Harbeth HL Compact 7ES-3 specific

[David Schalkwyk]

As usual, Alan has come up with not one, but two, wonderfully evocative ways of thinking about the nature of sound, reproduction and the place of Harbeth in the middle. The underwater cricket-ball is brilliant; and so are the two different popping packets, which give respectively high and low-Q sounds. I concur completely that Harbeths are low-Q, and marvellous for being so. But I'm troubled by a question:

Let's imagine that we make our two different pops: the satisfying low-Q one with the polythene plastic bag, and then the irritating loud, high-Q one with the crisp packet. We record both, and we play them through a pair of loudspeakers. How do we capture most accurately the DIFFERENCE between the sounds? With a low-Q speaker, or a high-Q one; or one somewhere in between?

This leads me to a further observation to add to my earlier comments about the differences between the P3s and C7s. I was always startled by the P3s' reproduction of percussive sounds, especially on jazz drums. They are immensely immediate and present, catching the leading transients of drums with disconcerting energy. This was so apparent that at times I'd feel there was a slight mismatch between the reproduction of drums, say, and piano or saxaphone. Compared to the P3s, the C7s don't have that startling, impressive, but slightly jarring reproduction of percussion, but they do reproduce the other instruments with greater ease, naturalness and richness. In short, they're better balanced. Do we see a degree of difference between the Q factors of the two speakers at work here? And is this due to the different Q factors of the RADIAL drivers?

Cheers

David

[A.S.]

...How do we capture most accurately the DIFFERENCE between the sounds? With a low-Q speaker, or a high-Q one; or one somewhere in between?

Ah, now that's the sort of question than makes me call out to my wife (now in bed, at the end of her long day) that I won't be long as I settle down to the computer to concoct an answer!

Ok, so we're clear that the bursting of the bags produces a difference in intensity of sound - and also of tone too. Now, before we can record those sounds we need to think through the process of recording to be sure we can draw valid conclusions on playback. First then, the choice of microphone. Most microphones are a little 'ringy' somewhere in the treble region - and that's deliberate and very much promoted by the mic sellers (and recording artists/producers) as 'sensational clarity' ... 'super clean' etc. etc.. In our terms, we'd say those mics were highish-Q, especially in the upper frequencies. If you want a low-Q mic you probably have to look backwards in time a bit to the very microphone designed by Dudley Harwood (our founder) when he was at the BBC. It's a ribbon, and unlike a modern capacitor microphone, it has just a strip of corrugated kitchen foil hanging limply in a very strong magnetic field. Yes, it is fragile and easily damaged, especially by wind. Modern mics have a plastic diaphragm pulled very tight and then welded under tension to the capsule case .... and, almost inevitable that tautness is their downfall - it causes peaks in the treble response. But they are very tough and durable. And produce a much higher output voltage than the ribbons. And they are less hissy too. You get an idea of microphone construction here (suggest p56). On page 63 you can see that a capacitor microphone is inherently peaky and entirely relies on some sort of internal damping to bring that peak down to an acceptable flatness. How well this actually works depends on such minute mechanical details of construction that only a few experts really understand how and why. (Such as B&K)

So, we'll take care over the selection of the mic. How about the room in which we record the bag popping? If the room is acoustically hard, like my kitchen or bathroom, then the reflections of the explosion off all the hard surfaces is going to greatly colour the sound picked up by the mic. So we really should go outside ..... but we will have to take great care to use our ribbon mic on a windless day. Finally, playback on the speakers. As we don't want the speakers to impart resonances (always associated with high-Q) we need our trusty low-Q Harbeths.

Incidentally - I was at the BBC anechoic chamber recently and took along and measured some 'modern' speakers not of our manufacture. I truly didn't appreciate how wonderfully low-Q and easy-on-the-ears our Harbeths are until I measured these others. The peaks in the response around the crossover region put these 'hot selling' 'must have' speakers in the super high-Q category and explain at a stroke why I cannot bear to listen to them or more than a minute or two. I hesitate to publish the frequency response curves because you probably wouldn't believe that any consumer (or professional) with normal hearing could buy them (and live with them) and the manufacturer would certainly recognise the tell-tale high-Q peaks in the response. Quite a shock I must say.

[David Schalkwyk]

Finally, playback on the speakers. As we don't want the speakers to impart resonances (always associated with high-Q) we need our trusty low-Q Harbeths.... The peaks in the response around the crossover region put these 'hot selling' 'must have' speakers in the super high-Q category and explain at a stroke why I cannot bear to listen to them or more than a minute or two. I hesitate to publish the frequency response curves because you probably wouldn't believe that any consumer (or professional) with normal hearing could buy them (and live with them) and the manufacturer would certainly recognise the tell-tale high-Q peaks in the response. Quite a shock I must say.

