Harbeth Monitor 30 Speaker Review

[Poseidons Voice]

Amir,

Thanks for the measurements. As expected, the directivity mismatch in this 2 way is most clearly exhibited by the mushroom cloud in the horizontal polar plot. Another way to look at it is to see the lateral off axis curves in the Stereophile review. They are awful. The speaker was designed pretty much with an eye on the direct axis (0 degree) to be as flat as possible but unfortunately that’s not good enough.

The early reflections DI and Soundpower DI support this as well.

Almost all 2 ways have a hole in the vertical off axis response and this one is no different. It’s not that big of a deal. Like Amir says, use absorption on the ceiling and most fellas can put a rug on the floor.

All in all, a rather expensive speaker given the sum of its parts.

And yes, I’ve listened to the top of the line Harbeth 40.2’s. My verdict? Pass.

Best,
Anand.

 

[MZKM]

Now if a company sells only direct then, in theory, they would charge $250 for the same $500 speakers without the dealer. Or since they sell fewer speakers in the direct model do they end up paying more for a lower volume of components to build those speakers and then also have to make more markup per speaker to cover their overhead taking it back up to the $500 price?

Then, for example, if Klipsch sells tons of speakers through places like Crutchfield, Best Buy etc. do they move enough volume that are they able to get the material prices down to a level where they can get the parts to make that $100 speaker for $60 and sell a $500 speaker for $300?

Compared to the big numbers of Klipsch and Polk speakers that get sold through major retailers can a little guy that is having wood cut and strapped together into flat-packs here in the US really shave enough off the cost that the big guys are paying to have their speakers completely assembled in China?

It just seems to me that the a chunk of the savings in the direct model is going to go away and a bigger chunk of savings in the flat-pack kit model are going to be eaten up with even larger expenses. There will likely still be savings but I'm not sure if they will be as great as the initial model I keep seeing where they claim the cut out the marketing budget and the dealer markup to get you speakers that would cost 2 to 4 times as much if bought in a store.

Don't get me wrong; I love DIY and direct speakers. I've built the Overnight Sensations and 5 of the Zaph audio ZA5.2 speakers and have been impressed with all of them. I just sometimes think the math is being made a bit extreme for the savings.

And when I bought my last brand name speakers I listened to every speaker that Magnolia HiFi had and only liked 2 pairs. Mine and the smaller version of mine. So, at least for me, the up to 4 times more expensive speakers didn't sound better (many sounded worse) so how much does price matter?

Kinda hit the nail on the head. Direct brands offer good value, but usually are not the conclusive best option in their price range; except for maybe Ascend Acoustics as their ~$2800 tower with RAAL and their new 2-EX have excellent measurements. I own bookshelves from HTD and RBH, and while they don’t make me have tears of joy from listening to them, I consider them very worthwhile for their price.

 

[BYRTT]

Why do you think the situation you described with difference in acoustic centres cannot be solved with phase EQ?

Take a look at my Harlech's, both woofers are playing up to 1800Hz but one is firing upwards, so acoustical centres of woofers and tweeter differ very much...

But I still managed to get pretty nice step response (measured at LP)...

Bit off topic here but will mean to get a textbook example for that response overlaid the other two sweeps for help learn how it should have looked, set X axis -1mS/+5mS, then go to tab "All SPL/Controls/Measurement action/Minimum phase version" and afterwards import that IR, it would probably look like a sharp vertical rise up to about 90% then a nice round over caused by ala B&K house curve out at LP, also Rephase can generate texbook education examples for REW, give it some tons of taps and output as 64 bits IEEE wav files plus its best for prescision always use same sample rate as your sweeps.

 

[Wombat]

FWIW

https://audyssey.zendesk.com/hc/en-us/articles/212347863-Midrange-Compensation

 

[veeceem]

Was it these speakers and Int-25 that gave the reviewer of Stereophile a new good experience :/

 

[Francis Vaughan]

You may have read some of Dr Toole's posts saying that placing a taped-on accelerometer on a random location on the speaker cabinet is not a reliable way to assess its performance. Above is the reason. You have no idea what the vibration modes are. You either need a full laser scan, or measure it with a mic array and compute for it. Not a simple subject.

