• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Harbeth Monitor 30 Speaker Review

kotmj

Member
Joined
Jun 14, 2021
Messages
38
Likes
15
666SB35fig3.jpg

Above, the most studied and measured loudspeaker of all times: the LS3/5a. It's just like a curve in Fletcher Munson.

It's really a dip and a bass boost.
 

q3cpma

Major Contributor
Joined
May 22, 2019
Messages
3,060
Likes
4,416
Location
France
1141px-Lindos1.svg.png


Look at the curve for 80dB. Doesn't it look like the frequency response of the Harbeth M30 reviewed in this thread?
The mostly agreed reference level is 85 dB SPL, a speaker should be flat at that SPL, then. Seriously, don't drown in the kool-aid, loudness compensation is something that should be done upstream from the speaker, not as a builtin equalization, as this makes no sense; cf RME ADI-2 mode or https://github.com/dpapavas/alsaloudness.

The BBC dip is really just the dip in the Fletcher Munson curve. You see it in all BBC-inspired designs and even in something like the Yamaha NS1000M, whose frequency response can be found here:
http://www.troelsgravesen.dk/Yamaha-NS1000.htm
The BBC dip must be off-axis and was to avoid their tiny speakers from sounding shrill in the specific environment that is the broadcasting van. Yamaha's "BBC dip" isn't actually one, it's tuning for a flat power response, Toole talks about it a bit in his book.
 
Last edited:

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
I did not read through all 23 pages of this thread. But I have been reading a lot of what Alan Shaw writes on his forum.

The response curve of Harbeths, like most BBC designs, respect the Fletcher-Munson curve. The aim is to have a "full, rich" sound at low to moderate volumes. Basically, a little like the "loudness" button on older integrated amplifiers.

A perfectly flat frequency response curve would sound, to a human being, lacking in bass (and therefore warmth) when played at low to moderate volumes. A thin anemic sound.

A speaker designed respecting the equal loudness curve would sound much better at low to moderate volumes.

For myself, I would never buy a speaker without a frequency response like Harbeths since I measure with my phone typically 60 dB in my listening position.
The correct solution is a proper loudness control - not a switch but something like a volume pot. AFAIR Yamaha amplifiers had those in the 80ies, and Marantz AVRs offer dynamic volume with adjustable threshold. The setting depends on both recording and playback level and therefore it makes no sense at all to build one fixed curve into a loudspeaker.:facepalm:
 

DSJR

Major Contributor
Joined
Jan 27, 2020
Messages
3,314
Likes
4,427
Location
Suffolk Coastal, UK
To Willem - Not sure the Klippel references went down well though and I post with link I made wasn't published ;)

By Alans admission, Harbeths of old were definitely warm and 'safe' toned (the M30 original as measured here was a special case in this version as explained much earlier in this thread - the 30.1 tilted the response almost to flat (approx. > +2dB at 10kHz over th eversion tested here I recall).. Starting with the P3ESR and continued with the SHL5+ introduced nearly ten years back, the responses were rather flatter and subjective bass 'tautness/damping' was improved (not sure a pure response plot would show this as I can't find responses on the SHL5 compared to the 5+ for example).

The M30 is twenty years or so old and designed initially as a refined replacement for the rather odd BBC LS5/9 as I said earlier (I compared these to my own 5/9's). The 30.1 was intended for a wider market and the different bass tuning performance over the slightly taller Compact 7ES3 won it many friends, thus justifying continuing the model. The current 'XD' range moves things forward again a little and it's likely according to Alan that not much more can be done with these designs as he caims to have taken them about as far as he can. With Spendor's Terry Miles on board, a different skill-set is now available without rocking the boat too much I suspect.
 

LightninBoy

Addicted to Fun and Learning
Joined
Jan 9, 2019
Messages
712
Likes
1,458
Location
St. Paul, MN
For myself, I would never buy a speaker without a frequency response like Harbeths since I measure with my phone typically 60 dB in my listening position.

Audiophile: I want my system to replicate a live performance
Also Audiophile: I listen at 60dB.

Seriously though, I think your phone is not accurate. 60dB is really quiet - for example an acoustic guitar 10ft away is louder than that.
 

kotmj

Member
Joined
Jun 14, 2021
Messages
38
Likes
15
The correct solution is a proper loudness control - not a switch but something like a volume pot. AFAIR Yamaha amplifiers had those in the 80ies, and Marantz AVRs offer dynamic volume with adjustable threshold. The setting depends on both recording and playback level and therefore it makes no sense at all to build one fixed curve into a loudspeaker.:facepalm:
Your reasoning I can absolutely follow, unlike some others above. I suppose it depends on your level of engagement with the gear. The majority of users want a plug and play solution. For instance, many still do not use room dsp.
 

kotmj

Member
Joined
Jun 14, 2021
Messages
38
Likes
15
Audiophile: I want my system to replicate a live performance
Also Audiophile: I listen at 60dB.

