[Early B.]

Danny -- can you explain why most of the speakers that are sent to you measure so poorly??? Are the designers not skilled in their craft? Are they fully aware they're creating poor designs? Is there an economic benefit to sloppy design?

The manufacturers have access to the same measurement tools as you do. And even if they're designing to a lower price point and use crappy parts, speaker designers can still design for a relatively flat response. What's going on?

[Danny Richie]

I think a lot of them simply calculate some values, price it out as cheaply as possible, and see if it hits their price points. Accuracy is lower on the priority list.

[Early B.]

I think a lot of them simply calculate some values, price it out as cheaply as possible, and see if it hits their price points. Accuracy is lower on the priority list.

Oh, wow. I was under the false assumption that speaker manufacturers hired professional designers, but if you don't desire accuracy, any college kid with a calculator can do it.

[Hobbsmeerkat]

Some manufacturers will hire professionals to do the crossover design, but not always.
I'd also wager it also has to do with manufacturers not vetting the performance of their chosen drivers either in or out of the box.
Just picking whatever is cheap & looks good enough and calling it a day. (That's not to say that expensive drivers will measure all that well either...)

We had the Neumi Silk 4 in a few months ago.
Fairly well behaved speaker for $150 a pair, but swapping out just 3 parts will get you an even better response.. (and would cost maybe $25-30 + shipping if going the budget route from Parts-Express)

[Early B.]

Seems like many speaker manufacturers are taking advantage of the general public's lack of knowledge about sound and build quality. It's easy to dazzle people with hyperbole rather than educate them and deliver a true quality product. I understand this approach for low-end speakers, but when you drop $3K or more on speakers that measure poorly, what did you really pay for?   

[toocool4]

Measurement is not everything, listen with your ears in your own system. If you like what you hear, buy it and if you don’t send it back.
If I bought speakers and it sounds good to me, but measures badly guess what I would not care.

[FullRangeMan]

I thought you knew that.
If one want precision, musicality and sense of presence in the room, should avoid using cross-over, even a simple two-way first order. Commercial Hi-Fi speakers always will have Tweeter beaming, Phase rotation, driver sensitivity loss, Harmonics loss at any price point, the more parts the cross-over has worse.

It should be noted that the US audiophile buy anything, as long as it has a famous brand name and fancy looking. The sound quality doesn't matter, he average joe doesn't know if what he's hearing is good or bad due low benchmark reference.

[Early B.]

It should be noted that the US audiophile buy anything, as long as it has a famous brand name and fancy looking. The sound quality doesn't matter, he average joe doesn't know if what he's hearing is good or bad due low benchmark reference.

I'm a US audiophile and I'm not offended by this statement.

[Early B.]

Measurement is not everything, listen with your ears in your own system. If you like what you hear, buy it and if you don’t send it back.
If I bought speakers and it sounds good to me, but measures badly guess what I would not care.

What you're saying works fine for most people. The stock radio in my car sounds good to me, but as an audiophile, good is never good enough. The main rig has to exceed my expectations of what good music sounds like. To get there, I need speakers that measure well because if they don't, I'll never get beyond good sound. That includes crossoverless speakers, FullRangeMan, because they measure just as poorly as other speakers.   

[jmimac351]

Seems like many speaker manufacturers are taking advantage of the general public's lack of knowledge about sound and build quality. It's easy to dazzle people with hyperbole rather than educate them and deliver a true quality product. I understand this approach for low-end speakers, but when you drop $3K or more on speakers that measure poorly, what did you really pay for?

A timely question... watch pieces of this interview the other day with Daryl Wilson.  If you watch the entire video, you can make up your own mind about hyperbole (I agree with you).  I like Daryl Wilson... he seems like a super nice guy...

1) Flat Frequency Response - Wilson Audio https://youtu.be/ptX7wdEet_k?si=B0ZYcUMJ_xA3uVaM&t=949

Question to Daryl Wilson: "Do you aim generally for an objectively measured flat frequency response when you test the loudspeakers?"

Daryl Wilson answer: "That's where we start. I think any loudspeaker worth its weight in salt can get a relatively flat frequency response." 

