Harbeth Monitor 30 Speaker Review

[tuga]

The Ultima Salon2 has a shelved-up "presence" plateau and I see no one complaining about it, even though it is probably worse from a perceptual/psychoacoustic perspective:

SoundStage! Measurements - Revel Ultima Salon2 Loudspeakers (12/2009)

 

[Thomas_A]

The Ultima Salon2 has a shelved-up "presence" plateau and I see no one complaining about it, even though it is probably worse from a perceptual/psychoacoustic perspective:

SoundStage! Measurements - Revel Ultima Salon2 Loudspeakers (12/2009)

I would say that the dip starts at 3,5-kHz-ish, i.e. not so different from the 3 kHz of the Harbeth.

 

[thewas]

I know all that but you are cheating because my comments were specifically about "on axis" response.

I am not cheating because listening window is almost identical to on axis response, so there is no loudspeaker which is flat on axis but has the BBC dip on the listening window, here for example for the here discussed Harbeth 30, the only ASR measured example of such a "BBC old school brand":

Above plot is generated from ASR measurements by https://pierreaubert.github.io/spinorama/

 

[tuga]

I am not cheating because listening window is almost identical to on axis response, so there is no loudspeaker which is flat on axis but has the BBC dip on the listening window, here for example for the here discussed Harbeth 30, the only ASR measured example of such a "BBC old school brand":

View attachment 188578
Above plot is generated from ASR measurements by https://pierreaubert.github.io/spinorama/

I depends on how wide the listening window is and how it's calculated.

 

[thewas]

I depends on how wide the listening window is and how it's calculated.

Surely, although for example the ASR NFS calculation of the LW is done based on the corresponding CEA standard. Still I haven't seen also in other sources which use different definitions (like for example Soundstage) a loudspeaker flat on-axis which suddenly has a BBC dip in the LW.

 

[mieswall]

I'm quite puzzled with Harbeth speakers, and of the way Stereophile measures them. When John Atkinson measured the Harbeth SHL5+ cabinet resonance he did it in the upper wall, that is the most rigid one, and he got a peak resonance of -3.6db@152 hz. Some years later he measured the Stirling LS3/6 unit, almost the exact equivalent to the SHL5+, but instead of the top panel he measured the side wall, achieving a monstrous +4.4db @160 Hz (louder than the very loudspeaker!). I'm pretty sure that if Atkinson would have measured a side panel instead of the top one in the Harbeth's, he would have gotten al least the same figures, if not higher.

I'm currently testing a pair of SHL5 in my home, and it is just shocking how those panels vibrate, with the side one vibrating MUCH more than the top one, an issue you can even sense just placing your hand on them. I even made the "cup of glass-and-ear" test, and is amazing how much sound you hear with it (with my Audio Physic Classic 20 I get a faint, almost inaudible sound with the same procedure; with my KEF LS50 Metas it is absolute silence). The thing is so weird, that in some specific locations in my room, even while directly facing the loudspeakers, the sound appears to be coming from a side window and nothing from the Harbeth: there are so much waves arriving from the loudspeaker's side walls bouncing into the window, that even that delayed sound is of a much higher amplitude than those coming from the loudspeaker itself. This is the first time I've heard such a bold anomaly.

I'm a bit troubled with this: I have many friends that just LOVE Harbeth loudspeakers. I personally find that their sound is...well... agreeable, but not faithful: besides the anomaly described, I found them "slow", without transient impact, that feeling that even brass instruments sound as they were made of wood, and notoriously rolled-off in highs (not because of direct sound that it is OK - I measured it-, but instead because of the contribution of the side panels I guess).

I discarded buying these second hand units. I have been skeptical of this old technology of the 60's-70's, but this detailed listening I've been doing for weeks just confirmed me this is not the product I want. No matter what Alan Shaw or magazine reviewers say.

 

[Chaconne]

I'm quite puzzled with Harbeth speakers, and of the way Stereophile measures them. When John Atkinson measured the Harbeth SHL5+ cabinet resonance he did it in the upper wall, that is the most rigid one, and he got a peak resonance of -3.6db@152 hz. Some years later he measured the Stirling LS3/6 unit, almost the exact equivalent to the SHL5+, but instead of the top panel he measured the side wall, achieving a monstrous +4.4db @160 Hz (louder than the very loudspeaker!). I'm pretty sure that if Atkinson would have measured a side panel instead of the top one in the Harbeth's, he would have gotten al least the same figures, if not higher.