Seriously though, I think your phone is not accurate. 60dB is really quiet - for example an acoustic guitar 10ft away is louder than that.
I have neighbours both sides of my house, separated by a single layer brick wall. I am always shocked at how loud others play their music when I visit them.
 

LightninBoy

Addicted to Fun and Learning
Joined
Jan 9, 2019
Messages
712
Likes
1,458
Location
St. Paul, MN
The correct solution is a proper loudness control - not a switch but something like a volume pot. AFAIR Yamaha amplifiers had those in the 80ies

Yes, I had one from the 90s. The loudness knob worked brilliantly.
 

kotmj

Member
Joined
Jun 14, 2021
Messages
38
Likes
15
To Willem - Not sure the Klippel references went down well though and I post with link I made wasn't published ;)

By Alans admission, Harbeths of old were definitely warm and 'safe' toned (the M30 original as measured here was a special case in this version as explained much earlier in this thread - the 30.1 tilted the response almost to flat (approx. > +2dB at 10kHz over th eversion tested here I recall).. Starting with the P3ESR and continued with the SHL5+ introduced nearly ten years back, the responses were rather flatter and subjective bass 'tautness/damping' was improved (not sure a pure response plot would show this as I can't find responses on the SHL5 compared to the 5+ for example).

The M30 is twenty years or so old and designed initially as a refined replacement for the rather odd BBC LS5/9 as I said earlier (I compared these to my own 5/9's). The 30.1 was intended for a wider market and the different bass tuning performance over the slightly taller Compact 7ES3 won it many friends, thus justifying continuing the model. The current 'XD' range moves things forward again a little and it's likely according to Alan that not much more can be done with these designs as he caims to have taken them about as far as he can. With Spendor's Terry Miles on board, a different skill-set is now available without rocking the boat too much I suspect.
He needs new woofers. That would open up a new era for Harbeth. He has ridden the Radial horse till it...
 

DSJR

Major Contributor
Joined
Jan 27, 2020
Messages
3,314
Likes
4,427
Location
Suffolk Coastal, UK
The BBC dip is really just the dip in the Fletcher Munson curve. You see it in all BBC-inspired designs and even in something like the Yamaha NS1000M, whose frequency response can be found here:
http://www.troelsgravesen.dk/Yamaha-NS1000.htm

HiFi Choice tested many samples of NS1000 and the originals had a smoothly rising response to 1kHz or so and then a very gentle drop into the tweeter. It would appear that as th edecades went on, the response flattened (controls at 'flat') a touch and in any case, backing off the mid control a bit and the tweeter a little less usually gave a fairly flat response from mid bass up
666SB35fig3.jpg

Above, the most studied and measured loudspeaker of all times: the LS3/5a. It's just like a curve in Fletcher Munson.

It's really a dip and a bass boost.

Real LS3/5A's with KEF B110 and T27 is more a 'Batman' curve than that shown above although the green plot looks nearer to me (note the Y axis which is slightly flattering). The 120Hz bump followed by a mid recess always there, but the approx. 1.5kHz peak has varied depending on samples but the current Falcon re-imagination was awful as measured in Stereophile. The B110 naturally dives after this right to the 5kHz resonance it suffers (you can see this in many original plots taken). The mesh over the T27 tweeter aids the post 10kHz lift and this was to aid hiss reproduction (as told me by one of the design team and a subsequent licencee of the design).
 

DSJR

Major Contributor
Joined
Jan 27, 2020
Messages
3,314
Likes
4,427
Location
Suffolk Coastal, UK
He needs new woofers. That would open up a new era for Harbeth. He has ridden the Radial horse till it...

The Radial cone gives very natural mids and isn't as 'flexible' as poly cones are. Have you heard the XD models - very different perceived balance and same cone... I'd say he needs a tweeter re-think myself ;)
 

kotmj

Member
Joined
Jun 14, 2021
Messages
38
Likes
15
The Radial cone gives very natural mids and isn't as 'flexible' as poly cones are. Have you heard the XD models - very different perceived balance and same cone... I'd say he needs a tweeter re-think myself ;)
I thought the intention behind the XD and Anniversary models were to make lots of cash for Shaw...

Shaw has always thought that the tweeter's contribution to the sound is minor. It projects much, much less power than the woofer. His words, not mine. Personally, I think Shaw has compromised high frequency hearing. Many musicians suffer from hearing damage as an occupational hazard; no reason a middle aged speaker designer wouldn't suffer the same.

