[caryking]

I must say, looking at the crossover on the Harbeth speakers is completely amazing.  These are extremely expensive speakers and are using some inferior crossover parts.

What does that say about the audiophile community?

[Rocket]

Hi,

There is great value in the direct speaker market. Especially if your handy and can do woodwork.  I've heard that the cost of the speaker is actually 10% of parts etc and the rest is manufacturing, dealer and advertising and associated costs.

There is a place for bricks and mortar stores I just prefer direct speakers sales as its the best bang for buck.

Cheers Rod

[Early B.]

Low-level crossover parts quality is typical for "audiophile speakers." Ninety-nine percent of consumers don't know or care about crossover parts, so it's not a priority and the gains are usually not significant enough to warrant the added cost. Also, quite often, higher-quality caps and inductors are too large to fit inside the speaker cabinet.

Aside from the lower cost, there is a practical reason for using cheap parts -- availability. Boutique parts are more likely to become scarce or no longer produced.

 

[Hobbsmeerkat]

Here's the picture i took that Danny used in the video:



Its not the worst crossover we've seen, but its far from great.
There's a fair bit of performance left on the table, but there's just no room left within the speaker for a custom crossover.

[ric]

I think it's unfortunate that most speaker manufacturers do this. I ended up replacing a few caps and a resistor on my Spatial's--huge upgrade(Jupiter caps), but also sizeable cost, but to me completely worth it! It would be great to have optional upgrades for speakers using quality parts, but I realize it makes it more complicated and would add to the cost, especially if they made modular type crossovers that could plug in and be interchangeable. For now, I guess we have Danny's upgrades, which certainly takes the guess work out of it, but I think if most people heard the difference in sound quality a few parts make, they would want to upgrade! Another plus for an outboard crossover it that it takes it out of the box where vibration can affect sound quality. A few IsoAcoustic oreas also help!

[mlundy57]

I think it's unfortunate that most speaker manufacturers do this. I ended up replacing a few caps and a resistor on my Spatial's--huge upgrade(Jupiter caps), but also sizeable cost, but to me completely worth it! It would be great to have optional upgrades for speakers using quality parts, but I realize it makes it more complicated and would add to the cost, especially if they made modular type crossovers that could plug in and be interchangeable. For now, I guess we have Danny's upgrades, which certainly takes the guess work out of it, but I think if most people heard the difference in sound quality a few parts make, they would want to upgrade! Another plus for an outboard crossover it that it takes it out of the box where vibration can affect sound quality. A few IsoAcoustic oreas also help!

From a practical manufacturing point of view, having multiple versions of the crossover would require an major increase in the number of speakers kept in inventory for immediate shipping. Which then requires more warehouse space for the additional inventory. Complexity increases pretty quickly as thee number of options increases. Which gets back to the big upside of building your own speaker from a company like GR. But even then, Danny has to try and keep all the parts options in stock.

[WGH]

There is a lot more going on behind the scenes than people realize. You look at the crossover and decide you could do better... maybe, but you might only succeed in making the speaker different.

Harbeth always gets glowing reviews so they must be doing something special with those "inferior" parts. I would postulate it is in synergy and parts matching, which takes time and a huge inventory. Like making a fine wine, you have to know how to to mix varieties to make a consistent product year after year. Most caps are made with a 5%-10% tolerance, sometimes you can get 1% matched at a premium but you don't know the true value of the caps you are replacing. Harbeth measures each assembled crossover to a match a reference, something you can't do.

Each tweeter and woofer are matched with complimentary characteristics in each cabinet and as a stereo pair. The entire assembled speaker is a meticulously matched system.

Harbeth Loudspeakers Factory Visit - Christian from Spitfire Audio talks with Harbeth's chief designer Alan Shaw
https://youtu.be/Ja9BJn0lMxI

Alan talks about the crossover at 19:30 into the video.

[S Clark]

Harbeth doesn't always get glowing reviews.  ....
Edit-- Actually, I went back and checked my notes from 10 years ago at a RMAF, and they were much more positive.  It was another widely acclaimed brand that I really didn't care for 

[Folsom]

Can you read one of the module numbers of the capacitors off as text in this thread, Hobbs?

Depending on the module of the capacitor they're either not very good, or actually really nice at low price. The sandcasts are inductive so you that Mills etc always better, and the inductors depend on the usage but generally speaking just are not air core inductor quality (there can be times they make sense to use).

The capacitors will determine whether or not I'd say it's a bad crossover or a decent/good crossover... either way their implementation isn't bad per se, as we know it isn't a company that is plain old foolish. And there's always a chance for better parts 

[Early B.]

