Harbeth Monitor 30 Speaker Review

[dougi]

Doesn't agree very well with Amirm's measurement.... we know how Amirm does his measurement. Do we know how Australian HiFi does theirs?

View attachment 135645

If you go to an example similar test you will see that the plot above is an in-room pink noise average. Hence likely to be different that Amir's above.

 

[witwald]

Doesn't agree very well with Amirm's measurement...

The Harbeth M30.2 that was tested was the 40th Anniversary model, which was available circa. 2017. On the other hand, the Harbeth Monitor 30 that was tested by Amirm is the 25th Anniversary model, which came along some 15 years earlier. As a result, these two test articles are quite different products in the Harbeth development cycle.

...we know how Amirm does his measurement. Do we know how Australian HiFi does theirs?

The measurement of the M30.2 Anniversary that was shown was from Graph 1 of the review, and it is a composite plot constructed from two different types of measurements. The caption for that figure stated: "Frequency response. Trace below 900 Hz is the averaged result of nine individual frequency sweeps measured at three metres, with the central grid point on-axis with the tweeter using pink noise test stimulus with capture unsmoothed. This has been manually spliced (at 900 Hz) to the gated high-frequency response, an expanded view of which is shown in Graph 2."

 

[Thomas_A]

Doesn't agree very well with Amirm's measurement.... we know how Amirm does his measurement. Do we know how Australian HiFi does theirs?

View attachment 135645

The M30, M30.1 and M30.2 are not the same speakers so not relevant to compare.

 

[DSJR]

Below, two plots lifted a while ago. First the original SLS5/9 donor model -

Below, a lifted Graham LS5/9 plot from a HiFi World review - as I've said before, the M30 as Amir tested was designed as a refined replacement for the above LS5/9 model, the tweeter of which could sound a bit scrappy and slightly isolated off-axis. The Graham LS5/9 does seem a touch smoother like the M30 although the plot I have isn't very good - I think the bass to lower mid balance is still very much 'up' as the plot is so crushed vertically -

Below, another M30 plot.

Below, the 30.2 Anniversary from Stereophile I think (so the 100Hz lift may not be correct). The lower kHz region still shows a gentle recess

 

[kotmj]

I wonder when recordings are made of say classical music, or the sort Chesky Records makes, at the levels these things are normally made (90-100dB?) and which are not really "mixed", that, when reproduced through speakers at say 70dB it would not be advantageous to have the speakers respect the Fletcher Munson curves?

Are there areas of use-cases where it is a good idea that speakers have this baked-in EQ? Say, small nearfield bookshelf speakers mostly for use in smallish rooms by non-audiophiles (who do not EQ their sources)? I get the feeling that 99.9% of those who buy things like Wharfedale Diamonds do not play them at the same loudness at which music is either recorded or mixed, and I don't think they apply EQ.

 

[kotmj]

Sonus Faber Chameleon (tiny bookshelves) courtesy of the NRC

 

[kotmj]

Sonus Faber actually has an anechoic chamber at their facilities, so they are not ignorant of the importance of measurements.

 

[kotmj]

The opposite position to take is that Toole reigns supreme: all speakers should have a completely level freq resp. All speakers should be like Revels.

 

[kotmj]

Sonus Faber Lumina III, a large floorstander.

It appears SF aims for a flat freq resp for large speakers meant for large rooms and FM-style resp curves for small speakers.

 

[kotmj]

Even within Harman, the guy who designed their TOTL speakers don't necessarily agree with Toole. This is from an interview with Greg Timbers.
------
Anechoic or listening? That is, where do measurements and listening fall in design? (Value, importance, role…)
This is a loaded question. I am going to answer it as I believe is correct. All designs require and deserve full Anechoic and other types of measurements. The cost, size or intended market for the device should not change the measurements taken. It doesn't matter how expensive a loudspeaker is, it should always be the absolutely best effort within the constraints agreed upon by cost, performance, appearance, etc. In many respects, cheap speakers are actually more challenging to do correctly. Once you know what your parts are capable of, you can work to achieve the highest level of performance possible within the guidelines. I am a firm believer that there is no magic in loudspeaker design, but I also strongly believe that we do not have all of the answers and that the ear is still the tool of choice. I can always hear differences that I can't measure but it doesn't make them any less important. It just means that we need to figure out how to measure what we hear. I have no use for blind and double blind listening tests the way Harman implements them. Sound systems and their environments are very complicated. No speaker is even close to sounding "real" so personal opinion is always a major consideration. Most blind tests are based on a series of assumptions that enable the test to be easy or practical to implement. Unfortunately, these assumptions often invalidate or color the results because they cover up or accentuate aspects of the loudspeaker design.