Thanks, Alan. You've established what I was angling at: whether preference for low or high Q was a matter of subjectively liking this rather than that, or accuracy of reproduction. So a low-Q 'speaker (and mic!) will reproduce a high-Q sound more accurately... That's interesting and counter-intuitive for the lay person. Surely, one would think, if I want to hear the brash resonances of a rock band, I need a speaker that is also brashly resonant. (I know there's a different thread on this.)

As to why people would want massively high-Q 'speakers, this intuition, propogated time and time again by the hi-fi press in their distinctions between a 'speaker's capacity to "rock" or not, is probably one reason. The other is more difficult to express, and may be conveyed by a story of what has happened to alcohol levels in wine over the past 30 years.

Thirty years ago, red wines (say, claret from Bordeaux) had alcohol levels of between 11 and 12.5%. With the influx of new world wines from hotter climates and a desire for more obvious fruit and softer tannins that make a wine drinkable soon after it is made, alcohol levels have increased to between 14.5 and 15.5%. That's close to port!

So the wine is now supposedly easier on the palate, it's softer and juicier. Only it's not. The added alcohol actually makes the wine more difficult to drink with food (the main aim of drinking wine), it makes it more difficult to drink more of it, and it finally deadens the palatte, not to speak of adding a whole unit per glass in the calculations of whether it is safe to drive home or not. And yet heavier alcoholic wines are what customers appear to want, despite their detrimental affects at every level except possibly a bit more rasperry flavour. Even clarets now contain 13-14% alcohol, and it's becomming impossible to get an Australian or South African Shiraz at less than 15%. I had a rose for lunch the other day that had 16% alcohol.

We seem to have a strange drive to convince ourselves that what we want or like is in fact what is horrible or bad for us. Go figure, as they say in the US. I've been wanting to establish a low-alcohol pressure group for wines. Perhaps they should give free membership to the low-Q club for 'speakers.

Cheers

David

[A.S.]

...We seem to have a strange drive to convince ourselves that what we want or like is in fact what is horrible or bad for us.

Before I comment I should just clarify one point .... when we say 'high-Q' or 'low-Q' this is a technical term relating to an electrical or mechanical systems behaviour at and around resonance frequencies. It should not be casually assumed that the subjective acoustic Sound Quality is necessarily bad with a high-Q speaker and necessarily good with a low-Q speaker. What we've been talking about is the intensity, the 'shrillness', the incisiveness of high-Q speakers. If you were tasked with designing a PA system for, say, the evacuation of an underground railway station, high-Q speaker with their punchy, attention-grabbing sound would be absolutely ideal and our trusty BBC-monitor low-Q sound so soft and relaxed that lives could be at risk. So, the speaker and the application have to be matched.

More general comment following your interesting notes on wine ..... (I didn't realise alcohol levels had increased, but then I am the sad, marginalised individual who likes Piesporter which you can rarely if ever find in restaurants or bars these days it is so far out of vogue...)

Two of the quirks of evolution that have brought us to the top of the food chain are curiosity and boredom. Curiosity about the environment (and people) around us and the boredom from repetition: we do like the buzz of things and people a-new. However, we are now being tantalised by a constant stream of tasty new goods, foods, clothes, music and sexy exciting people with one consequence: our nervous systems have withered through over-stimulation. Generation after generation we need more of a high just to get the juices going and we're hooked on this process. Just to think how one hundred years ago the Victorians could get a real high from a musical soiree around the piano with a few conservatively dressed friends!

[David Schalkwyk]

(I didn't realise alcohol levels had increased, but then I am the sad, marginalised individual who likes Piesporter which you can rarely if ever find in restaurants or bars these days it is so far out of vogue...)

One of the good things about German Riesling is that they've always had very low alcohol levels. In part that's because there's not enough sun to ripen them properly (i.e. produce lots of sugar in the berry), so they tend to be softer and sweeter, but with enough crisp acidity to balance the sugars. They sometimes have alcohol levels as low as 10.5%, and the really sweet ones can go down as low as 9%. These wines don't rock, but they're wonderfully rich, subtle, and give immense amounts of pleasure, just like Harbeths!

David

[David Schalkwyk]

It should not be casually assumed that the subjective acoustic Sound Quality is necessarily bad with a high-Q speaker and necessarily good with a low-Q speaker. What we've been talking about is the intensity, the 'shrillness', the incisiveness of high-Q speakers.

Getting back onto sound (or music) and away from wine, I can see the point that one would want different kinds of sound quality for different purposes. My original question was whether it would be better to have a low or high-Q speaker for the purpose of reproducing most accurately the difference between the explosion caused by crisp packet and the polythene bag. And Alan suggested that one would want low-Q speakers for this. That suggests that there is something objectively right about low-Q speakers for accurate reproduction (which is different from inducing people to evacuate a building). Or am I missing something here?