Let's add some context -- there's a hierarchy of attributes required for designing a good speaker or audio component. For speaker design, there are a million other factors more important than crossover parts quality. In fact, crossover upgrades are one of the last places to look to improve the sound. There's no point upgrading the crossover on a mediocre speaker (ask me how I know...). Some free tweaks can yield better results such as proper setup, toe-in, etc. So if a "high-end" speaker has cheap crossover parts, that tells you virtually nothing about how good (or not) it sounds.

 

[WGH]

The crossover in the Harbeth 40.1 is similar in parts quality.



I have heard Harbeth's at RMAF, the sound was always pleasant but also not my cup of tea. I would recommend a Harbeth to a music lover with a lot of money who could care less about extracting the smallest detail out of every recording and would like every recording to sound great no matter the source or quality.

[newzooreview]

Yeah, this video didn't make much sense. It's quite sloppy, in fact.

The data presented show what the presenter calls a very smooth frequency response. In fact, when the grills are placed on the speakers, the little "hump" at 4k also disappears. The presenter, however, fails to state what Harbeth states clearly and repeatedly: their speakers are designed to be used with the grills on. This is noted in most reviews and on the Harbeth site ( https://harbeth.co.uk/userguide/ ).

When the presenter displays the off-axis response curves, the little 4k hump is back. It appears to be that the measurements were made with the grills off, even though the presenter has already shown the data (but omitted the manufacturer's statements) regarding better measurements when the grills are left on as designed.

This sloppiness appears to be repeated in the spectral display plot: the woofer's range is noted to be "really smooth" but the 4k hump has reappeared here as well ("a little bit of stored energy from that peak…but it's not bad; not bad at all").

It very much appears that even after confirming what Harbeth clearly states, the measurements were made without the grills installed. And even then the little 4k is "not bad; not bad at all."

Why not do the measurements with the grills on when they work best that way (and when the manufacturer states that is how they are designed to be used)?

But then the video dives off the cliff into full woo woo. "It measures so well, why not go in and try to upgrade the parts quality?" Huh?

After criticizing Harbeth for designing a crossover that is very successful in smoothing out the response, the presenter criticizes the lack of space for putting different parts in the tiny speaker. I doubt that the manufacturer had "leave room in the cabinet for user-installed parts swapping" as a design goal, but nevertheless the presenter finds fault here.

And then the most amazing statement is made: the presenter states that the cabinets are designed to have a resonance and "have a little bloom down in that lower vocal region." Didn't we just look at a spectral decay and a frequency response that was described as really smooth in the woofer's range with only a 4k bump when measured (against the intended design) with the grills off? Where is this supposed bloom? What's the point of doing measurements if they are ignored two minutes later?

So, measurements are done that show very good performance and then the measurements are ignored and without even listening to the speaker a "bloom in the lower vocal region" is declared and the speaker is criticized because the type of parts used is, without a shred of evidence, declared "cheesy."

It may be that a bloom is audible that doesn't show up in the measurements. It may be that using different parts here, there, or everywhere would improve something in the sound (measured or audible). None of this is evaluated by listening or supported by any measurement. Consistent with the clickbait video title, the end result is sensationalist hot air.

Most boutique manufacturers try to build their reputations by making products that sound good and measure well. It's a shame to see the low road taken where the dominant approach is to attack and say "other people's stuff is crummy," especially when the observations (measurements or listening) don't support such claims.

[TKonrad.NOLA]

Yeah, this video didn't make much sense. It's quite sloppy, in fact.

The data presented show what the presenter calls a very smooth frequency response. In fact, when the grills are placed on the speakers, the little "hump" at 4k also disappears. The presenter, however, fails to state what Harbeth states clearly and repeatedly: their speakers are designed to be used with the grills on. This is noted in most reviews and on the Harbeth site ( https://harbeth.co.uk/userguide/ ).

When the presenter displays the off-axis response curves, the little 4k hump is back. It appears to be that the measurements were made with the grills off, even though the presenter has already shown the data (but omitted the manufacturer's statements) regarding better measurements when the grills are left on as designed.

This sloppiness appears to be repeated in the spectral display plot: the woofer's range is noted to be "really smooth" but the 4k hump has reappeared here as well ("a little bit of stored energy from that peak…but it's not bad; not bad at all").

It very much appears that even after confirming what Harbeth clearly states, the measurements were made without the grills installed. And even then the little 4k is "not bad; not bad at all."

Why not do the measurements with the grills on when they work best that way (and when the manufacturer states that is how they are designed to be used)?

But then the video dives off the cliff into full woo woo. "It measures so well, why not go in and try to upgrade the parts quality?" Huh?

After criticizing Harbeth for designing a crossover that is very successful in smoothing out the response, the presenter criticizes the lack of space for putting different parts in the tiny speaker. I doubt that the manufacturer had "leave room in the cabinet for user-installed parts swapping" as a design goal, but nevertheless the presenter finds fault here.