 

[antennaguru]

Here is a link to what I already wrote about the IMF Electronics Compact Monitor 2 vintage speakers that I refurbished, and which sound very similar to the Harbeth 30 series, are the same exact extereior dimensions, but lack the midbass bump - because they use Low Q aperiodic resistive bass reflex loading versus a High Q port. They can be bought occasionally for a few hundred dollars and Jerry at Falcon Acoustics will sell you a matched set of crossover capacitors for around a hundred dollars. I have heard both and prefer the IMFs to the Harbeths:

https://www.audiosciencereview.com/...ith-good-measurements.9637/page-3#post-821840

 

[DSJR]

The Stereophile mid bass is part of their testing regime if you read their description and not the speakers measured. Any speaker measuring flat on Stereophile is actually bass light! I'd expect a flat speaker to have a 3dB or so bump around the 100 - 120Hz area.

My second system boxes are a pair of early 1970's IMF Compacts, this era having 6" Elac bass and mid drivers with 'crackle doped' finish and an EMI cone tweeter. A bit bass light, the clarity is very good even today and this has led me to think of a three way with higher tweeter crossover point rather than a two way with crossover (issues usually) right where my ears aren't good. I'll deal with that when the time comes! The Super Compact from the mid 70's had a proper crossover network, a KEF B200 bass driver, not sure who made the mid driver with foam surround and the then ubiquitous Audax tweeter. Sound on these was a bit mid-recessed to better phase match their floor standing models for quad or surround sound use. Falcon supplied replacement close matched caps for the mid driver (bass just has a coil and tweeter a 1uF cap and Wembley Louspeakers re-foamed the bass drivers which may have altered the driver performance a little.

I've a soft spot for many IMF's if not all of them. The CM 2 wasn't tested in any of the 'Choice books I have but the CM 3 was.

 

[antennaguru]

Yes, agreed on the graphs. However the IMF CM2 still sounds a bit better than the Harbeth 30 series in my opinion. The Harbeth 30 series is also waaaaaay overpriced for a simple 2-Way-8!

 

[milosz]

When I was in college, I met a guy who had a pair of the IMF floorstanding 4-ways, TLS50's I think? They did not have the KEF racetrack woofer, they had round woofers, so I think they were TLS50's. I thought they sounded wonderful. Every once in a while I see a pair of IMF Reference Monitors of various sub-versions on eBay and I am tempted. I hate the boxy sound of most speakers - I am a panel speaker guy, I have a pair of Quad ESL-57's and a pair of Magneplanar MG-3.6's, with Rythmik open-baffle servo subs using 2 DEQX's to handle tri-amping the MG 2.6's and splitting off the subs too. Boxes smear the low end- they absorb energy then release it over time, I hate that. Transmission lines are the only enclosure type that seems to me to preserve the time-domain information in the bass nearly as good as boxless designs.

 

[antennaguru]

When I was in college, I met a guy who had a pair of the IMF floorstanding 4-ways, TLS50's I think? They did not have the KEF racetrack woofer, they had round woofers, so I think they were TLS50's. I thought they sounded wonderful. Every once in a while I see a pair of IMF Reference Monitors of various sub-versions on eBay and I am tempted. I hate the boxy sound of most speakers - I am a panel speaker guy, I have a pair of Quad ESL-57's and a pair of Magneplanar MG-3.6's, with Rythmik open-baffle servo subs using 2 DEQX's to handle tri-amping the MG 2.6's and splitting off the subs too. Boxes smear the low end- they absorb energy then release it over time, I hate that. Transmission lines are the only enclosure type that seems to me to preserve the time-domain information in the bass nearly as good as boxless designs.