2) Ribbon vs Dome Tweeters - Wilson Audio https://youtu.be/ptX7wdEet_k?si=Ye1Vpi4itB3g-m51&t=705

Question to Daryl Wilson: "Why do you prefer dome tweeters, soft dome tweeters over say, ribbon tweeters.  Have you ever experimented with ribbon tweeters?"

Daryl Wilson Answer: "Yes... they can be very flat... as soon as you go off axis even 1 or 2 degrees you start getting very extreme rolloffs..."

++++++++++++++++++

Wilson Audio Duette - Original Response vs GR Research Corrected Response:https://www.audiocircle.com/index.php?topic=187454.0

Daryl Wilson: "That's where we start. I think any loudspeaker worth its weight in salt can get a relatively flat frequency response." 

[FullRangeMan]

I'm a US audiophile and I'm not offended by this statement.

In my country happen the same thing with the current generation and millennials.

[JLM]

Speakers are auditioned in flawed ways:  too short of time; not at home; to meet expectations (not live/unamplified music), staring at the veneer/cabinet size; design type. 

[AllanS]

Which measurements are important.  Mr Danny posted this video the other day of a budget brand, complimenting the designers on achieving a commendably flat frequency response and well controlled spectral decay. But he also went on to explain how they accomplished this and the consequences of the choices they had to make to hit a price point.

Danny explains that achieving a flat frequency response is relatively easy to accomplish.  If a consumer is knowledgeable enough to go looking, the on axis frequency response is likely the only measurement they will find if anything is available.  It’s the easiest to understand, the easiest to manipulate by smoothing and, if only considering the in room response in your room, the easiest to correct for.  Note that I didn’t say easy to correct, just easiest.

The spectral decay is interesting. In this case it appears a primary time domain response guided some of the designers decisions and choices.  But time domain responses is the point at which things get very complicated very quickly and where most consumers eyes probably begin to glaze over.  You have to be invested in the hobby to even begin to grasp the concepts let alone measure your system and to determine what if anything to do about measured results.

If I’m understanding correctly, spectral decay is just a first level time domain response characteristic.  It measures what frequency is arriving at the mic and how quickly that frequency decays over time.  It doesn’t tell you if a given frequency arrived at the correct time.  I’m still trying to sort this out so may be wrong, but I think this is where impulse responses and phase angles are important and where the smearing Danny talks about comes from.

Speaker design choices directly affect both frequency and time domain responses.  The room contributes its own frequency and time domain responses. We measure the combination as in room response characteristics.  We can, to a certain degree, correct the room’s contribution through treatments, subs, and DSP.  I’m not sure about this: DSP may try to compensate for speaker deficiencies but cannot correct the deficiencies.

https://www.youtube.com/watch?v=r6nol7rZ5cg

[Early B.]

Which measurements are important.  Mr Danny posted this video the other day of a budget brand, complimenting the designers on achieving a commendably flat frequency response and well controlled spectral decay. But he also went on to explain how they accomplished this and the consequences of the choices they had to make to hit a price point.

Danny explains that achieving a flat frequency response is relatively easy to accomplish.  If a consumer is knowledgeable enough to go looking, the on axis frequency response is likely the only measurement they will find if anything is available.  It’s the easiest to understand, the easiest to manipulate by smoothing and, if only considering the in room response in your room, the easiest to correct for.  Note that I didn’t say easy to correct, just easiest.

Let's keep it simple and look solely at frequency response. If a flat frequency response is so easy to accomplish, why do so many speaker manufacturers get it wrong? Of course, a flat response can still sound...ummm... flat, but it's a starting point. I would like to know why the designer deviated from a flat response. There are plenty of good reasons to do so, including overall design choices. For instance, Klipsch designs their speakers to be tipped up in the highs. That's good to know. As audiophiles, we should want at least a basic knowledge of how a speaker should be designed properly, and Danny's videos are invaluable. However, I shouldn't have to send my $5K speakers to Danny so he can fix them. I'm not a measurements guy at all, but I'd like to see more honesty in this hobby. Third-party measurements, not marketing hype, is one way to get there.     