And then the most amazing statement is made: the presenter states that the cabinets are designed to have a resonance and "have a little bloom down in that lower vocal region." Didn't we just look at a spectral decay and a frequency response that was described as really smooth in the woofer's range with only a 4k bump when measured (against the intended design) with the grills off? Where is this supposed bloom? What's the point of doing measurements if they are ignored two minutes later?

So, measurements are done that show very good performance and then the measurements are ignored and without even listening to the speaker a "bloom in the lower vocal region" is declared and the speaker is criticized because the type of parts used is, without a shred of evidence, declared "cheesy."

It may be that a bloom is audible that doesn't show up in the measurements. It may be that using different parts here, there, or everywhere would improve something in the sound (measured or audible). None of this is evaluated by listening or supported by any measurement. Consistent with the clickbait video title, the end result is sensationalist hot air.

Most boutique manufacturers try to build their reputations by making products that sound good and measure well. It's a shame to see the low road taken where the dominant approach is to attack and say "other people's stuff is crummy," especially when the observations (measurements or listening) don't support such claims.

I concur......

[knotscott]

It's really kind of sad.  There's not a single component on that board I would choose to use on a modern speaker build.  The stuff I used 32 years ago was considerably better quality (IMHO of course).  At some level everything affects sound....most Harbeth owners are certainly at that level, or are trying to be (hidden components that are below the level they should be don't help).  How would you feel if you bought a Ferrari that you later found out had a motor sourced from Kia?

However good the current speakers sound, they would certainly sound even better without significantly altering the tonal balance character (likely clearer) with top notch, or at least upgraded crossover parts. Their crossover design work is admirable, but does seem a tad component excessive.  I'd have a hard time looking someone in the eye if I sold that quality level to them for that price.   I've always had a great deal of respect for Harbeth and many British speaker makers, but I find greed like this disturbing.  Parts choices like these say more about their pursuit of profit than their pursuit of sound quality.  Bad form in my book.

[Tyson]

Yeah, this video didn't make much sense. It's quite sloppy, in fact.

The data presented show what the presenter calls a very smooth frequency response. In fact, when the grills are placed on the speakers, the little "hump" at 4k also disappears. The presenter, however, fails to state what Harbeth states clearly and repeatedly: their speakers are designed to be used with the grills on. This is noted in most reviews and on the Harbeth site ( https://harbeth.co.uk/userguide/ ).

When the presenter displays the off-axis response curves, the little 4k hump is back. It appears to be that the measurements were made with the grills off, even though the presenter has already shown the data (but omitted the manufacturer's statements) regarding better measurements when the grills are left on as designed.

This sloppiness appears to be repeated in the spectral display plot: the woofer's range is noted to be "really smooth" but the 4k hump has reappeared here as well ("a little bit of stored energy from that peak…but it's not bad; not bad at all").

It very much appears that even after confirming what Harbeth clearly states, the measurements were made without the grills installed. And even then the little 4k is "not bad; not bad at all."

Why not do the measurements with the grills on when they work best that way (and when the manufacturer states that is how they are designed to be used)?

But then the video dives off the cliff into full woo woo. "It measures so well, why not go in and try to upgrade the parts quality?" Huh?

After criticizing Harbeth for designing a crossover that is very successful in smoothing out the response, the presenter criticizes the lack of space for putting different parts in the tiny speaker. I doubt that the manufacturer had "leave room in the cabinet for user-installed parts swapping" as a design goal, but nevertheless the presenter finds fault here.

And then the most amazing statement is made: the presenter states that the cabinets are designed to have a resonance and "have a little bloom down in that lower vocal region." Didn't we just look at a spectral decay and a frequency response that was described as really smooth in the woofer's range with only a 4k bump when measured (against the intended design) with the grills off? Where is this supposed bloom? What's the point of doing measurements if they are ignored two minutes later?

So, measurements are done that show very good performance and then the measurements are ignored and without even listening to the speaker a "bloom in the lower vocal region" is declared and the speaker is criticized because the type of parts used is, without a shred of evidence, declared "cheesy."

It may be that a bloom is audible that doesn't show up in the measurements. It may be that using different parts here, there, or everywhere would improve something in the sound (measured or audible). None of this is evaluated by listening or supported by any measurement. Consistent with the clickbait video title, the end result is sensationalist hot air.

Most boutique manufacturers try to build their reputations by making products that sound good and measure well. It's a shame to see the low road taken where the dominant approach is to attack and say "other people's stuff is crummy," especially when the observations (measurements or listening) don't support such claims.

You should watch other videos that Danny puts out.  They are all based around speakers that are sent to him by the owner of the speakers for analysis and possible upgrades. 