I'm currently testing a pair of SHL5 in my home, and it is just shocking how those panels vibrate, with the side one vibrating MUCH more than the top one, an issue you can even sense just placing your hand on them. I even made the "cup of glass-and-ear" test, and is amazing how much sound you hear with it (with my Audio Physic Classic 20 I get a faint, almost inaudible sound with the same procedure; with my KEF LS50 Metas it is absolute silence). The thing is so weird, that in some specific locations in my room, even while directly facing the loudspeakers, the sound appears to be coming from a side window and nothing from the Harbeth: there are so much waves arriving from the loudspeaker's side walls bouncing into the window, that even that delayed sound is of a much higher amplitude than those coming from the loudspeaker itself. This is the first time I've heard such a bold anomaly.

I'm a bit troubled with this: I have many friends that just LOVE Harbeth loudspeakers. I personally find that their sound is...well... agreeable, but not faithful: besides the anomaly described, I found them "slow", without transient impact, that feeling that even brass instruments sound as they were made of wood, and notoriously rolled-off in highs (not because of direct sound that it is OK - I measured it-, but instead because of the contribution of the side panels I guess).

I discarded buying these second hand units. I have been skeptical of this old technology of the 60's-70's, but this detailed listening I've been doing for weeks just confirmed me this is not the product I want. No matter what Alan Shaw or magazine reviewers say.

On the other hand, to me Harbeth speakers simply sound more like the real thing than any other speakers I've heard. I find measurements, construction theories/methods, etc. quite interesting. But in the end, what matters to me is the sound of unamplified instruments in a concert hall or recital space. I don't know how they do it and ultimately don't care, but Harbeths (and some of their BBC-style cousins) simply sound more real to me than any others I'm aware of.

 

[Vuki]

I'm quite puzzled with Harbeth speakers, and of the way Stereophile measures them. When John Atkinson measured the Harbeth SHL5+ cabinet resonance he did it in the upper wall, that is the most rigid one, and he got a peak resonance of -3.6db@152 hz. Some years later he measured the Stirling LS3/6 unit, almost the exact equivalent to the SHL5+, but instead of the top panel he measured the side wall, achieving a monstrous +4.4db @160 Hz (louder than the very loudspeaker!). I'm pretty sure that if Atkinson would have measured a side panel instead of the top one in the Harbeth's, he would have gotten al least the same figures, if not higher.

I'm currently testing a pair of SHL5 in my home, and it is just shocking how those panels vibrate, with the side one vibrating MUCH more than the top one, an issue you can even sense just placing your hand on them. I even made the "cup of glass-and-ear" test, and is amazing how much sound you hear with it (with my Audio Physic Classic 20 I get a faint, almost inaudible sound with the same procedure; with my KEF LS50 Metas it is absolute silence). The thing is so weird, that in some specific locations in my room, even while directly facing the loudspeakers, the sound appears to be coming from a side window and nothing from the Harbeth: there are so much waves arriving from the loudspeaker's side walls bouncing into the window, that even that delayed sound is of a much higher amplitude than those coming from the loudspeaker itself. This is the first time I've heard such a bold anomaly.

I'm a bit troubled with this: I have many friends that just LOVE Harbeth loudspeakers. I personally find that their sound is...well... agreeable, but not faithful: besides the anomaly described, I found them "slow", without transient impact, that feeling that even brass instruments sound as they were made of wood, and notoriously rolled-off in highs (not because of direct sound that it is OK - I measured it-, but instead because of the contribution of the side panels I guess).

I discarded buying these second hand units. I have been skeptical of this old technology of the 60's-70's, but this detailed listening I've been doing for weeks just confirmed me this is not the product I want. No matter what Alan Shaw or magazine reviewers say.