I agree and my main system has open baffle main speakers and very large sealed subwoofers (IRS towers). That's what I consider the best compromise as the open baffle speakers are super efficient, very dynamic, and disappear sonically while the subs plumb the depths with tremendous authority.

 

[DSJR]

Back in the mid 70's, we'd demonstrate and sell Crown amps exclusively with big IMF's. One or two here sneer at them but Ken Rockwell measured a few including the preamp (fingers dowen a blackboard sound was one audiophool comment when ARC sludge-box preamps became the sonic rage) and they still hold up very well half a century later and what comes out is basically what goes in! Four ohm and below drive is difficult for the smaller models though as protection kicks in.

I repeat, the M30's have changed out of all recognition from the BBC drop-in replacements of the model measured here. The Beeb have gone over to active Dynaudios mostly it seems for general 'noisebox' studio reproduction (I was told the odd Genelec is around). Now the M30 range is basically a domestic model, the response has flattened noticeably away from the LS5/9 donor model (which is still made by companies wanting to cash in on the legacy - the Rogers one looks awful in a HiFi News test and the Graham version is ripe and dull in balance if a HiFi Review squashed plot is to be believed). Below how the BBC wanted their two main monitors balanced - the M30 in original form smooths the highs a bit. I believe it's 5dB per major division. My Rogers 5/9's definitely sounded either mid-bass 'tubby' or 'laid back with sharp toned highs' depending on the room.

Below the Graham as measured in HiFi Review. Note the upper hundred Hz suckout.

The current M30.2XD isn't subjectively as bad as this although for myself now, I was staggered how much better over the previous model the slightly larger C7-XD model is, the mid bass under perfect control in the lengthy session I had with it.

I just hope people here can accept that any manufacturer should be allowed to learn from and further develop their products as the years and decades go on. Some makers regress sadly, but in my book, current model Harbeths seem excellent for the market they're in (which isn't necessarily 'ours'), the sound is perceptually their clearest yet (hell, I'm going to say all but on a par with many active models) and priced mainly to compete with now inferior equivalents from their 'BBC Legacy' competition, of which three well known makers immediately come to mind, either regurgitating old relics of the past, or trying to move on but with uncertain future now their final designer link with the past has moved on.

P.S. Rogers 5/9 measured in HFN - no idea how that terrible response equates to a similar sweep done on a Klippel - HF seems higher than ever in these...

https://www.hifinews.com/content/rogers-ls59-classic-se-loudspeaker-lab-report

 

[AC1]

Personally, I hated monitors with the BBC dip as mixing in the studio on those, one compensates for the dip in the mix. Then listening on speakers with no BBC dip, would sound like a "drill in your ears" with too much energy in the frequency range our ears are the most sensitive to (which I believe is the raison d'être for the so called BBC dip, but not good for mixing music on...).

Yes! Just try to listen to 'Jet' from the Anniversary Limited Edition of Paul McCartney & Wings' Band On The Run. The mids in that song are literally painful when listening to speakers that don't have that mid-range dip. While it might sound smooth and silky to many people, instruments simply don't sound complete and often overly polite (read: sans balls) when you take away a good portion of the mid-range.

 

[Pluto]

I'm amazed that this unsubstantiated, more or less legendary, response anomaly gets so many people worked up. If there is any substance at all to the "Gundry dip", it dates from a (largely mono) era when monitoring conditions in BBC cubicles (yes, cubicles) were often truly appalling. In those days a typical layout would have the panel SM sitting, often, no more than a metre from the glass separating the studio from the cubicle, with the loudspeaker at 90° to the head and frequently, little more than 1.5m away.

Under those conditions, it's hardly surprising that all kinds of anomalous tweaks got built in to things in an attempt to salvage something from the ruins.

 

[Willem]

Indeed it has absolutely nothing to do with recent Harbeth speakers.

 

[thewas]

Indeed it has absolutely nothing to do with recent Harbeth speakers.

All those "BBC old school designs" have still today some presence dip, if not on axis in the sound power due to the relatively large midwoofer crossed relatively high where they already beam before the tweeter without a waveguide takes over which is radiating wider.