Danny tends to break down his analysis into 2 sections - the measurements and the parts quality.  If something measures badly then he will often design a new crossover from scratch to get them to measure well.

On the other hand, if they measure well then there's no need to create a new crossover and so he focuses on parts quality instead. 

It's actually pretty logical, when you think about it.

[newzooreview]

You should watch other videos that Danny puts out.

I've watched every video. It's an endless negative marketing campaign. The measurements typically indicate something that could, theoretically, be resolved/improved. In some cases, at least, there is a new crossover design with new parts that are installed and measured. It's a fairly flimsy approach since nobody, including the presenter, ever talks about how the new parts/design sound, but something is measured at least.

In this case a series of measurements is made without the grill in place, even though keeping the grill installed is designed and observed to be beneficial, and these flawed measurements still repeatedly show what the presenter himself describes as a very smooth response and a very good performance (other than the 4k hump, which isn't there when the grills are used as intended).

It's actually pretty logical, when you think about it.

There's nothing logical about the video. Multiple claims are made in the video without evidence presented. In fact one of the claims, a "bloom" due to cabinet resonance, is mysteriously absent in the evidence presented. The other claim, installing different parts would improve performance, is backed by no evidence at all.

It is certainly appealing to feel that the "truth teller" is lifting the curtain on the "cheesy parts" that competitors are hiding in their products. But the "truth telling" is coming from someone with a profit motive who claims with pure arrogance that he doesn't have to listen to the modified versions of his competitor's speakers that he trashes on his YouTube channel. He just knows that his design sounds better. Shouldn't such assertions be challenged?

There is not another manufacturer on this forum that runs a routine online smear campaign against his competitors.
Strong claims incur a strong obligation for support from evidence.

The emperor has no clothes.

[Tubefly]

I've watched every video. It's an endless negative marketing campaign.

That is just plain false.

[Cheytak.408]

I've watched every video. It's an endless negative marketing campaign.
There is not another manufacturer on this forum that runs a routine online smear campaign against his competitors.
 
Strong claims incur a strong obligation for support from evidence.

The emperor has no clothes.

I am always amazed at people that get their rocks off slamming the efforts of others that are trying to "improve the breed".  I have heard Danny's crossover designs vs. original componentry and they always... yes ALWAYS... sound considerably better, often amazingly better.

In industry the bottom line is the bottom line.  It does not matter what we are talking about in the capitalist world.  An engineer or engineering team will develop a product that performs the way they want it to. Then it is given to the accountants and purchasers that are responsible for ekeing out the most profit possible.  That is what the consumer ends up getting. Often the end result sucks in comparison the the original design.

There is a term that used to be used in audio called POOGE: Progressive Optimization Of Otherwise Generic Equipment.  It is a valid exercise in extracting the last bit of performance from what was originally a good design that is executed with profit as the only goal.  Danny is a master at this when it comes to speakers.

I have known Danny for many, many years. He is one of the most honest, open and finest people that I have ever met.  There is no guile about him.  When he says "A,B,C" he means "A,B,C" and can invariably back up what he says with results. You do him a great disservce by posting what you did and make yourself look a bit foolish to everyone that knows or has dealt with him and GR Research.  Why do you thnk he has such a loyal customer base and gets "Best of Show" awards when he exhibits?

The emperor does have clothes, usually a pair of shorts and a T-shirt.

[77SunsetStrip]

Speakers are sent to Danny from individual owners for evaluation.  Must be something unsatisfactory or the owner would not send to Danny.  The overwhelming majority of speakers Danny measures exhibit something strict objectivists would call out as a problem, without doing any listening.  Criticizing Danny for not listening and presenting his impressions simply holds no water.  The point of suggested modification is to produce as smooth a response as possible, the accepted goal of every manufacturer.  There have been a very small number of speakers that measure well Danny clearly and honestly identifies.

Harbeth is not the only manufacturer using cheap crossover parts.  My eyes have seen plenty.  There are sufficient options for better quality parts without resorting to so called boutique level.  Even assuming Harbeth matched, tweaked, and sprinkled fairy dust the number of parts in that crossover is absurd.  More parts equals more complexity equals more variables to control.  That is a ridiculous engineering approach to speaker design.  If that many parts are required to mesh with the drivers, better find different drivers. 

Until a large number of customers unhappy with Danny's crossover upgrades register complaints, the naysayers should remember the Jim Croce song line "you don't tug on Superman's cape".

[timind]

This thread should get some traction. I owned the P3esr mentioned in the video, as well as the Harbeth C7es3. I get the Harbeth sound, and as someone mentioned above, if you want a nice easy to listen to speaker, Harbeth may be for you. They weren't for me. That said, they are way to expensive to modify with crossover changes. If you can buy the Harbeths, you can certainly buy speakers that suit your taste.