All sides were measured, one side (representative) measurement published. Level is not in relation to loudspeaker output level. Eg. Revel f208 is at - 5dB, Revel f30 is at - 1dB

 

[DSJR]

The idea originally was to keep the midrange frequencies totally clear of cabinet resonance artefacts and, if you think the SHL5 (not a neutral balanced speaker as the Plus version seems to be) vibrates, don't even go to the ancestor Spendor BC1 or derived Rogers LS3/6 (which the BBC authorised but in the end hardly ever bought apparently, preferring the BC1's instead for general 'sound box' duties).

Unlike many of the 'Me Too' two cubic foot with Bextrene 8" that followed, I'd suggest the Harbeth 5 midrange continues where the BC1-LS3/6 of the 70's left off. The midrange can be very natural, but older models were 'restrained' up top and were slightly more 'organic' in the bass than strictly correct. In lighter builkt and furnished timber style room builds which I gather absorb bass, it was a good 'voicing' where a more 'impressive' domestic speaker would be too shrill.

The 5+ model signalled a slightly different and more lively-balance direction for the brand and the C7 was for me a revelation over the tubby mid-bass ES3 predecessor. 90% of the changes were in the crossover rather than the box, so I'd not bother too much about the perceived side wall vibrations myself, as the 'sound' is determined by far more important things here. I'd add that Stereophile measured the 5+ on the top tweeter axis rather than the main one for some reason, although they realised the error in the write-up, but the Stirling 3/6 was properly done on the main tweeter axis.

 

[Waxx]

I know from my experience with the M30.2 and the HL5+ that those speakers are fairly neutral (not fully, but enough to be good) and very easy to listen too. I heared them a lot (my ex has those in her house) and they are hard to beat on that easy to listen to.

They are also not picky on what amp you put them on, she switched between a push pull tube amp (Prima Luna Dialogue HP) and a solid state amp (Bryston 4B SST) at the time we were a couple and later on i build some Hypex Ncore 400 kits to power the M30.2, which were as good for the speaker (but more neutral in sound). The speaker doesn't care, as long as there is enough power in the right ohm...

She did sell all her other speakers but the speaker she grew up with (a 1970's Tannoy Arden, but in a rebuild cabinet that is braced like it should and in 18mm plywood), and those were way more expensive (B&W 800 series, Kiplishhorn and so), but did not sound as good in real life as those Harbeths. And that is what i hear from a lot of Harbeth owners, they are so easy en enjoyable to listen to that it's hard to listen to others, even if they are more neutral or more expensive.

And i have to say, if i had to buy hifi speakers, not for monitoring use, i would probally also end up with a pair of those. But i prefer to build them myself as building and designing is half the fun for me. But they are definitly not neutral studio monitors, there you better go looking at Genelec, Neumann, ATC or Kii for me.

 

[Willem]

I would ignore observations on older models and second hand units. Technology has moved on, and second hand speakers may no longer meet the original specifications. Anyway, recent versions have measured very flat. I own a quite recent pair of P3ESRs as desktop speakers, and I am more than happy with them.

 

[DSJR]

Alan admitted on the HUG a year or two ago that older generations of his designs were deliberately set with the tweeter lower than 'technically flat' on axis. Age has nothing to do with it @Willem, from HL5 and first Compact and P3 onwards, the hf was deliberately 'laid back' subjectively. The Harwood designed HL IV measured similarly according to HFC measurements in the 80's, but I don't know this one personally. His idea of 'good bass' has been adjusted too from the 5+ onwards, but simply lifting the upper mids and hf can give that impression too.

The thing is - and this has been mentioned in other threads here, many audiophile buyers of such speakers don't read audio mags, they don't look online although that may be changing and the 'instant YouTube Gurus with six months total experience' aren't the fodder of such people. Buying an active speaker isn't really known to them it seems (unless it's a B&O) and they still like to choose an amplifier for all the well-discussed reasons. The higher prices have brought a different kind of audiophile to their customer base, into product cachet, valves and also looks, at almost any price, which I don't remember this brand attracting ten or more years ago. There's a scurry to as gently as possible advise without putting posters off the brand.

Harbeth reportedly now know full well how good class D amps can be, even cheap ones which have been used successfully at the factory with Harbeths and of course they used to make small active 'pro' models under the HHB Circle brand name (purple cones and all!). It'll be interesting to see how the brand progresses in the future, as new models will have to be rather different to genuinely improve (the ex-Spendor designer now working there used simple waveguides in their later model tweeters, so we'll see).