David

[Arlequen]

Hello Alan and hello Guys!

So .. anyone already listened the new series HL C7 series 3? And what is the difference from the previous (series 2) version? I missed you all just the time Alan presented the latest release of HL C7ES 3 and by the moment I haven't listened the new production

[Arlequen]

Hello Alan and hello guys

Now the new C7ES is well known .. could someone compare this new serie to the previosu one?
Where is better? ..

best regards

Curio

[A.S.]

I realise that your question was about sound but here is what I said earlier in this thread.

[Arlequen]

Ops! .. I got it Alan!
I didn't notice you already replied at this question
Btw I read around the net a very good review of the C7ESIII with a sonic parallel to the older II
I 'll try to sell mine and buy the new ES III or maybe I 'll think about the Super SHL5

Best regards ,

Curio

[hificafe]

Hello all,

last week I've bought Bewitch A30 (also known as Puresound A30) pure class A design with two valve rectifiers and a choke smoothed power supply. It uses Electro Harmonix 6550 output valves in an ultra-linear push pull design producing 30 watt / 8 ohm per channel (the output stage can also be switched to triode operation if preferred).

I wold like to ask you what is your experience with PP and Harbeth 7ES-3 (which I have) and/or 7ES-2?


Regs from Belgrade,
Sasa

[A.S.]

I thought I'd lost these video files from 2006 but recently as I've been archiving years of family videos onto hard disk* and I've rediscovered them on an old hard disk.

In the spring of 2006 I was developing what was to be the Compact 7ES3. All the critical listening of the the C7ES3, and crossover adjustment and in-room measurement was was undertaken in my 'study' i.e. a spare bedroom of 3m x 3m (I've just checked the dimensions) at home.

As I've mentioned before, it is much more challenging to design a speaker to sound natural and believable when the listener sits really close to it because nearfield listening seems to amplify dispersion, frequency response and colouration issues etc.. And, of course, in the nearfield, the environment local to the speaker has a great influence on sound. So, if a speaker can be made to sound great in the nearfield it's my experience that it will sound glorious when listened further away. But, I have so often observed, conventional speakers often sound acceptable in the farfield but extremely hard in the nearfield, which is not good for normal home listening in smaller rooms. I suspect that most speaker designers have the luxury of a huge well damped listening room at their facility and design accordingly in the farfield. But as our customers (and professional users) are likely to use the speakers in quite small rooms in the nearfield, I think our approach is more appropriate to the real-world user.

You will notice that I made no effort to clear the room of clutter. The left wall is a bank of hard reflective equipment and CDs. Yes, this did disturb the image a little (there is a door just to the right of the right side speaker which I kept open to disturb and counterbalance the right side image too). I'm sure you'll agree that this room is an acoustic nightmare with sound splattering off all those surfaces! And the speakers are very close to the rear wall and a heating radiator. But the unwelcome acoustic magnified every issue, which one by one over the months I solved so it actually helped. I have attached a picture of the set-up and I've highlighted points of interest. As you will always see in pictures of my work environment, my audio analysis equipment is always at hand - it is so important not to trust your ears but to use the equipment to keep your 'ears' on the ground - and to curb ones emotional attachment to the new creation. It is vital not to be seduced by the product you are designing and to maintain an objective detachment - measuring equipment guarantees that!

It would seem that I have several hours of 'spy cam' footage with sound, illustrating step by step how I listened, adjusted, measured and honed the design. They represent just one of many long and typical days during the C7ES3 design process going around the same measure, listen, tweak loop endlessly. Once we can transfer the video to a suitable web-based format and edit it down, we will make it available on line. I hope that it will help de-mystify the design process and show how we work in the real world, and hence are a little dismissive of money wasted in needless system tweaks. Speaker design is not complex but it is tiring trying to balance what your eyes (computer screen) and ears tell you is 'right'.

* After making various comparisons of hard disk performance, I have concluded that SATA is significantly faster than both USB2 and (surprisingly) Firewire400 and with a low processor overhead.

>

[digiphobe]

I recently posted a short review (favourable) in AudioAsylum in regards to the Compact 7's. There are those that actually question the retail price of the C7; imagine that. Although admittadly costly, I find that they're one of the few audio products that actually keep me listening, as opposed to boring, or offending.

[Gan CK]

I recently posted a short review (favourable) in AudioAsylum in regards to the Compact 7's. There are those that actually question the retail price of the C7; imagine that. Although admittadly costly, I find that they're one of the few audio products that actually keep me listening, as apposed to boring, or offending.