The quoted stuff on supersonic waves etc seems to be describing the phenomenon of evanescent waves. Cabinet sides forced by the pressurised air in the cabinet aren't really subject to this. Where it gets really important is in things like speaker cone design. The speed of sound in a speaker cone, met with the shape of the cone, and interfacing to the air, means you have evanescent wave phenomena that do have a real impact.
As to cabinet walls, I agree a single accelerometer isn't enough, but jumping to a laser interferometer or the like is massive overkill. Want to see the modes? Turn it on its side, sprinkle some fine sand on it. Luthiers have been performing high quality analysis of vibrating plates for centuries without FEA or laser interferometry. Again, learning about how speaker drivers behave has taken these tools. But cabinets are flat slabs of wood. They are not rocket surgery.
http://www.sites.hps.cam.ac.uk/whipple/explore/acoustics/ernstchladni/

 

[anmpr1]

Most loudspeakers (with dealers) ...meaning if it costs them $100 to build (including stuff like shipping) ...results in 150% profit.

...I know PS Audio stated their upcoming reference DAC would use a ~5x markup (he stated rough material cost and rough retail cost).

A 150% net profit (i.e., after all costs) would surprise me.

I thought PS was selling direct now? The idea that an outfit like PS is going to sell 5 times over their total cost + retail share, direct to consumers, is absurd. On the other hand, maybe they have the fan base that will accept that kind of spread in order to own Paul and Stan's special kind of magic for their average quality gear.

 

[Willem]

The discussion of prices is quite a US specific story because Harbeth's US importer charges a major mark up compared to European prices. US price for the M30.1 seems to be $5500 for the cheapest veneer, whereas the cheapest veneer in the UK costs about 3000 pounds and a similar 3500 euro in the Netherlands. Note also that the US price does not include local sales tax, whereas the UK and Dutch prices include Value Added Tax of 20% and 21%. So US prices are almost twice as high. Admittedly distribution costs are a major part of the final price of any product, but to me this is a bit over the top, so I can understand the comments about the price.
Conversely, I don't mind paying for products made for decent European wages and in proper labour conditions.

 

[martijn86]

Tom Holman of the T in THX recommended that the same dip be implemented in Audessey Room EQ! A mistake in my opinion.

At least in Audessey you can turn it off (with the MultiEQ app). Although I recommend leaving everything above 400Hz or so untouched by Audessey. That covers room mode and port tuning for most speakers and rooms but preserves the imaging of your speakers better.

 

[anmpr1]

The discussion of prices is quite a US specific story... Admittedly distribution costs are a major part of the final price of any product... Conversely, I don't mind paying for products made for decent European wages and in proper labour conditions.

Why certain loudspeakers cost as much as they do is a big question. It can't be simply materials...a box and some drivers. I understand about labor costs, factory expenses, taxes etc. For a product actually designed and built in GB, I guess it makes some sense. Or a Klipsch Heritage loudspeaker hand made by craftsmen in Hope Arkansas. I get that.

What I don't understand is something like the JBL L100 classic, made in Indonesia. For that, and Harman's distribution network, these should be selling for $500.00 and not four thousand. I'm not knocking Indonesian workmanship. Far from it. You can buy a beautifully finished electric guitar made in Indonesia--five hundred dollars or less shipped anywhere in the world, purchased through a guitar store that takes their own cut. I imagine it's harder to build a guitar than a loudspeaker. But that's Indonesia labor and manufacturing.

 

[Willem]

To be fair, a Harbeth speaker is a bit more than just a box. The quality of the veneers and the woodwork is quite exceptional (whether you want or need that is something different, of course). And their Radial woofer is not some bought in Chinese mass product, but produced in-house in relatively small batches in a proprietary process.
I have often wondered if manufacturers like these would not be better off selling direct. Dealers (and in the US case also the importer) take a huge cut, and I am not so sure they always give the customer value for money. How many dealers demonstrate level matched, and how many refrain from pushing vinyl, tubes and other forms of wasting money?

 

[Juhazi]

About these box resonances, CSD measurement of on-axis (my guess) sound shows long decay of frequecies that match with Atkinson's tape measurements of 30.2. I believe that's more important than a small wiggle in spl measurement. By the way it was side and top panel resonating in Atkinson's test.