 

[Waxx]

Alan admitted on the HUG a year or two ago that older generations of his designs were deliberately set with the tweeter lower than 'technically flat' on axis. Age has nothing to do with it @Willem, from HL5 and first Compact and P3 onwards, the hf was deliberately 'laid back' subjectively. The Harwood designed HL IV measured similarly according to HFC measurements in the 80's, but I don't know this one personally. His idea of 'good bass' has been adjusted too from the 5+ onwards, but simply lifting the upper mids and hf can give that impression too.

The thing is - and this has been mentioned in other threads here, many audiophile buyers of such speakers don't read audio mags, they don't look online although that may be changing and the 'instant YouTube Gurus with six months total experience' aren't the fodder of such people. Buying an active speaker isn't really known to them it seems (unless it's a B&O) and they still like to choose an amplifier for all the well-discussed reasons. The higher prices have brought a different kind of audiophile to their customer base, into product cachet, valves and also looks, at almost any price, which I don't remember this brand attracting ten or more years ago. There's a scurry to as gently as possible advise without putting posters off the brand.

Harbeth reportedly now know full well how good class D amps can be, even cheap ones which have been used successfully at the factory with Harbeths and of course they used to make small active 'pro' models under the HHB Circle brand name (purple cones and all!). It'll be interesting to see how the brand progresses in the future, as new models will have to be rather different to genuinely improve (the ex-Spendor designer now working there used simple waveguides in their later model tweeters, so we'll see).

Wel, to answer your question, Allan Saw said more than once on the Harbeth forum (free accessible after registering, even if you don't own a product) any amplifier that is good should work with his speakers, and he says a good integrated like a Yamaha A-S201 is as good or sometimes even better as those super expensive esotheric ones. He also suggest standard cabling and is not like some others promoting snake oil stuff.

And yes, the speakers are expensive. But building wooden speakers in Europe is expensive with all the needed hand labour, especially in the UK. And he claims (i don't know if that is true) that each pair is tuned seperatly on measurements in this

.

 

[DSJR]

My first pair of HL5's had crossover film caps which had been individually measured. Not sure if this is necessary these days as it's many years since I visited the factory.

Yes, Alan has done his level best to make his speakers as amp-agnostic as possible and a while ago boasted at the inline loss of the 40's three way crossover (no mention of frequency though), but many of the new owners are still on the audiophool side of audiophiledom IMO and need to be gently weaned off the negative aspects of some of the wilder 'enthusiast-hobby' source and amp products which they appear to often own. It takes time of course and a 'softly softly' approach is the Harbeth way it seems, but hopefully they'll keep on.

 

[ahofer]

And that is what i hear from a lot of Harbeth owners, they are so easy en enjoyable to listen to that it's hard to listen to others,

This is my feeling as a SHL5+ owner, although I also enjoy the JBL L830's in the system in my other house.

Shaw maintains that a "lack of coloration" in the mid and presence range is the secret to the low-fatigue aspect of the speaker, but he's cagey as to whether that is in FR or distortion measurements. I also find the stereo illusion a bit more spacious than other speakers I've been able to compare them to. I'm still uncertain as to what those subjective impressions might correspond to in measured performance.

 

[DSJR]

I, like everyone else outside the factory, have never seen a 'raw' Radial-Coned Harbeth driver measurement, but JA commented (I think in a now gone review of the SHL5 fifteen or so years ago) how well behaved the driver appeared to be, the crossover not needing to smother nasty resonances as happens with many metal cone drivers (I see the popular one with copper coloured phase plug and all I can see in my mind is a 10dB+ peak at several kHz or so that a crossover can never fully remove). The radial cone is certainly much 'stiffer' than the awful poly thing the LS5/9 uses, which is apparently very weak especially at the 'neck' of the cone and easily deformed all over.

The classic Bextrene BC1 and similar midrange quality BC2/BC3/SA2 large-coil version does have something good going on in the mids, even if the bottom and top of the range isn't as good perhaps as modern drivers can be. Got to say the 8" radial driver has to me anyway, a similarly good quality midrange 'sound.' Where Alan's talents really lie is at the crossover and he mentioned just recently what we old Linn dealers recognised years ago - fine tuning a tweeter to the top of the main driver is CRITICAL in all manner of ways not always shown in an on-axis response plot!

