of the cabinet making business and not one cabinet maker in ten would even attempt it.

>