I think that it easy easy to be harsh on hi-fi speaker manufacturers, but it's not easy to make a profit manufacturing anything. It's ok for us DIY folk to buy a couple of speaker drivers, some crossover components and stick them in a box, but a manufacturer needs a guaranteed supply of consistent divers/components that will stay in production for many years (especially if they want to service 30 year old products), they need to employ people to do the work - that means interviewing them, paying there income taxes etc, sick pay, toilet/canteen facilities, HR department, health and safety supervisor, then you've got all your office staff, some guy to stand by the photocopier, cleaners - the list goes on. When the business starts there's an up-hill struggle to get established, all the costs of going to the shows, then the research and tooling costs, then when you've been going for a few decades you might find that you've ended up with loads of incompetent staff that are just dead wood (possibly because of nepotism and cronyism) .
Everything is designed to a price point to be profitable. Speakers are always a set of compromises. Even with cost no object you are still making compromises and the design will focus around what the designer thinks is most important. They are not all bad they are just all "speakers"
Rob 🙂
Rob 🙂
The Peerless TC9 has a shorting ring/cap which is very impressive for a $15 driver obviously designed for a sound bar or TV application. It also has a vented spider and a decently rigid polymer/FG reinforced basket. I think its the best sounding limited LF fullrange for under $50.
Some of the cheaper SB drivers also have copper in the motor. The neo magnet SB26STCN tweeter has a copper cap, no ferrofluid and a vented FG VC former for around 35 dollars. I'd say thats hard to beat for a decent compact 1" soft dome that actually sounds as good as most common $100 soft dome range.
Most people think small plastic bodied neo tweeters are a cheap compromise, but if there isn't any ferrofluid to hide the small details, you end up wth less CTC spacing and therefor a higher possible xover point. The obvious benefit of no FF is the tweeter not ending up in the landfill once it eventually dries up over time, as FF is a bandaid for a poor design. Most tweeters used in commercial designs have FF.
Ferrofluid and hiding details?
Yes, there are quite some very affordable speakers these days with shorting rings 🙂
Yes, there are quite some very affordable speakers these days with shorting rings 🙂
Ferrofluid is detrimental to low level detail retrieval. Most of the best, highly regarded tweeters don't use Ferrofluid.
I think there will probably be caveats.
If you gave him a DIY design then I doubt he'd measure and optimise for free, as there is no market to sell upgrade kits.
I have absolutely not that experience with ferrofluid.
But most ferrofluid tweeters do have a lower Qt as well as a difference in sensitivity, plus sometimes a slightly different frequency response.
If one doesn't compensate accordingly, it's quite obvious why they sound different.
What I have experienced, is that most people don't compensate accordingly.
But most ferrofluid tweeters do have a lower Qt as well as a difference in sensitivity, plus sometimes a slightly different frequency response.
If one doesn't compensate accordingly, it's quite obvious why they sound different.
What I have experienced, is that most people don't compensate accordingly.
Re:
"I've wondered about this as well. I have no idea what I'm talking about, but I reckon lots of what people are saying here holds true: cost cutting measures, ruler flat response not being the only route to a generally pleasing sound, etc."
When we say ruler flat, I think we can safely that that ruler flat on axis, is not as important as we once thought it was.
Here's what Sean Olive said, a couple of years ago, about what matters, based on their research in the late 20th century
For brevity, I draw your attention to main slides at the end-
The whole review-
Initially I posted some graphs of some real.. err... under-achievers. Then I thought- Nah- you can watch them here:
Re: "When Danny Richie mods these commercial speakers, he is generally only sent a single loudspeaker. So, if there are any deviations between even a single pair, Richie is none the wiser.
My thinking is that at the price these speakers are often sold for, such deviations are inevitable, so kneading out minor kinks in the response of a speaker is pointless, due to the uncertainty that every unit produced will have that exact response.
Heck, maybe (probably not) designers even proactively account for this inevitability, and design their crossover such that the speaker will still sound relatively fine even if the drivers within don't perform exactly as intended"
I think this is a good point. As a DIYer, I can massage any kind of frequency response I want. But to bring that into mass production?