I'm now (having to) starting to look more closely at dispersion patterns in the crossover region (many old designer-hands here are decades ahead of me in this, but the well respected speakers here do show advances in lower kHz driver integration). In th eold Linn days for me, if the tweeter was only slightly out in terms of level, the speaker sounded (more than usually) 'odd' and the active crossovers with steeper slopes magnified this due to obvious 'steps' forming at the driver transition frequencies (I apologise for not putting it better). Getting the drivers as well level-matched as possible, the overall 'vibe' in the sound was much better - but this is but one adjustment a designer can make and it's here where Harbeth, presence dip and all (much reduced in current models), can really work for many, when compared to the boom-shriek-n-tinsel boxes passing as 'HiFi' these days. The current XD models may still not 'disperse' as well as a properly designed waveguide, but they're bloody good for a trad passive speaker range if you can afford them without wincing at the cost and I spent hours playing some of the new models at low levels and not having to turn 'em up or strain to hear what was going on.

 

[witwald]

The midrange can be very natural, but older models were 'restrained' up top and were slightly more 'organic' in the bass than strictly correct. In lighter built and furnished timber style room builds, which I gather absorb bass, it was a good 'voicing', where a more 'impressive' domestic speaker would be too shrill.

The "voicing" referred to above appears to be much like the use of the bass and treble tone controls on an audio amplifier. It seems to be counterproductive to "bake in" a tonal imbalance into a loudspeaker. As to other domestic loudspeakers that sounded shrill, could not that be simply explained by a very boosted tweeter treble response? That's a well-known design deficiency to which many loudspeakers, both inexpensive and expensive, appear to succumb to on a regular basis.

 

[tuga]

The "voicing" referred to above appears to be much like the use of the bass and treble tone controls on an audio amplifier. It seems to be counterproductive to "bake in" a tonal imbalance into a loudspeaker. As to other domestic loudspeakers that sounded shrill, could not that be simply explained by a very boosted tweeter treble response? That's a well-known design deficiency to which many loudspeakers, both inexpensive and expensive, appear to succumb to on a regular basis.

It could also just be marketing hype...
Reviewer enters the room at a show, sees BBC-like speakers, writes down "lifelike midrange" on his notebook and leaves.
He then enters the next room, sees a pair of Revels, writes down "accurate" on his notebook and leaves.

 

[DSJR]

The BBC aimed for a neutral 'accurate' balance at one time, but later on, the sound balance and sound editing engineers wanted a non flat response, that's what they got in the 5/8 and 5/9 I gather...

I gather far eastern rooms are very lightly constructed (hearsay but I've been told this a few times) and a warmer less taut bass can work well here (group delay rather than purely an eq thing I feel). Alan's skill has been to get the highs to 'sweetly' integrate. This kind of subtle 'voicing' done in the past is no different to the stark and brightly lit Dynaudio balance (the Confidence 60's costing a fortune send me from the room after half an hour... XD Harbeths don't sound overly 'warm' or 'nice' now apart possibly from the slightly 'full bodied' toned M30.2.

At least we have a choice and with my age/infection related hearing defects now measured and plotted, my future choices will be to look at wider hf dispersion with no 'dip' at the crossover point...

 

[MattHooper]

As I've mentioned before: I've heard some of the most natural sounding reproduction of the human voice from some Harbeth speakers.
Previous to owning a pair I remember being startled by the realism of voices from Harbeth speakers in a few cases. One included an audio show where there were inevitably tons of rooms playing quite well recorded vocals "see how realistic our speakers are?" I would close my eyes and compare the sound of the reproduced voice to people's voices talking in the room and it would always show how artificial - hardened, or thin, or electronic-sounding - the reproduced voice was compared with the actual flesh-and-blood voices of people in the room or nearby. The only speaker that came close to surviving this comparison were the Harbeth speakers (30s, I believe). Track after track the voices, with eyes closed, sounded more like the real thing even having real people talking to compare.

I seem to remember Shaw talking about the human voice being his main reference, and comparing the sound of his speakers with real voices.