He doesn't get help with voicing his speakers either. He doesn't ask for anybody's opinion when designing his speakers.
 

DSJR

Major Contributor
Joined
Jan 27, 2020
Messages
3,314
Likes
4,427
Location
Suffolk Coastal, UK
Not for me to be a mouthpiece, but have you seen the prices of Harbeth's spiritual competition? Rogers (re-launched and getting in on their legacy), Graham and Spendor since the re-financing a few years back in recent times were substantially more expensive and arguably not as good (subjective opinion). Only Stirling was a lot cheaper with their LS3/6 but these have climbed in price in more recent times. The 40th Anniversary showed there was a market happy to pay a premium for visually (and technically) tweaked examples and the XD range continued this (I was stunned at the C7-XD which used to sound too lush, but to me not now). Best I say no more, but they're still not silly priced - you're welcome to disagree :)
 

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,446
Location
NYC
1141px-Lindos1.svg.png


Look at the curve for 80dB. Doesn't it look like the frequency response of the Harbeth M30 reviewed in this thread?

Hi @kotmj , welcome to the forum =]

A speaker's frequency response should not follow the shape equal loudness contours, even if it is intended for low listening levels. The equal loudness contours are baked into our hearing and are handled in music's production. You may not have noticed the scaling of that graph.For context, here's what that 60dB curve looks like compared to the harbeth monitor 30smeasurements.

harbeth equal loudness.png

There isn'te quite a match =]

What does matter a bit is the difference between the curves at various listening levels. So there is definitely logic to having a speaker that has more bass than flat if you know your audience tends to listen at quieter SPLs.

I should further add that image from wikipedia appears to be slightly different from other graphs of the standard. If I trace the lines there compared to the one in Dr Toole's book, they are slightly different. In addition, there are some critiques to this standard too when it comes to small differences, and that there are some newer attempts at equal loudness with slightly different curves.

The only really consistent thing is that you want a bit more bass at low volumes.
 

mSpot

Senior Member
Joined
Nov 15, 2018
Messages
404
Likes
519
He needs new woofers. That would open up a new era for Harbeth.
On Harbeth's forum Shaw has said that with the XD release, he feels has reached the limit of what he could accomplish with their current driver design. Harbeth does have new drivers in development.
 

milosz

Addicted to Fun and Learning
Forum Donor
Joined
Mar 27, 2019
Messages
584
Likes
1,643
Location
Chicago
Lots of people love these Harbeth speakers. I have yet to hear them.

The only BBC speaker that I heard was an original pair of LS3/5a's. I heard them in a small room- nearfield- with decent solid state amplification and a CD. They sounded OK, but I do not understand what all the fuss was about. Imaging was good and they sounded smooth, but LS3/5a's have such a cult status and sell for such high prices - I don't really get it.

The "Harbeth sound" seems to be an elaboration on the BBC sound.
 

witwald

Senior Member
Forum Donor
Joined
Jul 23, 2019
Messages
478
Likes
502
The response curve of Harbeths, like most BBC designs, respect the Fletcher-Munson curve. The aim is to have a "full, rich" sound at low to moderate volumes. Basically, a little like the "loudness" button on older integrated amplifiers.
I understand that Harbeth loudspeakers are designed to have flat on-axis frequency response. Here is a test measurement of one such Harbeth loudspeaker, the M30.2 Anniversary Edition, taken from an Australian HiFi review (here). The response is quite flat.
NFzayx3SRQRimJh8d2W3k6-970-80.jpg.webp
 

jhaider

Major Contributor
Forum Donor
Joined
Jun 5, 2016
Messages
2,823
Likes
4,522
The response curve of Harbeths, like most BBC designs, respect the Fletcher-Munson curve. The aim is to have a "full, rich" sound at low to moderate volumes. Basically, a little like the "loudness" button on older integrated amplifiers.

Except…that’s not really in evidence in the measurements. The dominant feature in the measurements, IMO, is the large dispersion disruption caused by geometry errors: largish woofer and a flat waveguide for the tweeter.
 

milosz

Addicted to Fun and Learning
Forum Donor
Joined
Mar 27, 2019
Messages
584
Likes
1,643
Location
Chicago
I understand that Harbeth loudspeakers are designed to have flat on-axis frequency response. Here is a test measurement of one such Harbeth loudspeaker, the M30.2 Anniversary Edition, taken from an Australian HiFi review (here). The response is quite flat.
NFzayx3SRQRimJh8d2W3k6-970-80.jpg.webp

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

Harbeth Monitor 30 Speaker CEA 234 Spinorama Dashboard Audio Measurements.png
 
Top Bottom