Are we being unfair when we criticize single samples? My sister in law, who is a professional statistician, won't even look at reviews from Choice, Consumer Reports or any websites that you and I may use for purchase recommendations. "That's a sample size of 1!" she would say. That's a bit extreme, but I see her point.
I'm reminded of what Floyd Toole's said, during a lecture at the Centre of Interdisciplinary Research for Music Media and Technology @ McGill University in 2015. He shows the spec of a speaker, with a +/- 1dB frequency response.
Listening window 33Hz - 20 KHz +/- 1.0 dB
Now THAT is as close to ruler flat as one can get.
Floyd went onto say-
"Now this is an interesting one. +/-1 dB. Now that's getting serious. That's an audacious challenge. And they... just... squeak in, just squeak it (the +/-1 dB spec) barely. Boy! That's a.. that's a brave thing to do. And I'll tell you why it's a brave thing to you. One of the fundamental problems of the loudspeaker industry is.. that... the consistency... it's one thing to design a golden prototype.. and you might get lucky and do a really good job on that, it's a very different thing to clone that in mass production. And uh, you know, when we do some products, this is one, where we guarantee that it's within spec.. and the only way we can do is to guarantee is that we spend a LOT of money in the production process. (retail price is US$16K/pair in 2015) That's part of what you're paying for, It's the consistency."
Here is a base-line for a cheap, but reasonable loudspeaker US$100 ea-
"Loudspeaker transducers are minimum phase devices, now that is really technical jargon, but I'll quickly explain, it means basically that if the frequency response, the amplitude response, is smooth and flat.. smooth, no bumps, then there's no resonances. If you see a bump, it will ring. Now if you can have an equaliser which can match the little bumps and get rid of the bumps, then the resonances is gone. There's no more ringing. You've corrected the problem. So this is one of the reason why active loudspeakers, with built in amplifiers and digital electronics, are, I would hope, the future, because that's the only way we could really make SUPER good loudspeakers..."
"The M2 is one of them, it used dedicated electronics"...
Here's the M2, as measured by JBL/Harman, pre-Klippel NFS
Same model, as measured by Erin, with his Klippel NFS:
Reference:
https://www.erinsaudiocorner.com/loudspeakers/jbl_m2/
OK, it's it's not quite as smooth as Harman's graph. Now before we debate whether its a design issue or unit to unit consistency issue, or production issue, a Harman vs Klippel's NFS measurement resolution issue (all possible reasons for a very smooth vs less smooth measurement, particularly around 400-600Hz, here's a newer (post 2015) JBL speaker, as measured by Erin:
Reference:
https://www.erinsaudiocorner.com/loudspeakers/jbl_4367/
So the 4367 is is arguably better than the M2.
Btw, Floyd Toole's lecture at the Centre of Interdisciplinary Research for Music Media and Technology @ McGill University in May 2015, is definitely worth watching, if you haven't already.
Now before I go criticizing commercial speakers that I think are designed poorly, let's me reconsider some reasons why this may be less than ideal
1) Traditional design (on axis measurements, popular 1970s-1990s (when we could measure, and tune by tear)
2) Older design (passive/active textbook filters- when we know about basic electrical circuit theory)
3) Archaic design - no measurements, eg. tuned by ear, in the designers own room (when we couldn't/didn't measure)
4) Even if I had a golden reference/prototype design- there are constraints when this are put into mass production with consistency, and thus sample to sample variability.
As DIYers, as much as we criticize, should we also commend?
I purchased this:
https://www.erinsaudiocorner.com/loudspeakers/kali_in-8v2/
And did a teardown. Could I bring into production something that is competitive, for you to buy, that is better?
Kudos to Kali.
Won't play loud enough for you intended listening area? What about this?
Here's one from Dolby.
https://www.spinorama.org/speakers/Dolby CS 128/Dolby/index_vendor.html
Before Danny suggests that he could "improve" this one with his boutique crossover parts that he sells, perhaps he should take a more than a few heavily smoothed measurements. 😊
Attachments
Last edited:
Toole & Olive have done work of great significance. Thanx for bringing up some examples of spinorama measurement sets.
But everything has to be taken with a grain of salt. For instance they generally hold with the(to paraphrase) “all amps sound (much) the same if….”