Sad to say, many so called audiophiles don't really listen to music. They are only interested in listening to exaggerated 'bang' & artificial 'wow' elements in music. Things such as correct timbre, tonality, emotion & artistry of music are deemed as unimportant. Just as somebody from this Usergroup has highlighted before, many high end loudspeaker manufacturers don't place any emphasis on correct timbre & tonality as their design criteria. But, if timbre or tonality is not correct, how can music sound like they are supposed to in the first place.

As far as i know, Harbeth is the only loudspeaker manufacturer in the world to use the human voice as their reference in voicing their loudspeakers & there is absolute logical sense to do that. After listening to & having owned countless speakers, i totally agree with Harbeth's claim that they make the world's most natural sounding loudspeakers. Probably i'll add in to say that Harbeth also makes the most correct & musical sounding loudspeakers. Period! Value? What value are we talking about when music sounds this natural & correct with virtually no peers from any other loudspeakers at any price? To me, its priceless.

I know that's a very strong statement to make but that's really how i feel.

[A.S.]

The issue of 'best value for money' is a complex one and can not be measured scientifically. When I was starting out here at Harbeth I was hugely influenced by the book 'Zen and the art of motorcycle maintenance' which explores the philosophical meaning of the word 'quality' and our concept of 'value'. In short, it said that products (and even people) sometimes had this inherent 'quality' which is how many observers would rate the product but that it was in fact impossible to pin down the precise definition of this 'quality' even though it was the widely held view. I guess that you either get it or you don't. I am not - as I've said many times - an audiophile nor do I believe that we are of much interest to audiophiles. Music lovers, yes, very much so.

From my side a product that demonstrates the mysterious 'quality' and hence offers great value is where self-evidently it has been designed and made in a way that is not usual or normal for that class of product, which may be expensive or difficult to make, has unusually tight control over tolerances and materials and which perhaps above all, is built to last.

As an example of an unusual construction technique that features in Harbeth speakers (and adds tremendously to the difficulty of making the cabinets and their cost) have a look at the attached front-and-back shots of the M40/M40.1 cabinet. The front and back are made as completely separate pieces, not glued in place but each held with 14 screw with just the right tension to control potential resonance. Who else would use such an expensive assembly technique these days unless they really cared about quality? It certainly pushes up the retail price.

From what we hear so often, when a music lover has gravitated to Harbeth that's the end of the process of fiddling around with speakers. It's a done deal as they say. That, to me, is the real value for the consumer and makes his Harbeths long term, the cheapest speakers he'll ever buy.

[Gan CK]

From what we hear so often, when a music lover has gravitated to Harbeth that's the end of the process of fiddling around with speakers. It's a done deal as they say. That, to me, is the real value for the consumer and makes his Harbeths long term, the cheapest speakers he'll ever buy.

Absolutely no doubt about that Alan!

[digiphobe]

Yes Gan, timbre and tonality are big ones for me also. Most of the previous loudspeakers that I've owned tended to sound either "lean" or "threadbare", lacking in richness and warmth, but not so with the Compact 7s' which seem to reproduce the body of the music much better.

[digiphobe]

Thanks Alan. One other aspect of value for me, is that I no longer have to engage the willing suspension of disbelief switch. It's easier to accept that what I'm hearing is a reasonable facsimile of the artist's initial intent.
"The cheapest speakers you'll ever buy" - that could be the new sales slogan.

[Shutterbox]

..the M40/M40.1 cabinet. The front and back are made as completely separate pieces, not glued in place but each held with 14 screw with just the right tension to control potential resonance....

Alan,

So the torque applied to the screws does affect the resonance, and thus the sound of a cabinet, and eventually the sound of a set of speakers? Is there a tight QC on this process so that all cabinet, belonging to a same model, will have a tight tolerance as far as screw torque is concerned?

If this QC does exist, is it also applied to other models too?

[A.S.]

There is no need to be concerned about finding a specific torque setting. "Fairly tight" is perfectly good enough and probably preferred to "exceedingly tight". In fact, it is better if each screw is at a very slightly different torque (as they will be, ex-Harbeth due to minute differences in the density of the wooden bearers) as this will randomise any resonance control benefits. That's a good thing.

The point is not to worry about this but to draw the comparison with what other manufacturers offer - namely rigidly pinned and/or glued baffles and/or backs. Before you audition a speaker it pays to have a close look at the construction method, specifically is the back and/or baffle rigidly glued in and approximately how thick are the side walls likely to be. If the box is rigid - the vast majority are - then that implies a certain type of system sound which to my taste can often sound rather cold.

Never forget that just as with a stringed instrument (violin, cello even piano) the speaker box is a critical part of the perceived sound.