Anyway I wonder why time=0 shows no dB at all above 1kHz in this NFS measurement...

compared to Atkinson's CSD of axial sound, which doesn't show low freq at all - obviously different parameters for plotting!

Notice! Low freq decay can be room modes or other artefact, amir doesn't have an anechoic chamber. That is most likely the reason Atkinson doesn't want to show what happens below 1kHz. . An amateur should measure speaker's decay in extreme nearfield - an inch from membrane of each driver. Room measurements (listening spot) with longer decay with specific settings are under another tab in REW measurement program.

 

[Sal1950]

There is a lot of controversy around the so called "BBC dip" where around the crossover frequency the on-axis response drops fair bit. Tom Holman of the T in THX recommended that the same dip be implemented in Audessey Room EQ! A mistake in my opinion.

Thankfully now switchable with the new Audyssey Editor app.

 

[Labjr]

To be fair, a Harbeth speaker is a bit more than just a box. The quality of the veneers and the woodwork is quite exceptional (whether you want or need that is something different, of course). And their Radial woofer is not some bought in Chinese mass product, but produced in-house in relatively small batches in a proprietary process.
I have often wondered if manufacturers like these would not be better off selling direct. Dealers (and in the US case also the importer) take a huge cut, and I am not so sure they always give the customer value for money. How many dealers demonstrate level matched, and how many refrain from pushing vinyl, tubes and other forms of wasting money?

The retail furniture business has some of the highest markup and it seems like value than hi-fi. Two MDF boxes shouldn't cost several thousand. Even with a nice veneer. IMO B&M hi-fi dealers selling esoteric gear are a thing of the past. They don't add any value to the purchase. They're only pushing the the latest thing to be reviewed with the highest margins. The manufacturers, along with the review publications, and dealers are all in together, creating this illusion of value that doesn't exist. It's like politics. "Shut your mouth, do what I tell you, and we'll all make money!"

 

[Willem]

And for the current M30.2 Australian AV hub summarized their response measurements with the words 'superbly flat': https://www.avhub.com.au/product-re...2-anniversary-loudspeakers-review-test-496810

 

[BDWoody]

US price for the M30.1 seems to be $5500 for the cheapest veneer, whereas the cheapest veneer in the UK costs about 3000 pounds and a similar 3500 euro in the Netherlands. So US prices are almost twice as high.

So, 3000 pounds and 3500 euro is about $3,850. Closer to 50% more...but still a lot more for sure.

 

[Krunok]

About these box resonances, CSD measurement of on-axis (my guess) sound shows long decay of frequecies that match with Atkinson's tape measurements of 30.2. I believe that's more important than a small wiggle in spl measurement. By the way it was side and top panel resonating in Atkinson's test.

Anyway I wonder why time=0 shows no dB at all above 1kHz in this NFS measurement...

compared to Atkinson's CSD of axial sound, which doesn't show low freq at all - obviously different parameters for plotting!

Notice! Low freq decay can be room modes or other artefact, amir doesn't have an anechoic chamber. That is most likely the reason Atkinson doesn't want to show what happens below 1kHz. . An amateur should measeure speaker's decay in extreme nearfield - an inch from membrane of each driver. Room measurements (listening spot) with longer decay with specific settings are under another tab in REW measurement program.

Floyd Toole: "Putting an accelerometer on the wall of a cabinet is not a reliable indicator of the audibility of a resonance (sorry John Atkinson)."

https://www.avsforum.com/forum/89-s...viewed-speakers-ever-made-5.html#post54570242

 

[Soniclife]

So, 3000 pounds and 3500 euro is about $3,850. Closer to 50% more...but still a lot more for sure.

And it works just as badly or worse in the other direction for US kit. I find these price discussions largely pointless for this reason.

 

[maty]

Just now,

https://www.google.com/search?q=Harbeth+Monitor+30+test+review

 

[Juhazi]

Floyd Toole: "Putting an accelerometer on the wall of a cabinet is not a reliable indicator of the audibility of a resonance (sorry John Atkinson)."

Yes, that's why I showed CSD (cumulative spectral decay) measurements. What does FT tell about audibility of them?