Leaving many of us to wonder. And suggesting that many of out favorite amplifiers don’t count. So that makes their examination incomplete. And if we are being anal, the results have yet to be indepentently verified, so it is not officially scientifically valid. Seems ASR is asking the question too: https://www.audiosciencereview.com/...dies-to-dr-toole’s-and-dr-olive’s-work.26855/
And the same search turned this headline up: https://www.audiosciencereview.com/...ama-model-incomplete-and-limited.36934/page-2
Not that i take ASR seriously but sometimes they have somethign interesting.
The single most important take away i took from reading the book a number of times, and personally attending a Floyd presentation on the same including questions was:
And as to Danny: Floyd thinks that anything outside an anecoic chamber is suspect… and even then one has to take great care…
dave
But everything has to be taken with a grain of salt. For instance they generally hold with the(to paraphrase) “all amps sound (much) the same if….”
Leaving many of us to wonder. And suggesting that many of out favorite amplifiers don’t count. So that makes their examination incomplete. And if we are being anal, the results have yet to be indepentently verified, so it is not officially scientifically valid. Seems ASR is asking the question too: https://www.audiosciencereview.com/...dies-to-dr-toole’s-and-dr-olive’s-work.26855/
And the same search turned this headline up: https://www.audiosciencereview.com/...ama-model-incomplete-and-limited.36934/page-2
Not that i take ASR seriously but sometimes they have somethign interesting.
The single most important take away i took from reading the book a number of times, and personally attending a Floyd presentation on the same including questions was:
And as to Danny: Floyd thinks that anything outside an anecoic chamber is suspect… and even then one has to take great care…
dave
(1) Some manufacturers spend quite a bit of time to measure and tune their production loudspeakers to match their golden design. Then they keep track of the serial numbers in case they need maintenance or repair down the line. This all costs money, but surprisingly not so much if it's done as a "production plan" not a post production QA repair.
(2) That guy measured the Maggie 1.7s with some mean dip... but I wonder, how did he get that measurement? Was in a close in @ 1 meter measurement or an in room power measurement? Does he take into account how the speakers are designed to power the room?
(3) In the 80s I bought an equalizer with a built in spectrum analyzer, pink/white noise generators and a calibrated mike. I proceeded to equalize my speakers flat from my listening spot. I hate it. So, my next step was to use the analyzer, white noise and mike to move my speakers around for the smoothest sound and then put a finishing touch by ear. It turns out they ended up very close to the starting point. After that, I put the equalizer away and never used it again.
(2) That guy measured the Maggie 1.7s with some mean dip... but I wonder, how did he get that measurement? Was in a close in @ 1 meter measurement or an in room power measurement? Does he take into account how the speakers are designed to power the room?
(3) In the 80s I bought an equalizer with a built in spectrum analyzer, pink/white noise generators and a calibrated mike. I proceeded to equalize my speakers flat from my listening spot. I hate it. So, my next step was to use the analyzer, white noise and mike to move my speakers around for the smoothest sound and then put a finishing touch by ear. It turns out they ended up very close to the starting point. After that, I put the equalizer away and never used it again.
Dave.
As you probably know, as I was taught many years ago, "As soon as something is published in a textbook, it's already out of date"
That might sound like a cynical comment, but what that sentiment means, is that by the time all the research and data is gathered and collected, peer reviewed, and analyzed and digested, written into chapters, edited, proof read, gone to the supervisor, backwards and forwards and edited, then to the publisher, etc. well that's usually in time period of YEARS to DECADES. So by the time it's published into a book form, it's already out of date.
So I assert that half of what is published in Floyd's seminal textbook is wrong. Or more fairly, in 2023, half of what is that book might be be not quite right. The problems is that we just don't know what half.
It's like the thing about THD. Ok so we know it doesn't correlate well with listener preferences. Let's be honest. It's a crapshoot.
But does that mean all forms and levels of harmonic distortion are irrelevant? I don't think that's the same thing.
Here's my post about it:
https://www.diyaudio.com/community/...istortion-speaker-drivers.294787/post-7094796
Now if you're an academic or professional in this field, you can access to latest evidence. Or maybe you work in the industry, and you have access to valuable trade secrets. The rest of us simply do not know. And besides, we lay people, when we get down into the weeds, we get stuck.