[jmimac351]

I thought you knew that.
If one want precision, musicality and sense of presence in the room, should avoid using cross-over, even a simple two-way first order. Commercial Hi-Fi speakers always will have Tweeter beaming, Phase rotation, driver sensitivity loss, Harmonics loss at any price point, the more parts the cross-over has worse.

It should be noted that the US audiophile buy anything, as long as it has a famous brand name and fancy looking. The sound quality doesn't matter, he average joe doesn't know if what he's hearing is good or bad due low benchmark reference.

No crossover...

Here's a speaker I own... EgglestonWorks Fontaine.  There is no crossover on the bottom 2 woofers (they play FullRange, Man) and an L-pad setup on the Esotar tweeter.  The tweeter and woofers, from the factory, were wired out of phase, which caused a huge suckout. It sounds like azz with this configuration. Just putting the tweeter back in phase with the woofers made it sound better.

I get doing "full range" with a speaker designed to do it... but this is probably one of the best examples of a speaker that leaves the factory with horrible measurements.  There are people holding Klipsch signs outside my house asking to be my friend, but I don't have the heart to tell them the speaker measures better now. 

Original Measurements



Overlay - Original Measurement / Tweeter Polarity Corrected / Tweeter Polarity Corrected & Notch Filter Added

[Hobbsmeerkat]

TBH, flat-measuring speakers can be easy to achieve, but it's going to rely on selecting well-behaved drivers in a properly designed cabinet and ensuring that your measurements are done correctly, (and time aligned, if using a crossover design software like XSim) and that your drivers can actually meet one another at the desired crossover point.
We've seen manufacturers who selected woofers which had breakup issues at or before the crossover point, and that's not really fixable.
We've also been sent custom speakers where the woofer naturally rolls off before it can ever reach the chosen tweeter.

It's why things like canned/calculated crossovers tend to be such a mess 99% of the time.. There are several factors that are often not being taken into consideration..
Baffle-step, edge diffraction, acoustic spacing, cone breakup, just to name a few.

A lot of times for manufacturers, the main things to consider for the average person are price and looks, everything else is often secondary, so long as there are no glaring flaws. (Though that's not always true)
If a speaker is going to retail for $1000, they have maybe $200-250 in the drivers, crossovers, cabinet and packaging. 
It's all a balance of tradeoffs, even with full-range drivers.

In the end, especially within the retail space, it's largely about what appeals to the most end users.

[Freo-1]

I've become quite biased regarding speakers.  For me, I keep going back to speakers from ATC and Dynaudio.  I especially like the ATC speakers.  They use very high quality drivers, with excellent dispersion.  The ATC dome midrange sounds wonderful.  They measure well, and can handle dynamic transients better than many other speaker brands.  In other words, they play louder cleaner (with less distortion).

[jmimac351]

I've become quite biased regarding speakers.  For me, I keep going back to speakers from ATC and Dynaudio.  I especially like the ATC speakers.  They use very high quality drivers, with excellent dispersion.  The ATC dome midrange sounds wonderful.  They measure well, and can handle dynamic transients better than many other speaker brands.  In other words, they play louder cleaner (with less distortion).

Stop.  Please.  There is a pair of ATC speakers (ATC SCM12 Pro Studio) 15 minutes away from my couch and I really don't need another pair of speakers... or do I? 

[Freo-1]

Stop.  Please.  There is a pair of ATC speakers (ATC SCM12 Pro Studio) 15 minutes away from my couch and I really don't need another pair of speakers... or do I? 

But of course.  How can you not?   

[Tyson]

It is disappointing that so many speakers measure badly.  I think it's a combo of a few different things. 

For big companies, I think they are more manufacturing and not so much design outfits. 

For small mom and pop companies I think they don't have a lot of resources and try to use design software and don't actually understand how to measure properly a driver in a box. 

The good thing about channels like Danny's (and Erin's audio corner) is that they can and do measure everything and publish the results.  This helps in 3 ways:

1. Consumers are more informed and can make better decisions
2. Manufacturers see the results and (maybe) are motivated to do better
3. Manufacturers see the methods used to measure and can use that understanding to create better products