Like the amplifier thing. There's no doubt in my mind that different amplifiers can and do sound different in passive speakers... subjectively I've heard it, There is a lot that's going on because it's not a purely resistive load and speakers are non-linear devices. The static measurements aren't fully characterizing what is going on, I get that.
But in an active speaker when the load is a cakewalk, and the amplifier is doing a much smaller bandwidth, the differences might disappear. Or at least they are harder to detect. You know, I have no problem driving subwoofers with 400W Class D. Treble? Now people will say... well they're sighted listening tests. "Don't talk to me until you've run the gauntlet with double blinded listening trials". Yeah, so I think that's a better methodology.
But is it the gold standard upon which we should hang our hat on? Maybe not. Maybe it's not definitive. Maybe there's something else.
Allow me to explain why. (And I'm not going to use any car metaphors or similes here) In the medical world, there's something called randomized clinical trials. It's where we take an intervention and compare it no intervention. (skip this part if this is old news to you, re-enter here #)
So in the days of antibiotics (tablets) we need to compare against something that had no active ingredient. So we have there pill with a bit of salt or water (no active ingredient) that the control group (the group not having active treatment) take. Suppose you have 1000 participants in your study, you divide your study participants, into two groups of 500 people.
Now if 80% of the intervention (eg. treatment) group gets better after the intervention. For some reason, 10% of the control group, who were taking nothing of value, also get better (more on that later). We can say that the real difference is 70%. "Hmm, that's good- Its more than the flip of the coin. It's worth it. Let offer this as a treatment"
# In recent years researchers have discovered a funny thing with the placebo. A sugar/salt/water pill that the intervention group takes that actually makes them WORSE! It typically goes something like this. The intervention group has the intervention/treatment being studied. The placebo group has the salt/sugar/water pill. Another group has nothing (business as usual). The intervention group has a headache, or abdominal pain, or nausea, vomiting etc. "Yeah yeah, I don't have to be a doctor to know all about that. It's called "side effects"
But what's interesting that there are actual objective/physical changes that are measurable in the group that took the salt/water pill, compared to the group that took nothing at all. These differences can be quantified. Things like like reaction time, biochemical markers, on blood tests. And not by chance. Statistically significantly amongst the group.
What?! How is this even possible?
This is called nocebo. And it's an emerging field of study.
And now, if you think I've gone completely off tangent, let me reassure you, it's only circumstantial*
Artificial sweeteners, despite having no calories, can, and DO, cause weight gain.
Reference:
https://gimletmedia.com/shows/science-vs/2oh9ne
(2017)
Now some of you may already think this to be true, intuitively, since the days articial sweeteners came out in the 1980 IIRC.
("artificial sweeteners can't be good" "it's it too good to be true, it probably is")
Now you're not going to get an argument from me... but bear with me.
In the absence of a medical disorder, the medical/physiology/nutritional community thought that weight gain was predominantly an Energy In > Energy Out. One eats/drinks/consumes too much calories (in the form of proteins, carbohydrates and fat) and doesn't use up enough energy (in the form of physiological function, activity, exercise). Of course the diet/weight loss community will say, "oh no, it's more complex than that- it's about what you eat, eat this, don't eat that, it's this vitamin, it's hard mineral, it's about your body shape, you're not doing the right kind of exercise... etc."
But what's the corollary of this-
If it tastes sweet, then yes, it's fooled your taste buds.
But it turns out it also fools your body into thinking it IS sweet and causes some kind of biochemical signaling that starts a cascade of gain weight.
References:
https://pubmed.ncbi.nlm.nih.gov/30983091/
https://www.endocrinepractice.org/article/S1530-891X(21)01157-5/fulltext
So Beware of food packaging saying
"Low fat" (contains more sugar)
"No sugar" (contains something to make it taste sweet)
*cirumstantial"
https://en.wikipedia.org/wiki/Circumstantial_speech
In the 2015 Denmark's 4th largest export was electronics and electroacoustics, behind oil and energy (3), farming (2) and medical and pharmaceuticals (1)
This week the market cap of a single Danish company Novo Nordisk exceeded the entire country's GDP... semiglutide, that anti-diabetes and anti-appetite drug, is set to become Denmark's most valuable export (if it hasn't already)
Is obesity is a form of violence? Coming to a country near you...
As you probably know, as I was taught many years ago, "As soon as something is published in a textbook, it's already out of date"
That might sound like a cynical comment, but what that sentiment means, is that by the time all the research and data is gathered and collected, peer reviewed, and analyzed and digested, written into chapters, edited, proof read, gone to the supervisor, backwards and forwards and edited, then to the publisher, etc. well that's usually in time period of YEARS to DECADES. So by the time it's published into a book form, it's already out of date.
So I assert that half of what is published in Floyd's seminal textbook is wrong. Or more fairly, in 2023, half of what is that book might be be not quite right. The problems is that we just don't know what half.
It's like the thing about THD. Ok so we know it doesn't correlate well with listener preferences. Let's be honest. It's a crapshoot.
But does that mean all forms and levels of harmonic distortion are irrelevant? I don't think that's the same thing.
Here's my post about it:
https://www.diyaudio.com/community/...istortion-speaker-drivers.294787/post-7094796
Now if you're an academic or professional in this field, you can access to latest evidence. Or maybe you work in the industry, and you have access to valuable trade secrets. The rest of us simply do not know. And besides, we lay people, when we get down into the weeds, we get stuck.
Like the amplifier thing. There's no doubt in my mind that different amplifiers can and do sound different in passive speakers... subjectively I've heard it, There is a lot that's going on because it's not a purely resistive load and speakers are non-linear devices. The static measurements aren't fully characterizing what is going on, I get that.
But in an active speaker when the load is a cakewalk, and the amplifier is doing a much smaller bandwidth, the differences might disappear. Or at least they are harder to detect. You know, I have no problem driving subwoofers with 400W Class D. Treble? Now people will say... well they're sighted listening tests. "Don't talk to me until you've run the gauntlet with double blinded listening trials". Yeah, so I think that's a better methodology.
But is it the gold standard upon which we should hang our hat on? Maybe not. Maybe it's not definitive. Maybe there's something else.
Allow me to explain why. (And I'm not going to use any car metaphors or similes here) In the medical world, there's something called randomized clinical trials. It's where we take an intervention and compare it no intervention. (skip this part if this is old news to you, re-enter here #)
So in the days of antibiotics (tablets) we need to compare against something that had no active ingredient. So we have there pill with a bit of salt or water (no active ingredient) that the control group (the group not having active treatment) take. Suppose you have 1000 participants in your study, you divide your study participants, into two groups of 500 people.
Now if 80% of the intervention (eg. treatment) group gets better after the intervention. For some reason, 10% of the control group, who were taking nothing of value, also get better (more on that later). We can say that the real difference is 70%. "Hmm, that's good- Its more than the flip of the coin. It's worth it. Let offer this as a treatment"
# In recent years researchers have discovered a funny thing with the placebo. A sugar/salt/water pill that the intervention group takes that actually makes them WORSE! It typically goes something like this. The intervention group has the intervention/treatment being studied. The placebo group has the salt/sugar/water pill. Another group has nothing (business as usual). The intervention group has a headache, or abdominal pain, or nausea, vomiting etc. "Yeah yeah, I don't have to be a doctor to know all about that. It's called "side effects"
But what's interesting that there are actual objective/physical changes that are measurable in the group that took the salt/water pill, compared to the group that took nothing at all. These differences can be quantified. Things like like reaction time, biochemical markers, on blood tests. And not by chance. Statistically significantly amongst the group.
What?! How is this even possible?
This is called nocebo. And it's an emerging field of study.
And now, if you think I've gone completely off tangent, let me reassure you, it's only circumstantial*
Artificial sweeteners, despite having no calories, can, and DO, cause weight gain.
Reference:
https://gimletmedia.com/shows/science-vs/2oh9ne
(2017)
Now some of you may already think this to be true, intuitively, since the days articial sweeteners came out in the 1980 IIRC.
("artificial sweeteners can't be good" "it's it too good to be true, it probably is")
Now you're not going to get an argument from me... but bear with me.
In the absence of a medical disorder, the medical/physiology/nutritional community thought that weight gain was predominantly an Energy In > Energy Out. One eats/drinks/consumes too much calories (in the form of proteins, carbohydrates and fat) and doesn't use up enough energy (in the form of physiological function, activity, exercise). Of course the diet/weight loss community will say, "oh no, it's more complex than that- it's about what you eat, eat this, don't eat that, it's this vitamin, it's hard mineral, it's about your body shape, you're not doing the right kind of exercise... etc."
But what's the corollary of this-
If it tastes sweet, then yes, it's fooled your taste buds.
But it turns out it also fools your body into thinking it IS sweet and causes some kind of biochemical signaling that starts a cascade of gain weight.
References:
https://pubmed.ncbi.nlm.nih.gov/30983091/
https://www.endocrinepractice.org/article/S1530-891X(21)01157-5/fulltext
So Beware of food packaging saying
"Low fat" (contains more sugar)
"No sugar" (contains something to make it taste sweet)
*cirumstantial"
https://en.wikipedia.org/wiki/Circumstantial_speech
In the 2015 Denmark's 4th largest export was electronics and electroacoustics, behind oil and energy (3), farming (2) and medical and pharmaceuticals (1)
This week the market cap of a single Danish company Novo Nordisk exceeded the entire country's GDP... semiglutide, that anti-diabetes and anti-appetite drug, is set to become Denmark's most valuable export (if it hasn't already)
Is obesity is a form of violence? Coming to a country near you...
Last edited:
You answer that yourself in more or less the same sentence.
Bad correlation doesn't automatically mean that's not relevant anymore.
It means that there is an important nuance that has to be investigated further to find out HOW relavant something is.
That is FAR to simplified as a general statement.
Certain parts of physics were discovered centuries ago and are still relevant and true today.
I've been following GR-Research for a few years and I'm a fan. I think it's pretty clear that he's mostly trying to bolster interest in DIY audio. And yeah, he sells DIY audio...
I give his channel a lot of credit for pulling me back into DIY in a big way even though I've never purchased any GR-Research product. Anything that keeps Madisound, Parts-express, etc. in business is a good thing imo.
No matter what he says in the videos we still get to look inside the enclosures and see the drivers and crossovers.
I give his channel a lot of credit for pulling me back into DIY in a big way even though I've never purchased any GR-Research product. Anything that keeps Madisound, Parts-express, etc. in business is a good thing imo.
No matter what he says in the videos we still get to look inside the enclosures and see the drivers and crossovers.
I don't always agree with him (which is fine), but yes I totally agree with this as well.
Even from a "review" point of view, some people brush this off a unimportant, but it sometimes can tell A LOT about a product or company. In some cases you can even already predict how certain things will perform.
Or how well they have their stuff together, or not at all in some cases!
Besides that, pulling things apart is always interesting and fascinating!
Release that curious inner nerd in you!!!! 😀 😀
Btw, I find the whole commercial argument of it very poor.
In this world there is always some money or politics involved, left or right.
Nothing wrong with it as long as it's transparent and people don't talk from one side only to clearly sell things.
Klippel of all companies, is a commercial company by itself.
There are other ways of providing similar data for a lot less money.
I don't know many people who ever thought it ever was?
Except for the ones who don't understand the concept of resonances and interference 😇
(I am not talking about your average DIY'er, but in a general sense btw!)
To give a bit of an insight in the kitchen.
Last week I had a talk with a friend about a (maybe) potential new loudspeaker product.
He already had a first run just to get some idea.
The most difficult part about mass production:
- Finding ways to make it profitable. So spending 10 minutes on a single screw ain't gonna work (which is fine in DIY)
- Dealing with some UL/CE Safety regulations (plus the involved costs OMG!!), which you don't have with DIY.
- Dealing with customers expectations, which you also don't have with DIY.
In this particular example, the product falls a bit between serious listeners who also want the convenience of changing things with a smart phone for example.
So in this case you spend so much time on the acoustics and all, but people complain that a knob is the wrong color, or doesn't integrate smoothly with app "something" etc etc.
Therefor: they don't like the product.
I have also a lot of experience in the Pro side (sound-reinforcement etc).
You will be blown away how many customers just don't get it.
So it's extremely hard to come up with something new.
I think I mentioned this before, but humans are creatures of habit.
Even more so today. At least in general.
Yes. One of those situations wher eyou have to know more than they do/did. Escher-like situation.
dave