I wonder how much distortion is associated with the ears and brain in the average listener?

Did you actually read the beginning of this thread?

Quote:
I read that loudspeakers have 2% distortion at average listening levels. TWO PERCENT !! Power amps often have .01% distortion.

You can read something but that does not make it true. Who said it? What data/measurements was it based on?

Kal

Quote:
For starters, it's mentioned on Wikipedia - twice. First under the topic "audio system measurements" in the "total harmonic distortion" section. Then under the topic "loudspeaker measurement" in the "distortion measurement" section. These reports say that loudspeaker distortion actually rises to FIVE PERCENT during playback.

If a power amp was rated at 1% distortion (fundamental), it would be considered faulty-broken.

Wikipedia is only as good as the references that support the entry. Can you provide it?

Quote:
For a science-based look try:
http://geocities.com/la1zka/hifi/hornpage.html

Old news based on PA drivers and horns (with their inherent distortion).

Now, I am not disputing the point that loudspeakers and, indeed, other transducers have higher distortion than do modern electronics. But that is not a secret. The old speaker reviews in Audio magazine documented that decades ago.

Kal

Quote:
If a power amp was rated at 1% distortion (fundamental), it would be considered faulty-broken.

The classis tube power amplifier, spec'd at 1% T.H.D.

http://home.indy.net/~gregdunn/dynaco/components/ST70/index.html

Does no one remember that it's not how much harmonic distortion there is but how much of what kind of distortion there is?

Quote:
You can read something but that does not make it true.

Quote:
The article on Wikipedia stands because no one has *disproven* these facts !!

That's hardly science. Wikipedia is a consensus of those who participate.

Quote:
Do you think someone made these numbers up ?

Nope but I still would like to see data.

Quote:
Why can't Stereophile publish figures of loudspeaker distortion to re-but all this ?

Dunno but I suspect that JA has dealt with the issue before. I hope he comments or directs us to his comments. The only thing I could find is http://stereophile.com/features/100/index6.html

Quote:
As for the published graph article, it wasn't just about "old PA drivers", if you read the entire piece. It was showing how direct radiators work as well.

I didn't see reference to anything but horns and direct radiator PA drivers.

You can infer what you will. Doesn't prove anything.

BTW, if you want to read speaker distortion measurements, try these http://www.audiovideonews.com/speakermeasurements.html.
Most of them are fairly innocuous.

Kal

Quote:
As for power, I would never listen to a tube amp with *that* much distortion.

Two things. The 35 watt Dynaco didn't produce 1% T.H.D when it played at less that rated power. Also, it is the structure of the distortion component, low order to high order that generally makes for a bothersome sound or a not so bothersome sound. Most tube amps have very little in the way of high order harmonic distortion and they possess a "euphonic" degree of low order disortion. (A simple explanation of "tube sound" is this distribution of harmonic distortion components.) Depending on the type of tube and the output configuration of the tubes, they can serve to cancel some harmonic content.

Someone can correct me if I'm wrong, but other than the basic nature of a push pull amplifier, that sort of cancellation is not possible when bipolar outut transistors are employed.

So, while 2-3% T.H.D. is substantially higher than 1% rated T.H.D. as you first posted, you really need to know the make up of the distortion component before you decide "how it will sound". And, of course, moving to a source component from an amplifier to draw your distinction makes a huge difference in how much distortion is going to affect your perception of the sound. A bad source will always be a bad source.

Most dynamic drivers on their own have a very low amount of high order harmonic disortion and therefore what you hear from a loudspeaker cannot be equated with what you hear from an amplifier. Like a tube amplifier, the distortion produced by a driver in a well designed and correctly damped enclosure will be mostly of the "musical" variety that tends to complement the music instead of aggravating the signal.

Quote:
Kal: try the Soundstage! figures.

???

Quote:
Or just listen to a new-generation rear-loaded horn.

I do, periodically, at shows.

Kal

Quote:
All audiophile loudspeakers - save rear-loaded horns - have very high distortion, harmonic and otherwise.

You may have forgotten about electrostats.

> I read that loudspeakers have 2% distortion at average listening levels.

This is not correct as a general statement for a reasonable loudspeaker. Here is the specification of a good small loudspeaker:

http://www.klein-hummel.com/klein-hummel...-monitors_O300#

To get 2% distortion one needs to either play loud enough to approach "hitting the stops" or play deep enough to approach "hitting the stops".

> TWO PERCENT !! Power amps often have .01% distortion.

Indeed. This is why people with a genuine interest in accurate reproduction purchase cheap amplifiers (by audiophile standards) and spend almost all their budget on speakers.

> Why isn't Stereophile telling us this ?!

That is a question for Stereophile. However, I will say that Stereophile is a publication targetted at audiophiles (people with an interest in luxury home audio products and the beliefs that are associated with them) and not people with a professional interest in sound like musicians, sound engineers and the technical/scientific community in general. Confusion is almost certain if you do not make the distinction.

> And why aren't more designers building compression-driver based horns to overcome this ugly fact ?

If you want low distortion without a lot of signal processing then you need to compress/move a lot of air a small amount. Horns do the opposite and move/compress a small amount of air a lot in the throat and this generates distortion because the pressure-volume compression curve for air is nonlinear (Boyles law and all that from school Physics classes). If the pressure change associated with sound is a significant fraction of the atmospheric pressure you will get significant distortion.

In truth things are not quite as clear cut because a good gentle horn/wave guide can make a bigger improvement to acoustic efficiency (play louder) than it degrades the distortion (it also has other benefits and problems). Hence for a given driver and a given SPL a waveguide will often give a lower distortion which is one of the reasons they are widely used in accurate loudspeakers (e.g. recording monitors). They are fairly rare on audiophile speakers.

> Are they *satisfied* with 2% distortion ??

High distortion occurs at low frequency and/or high SPL. This is best reduced by increasing the area driving the sound but the trend in home audio designs has generally been towards smaller speakers. Speaker design is a compromise of all sorts of factors but the signs are that low distortion is not a heavily weighted parameter for home audio.

Quote:
Kal:

We are referring to the same site. But even here they say in "How we test Loudspeakers" (THD+N section), that distortion measurements for loudspeakers are "many times" that of electronics. But there's more. In their Deviation from Linearity section, they say that most speakers start to show "serious deviations" at 95db.

How many more reports do you need to accept these ugly loudspeaker facts ?

C'mon. We have known these 'facts' for decades. What we are talking about is their significance.

Quote:
And you think the distortion figures here are harmless ? I don't.

While less distortion is always preferable, I still have not seen evidence of the effect of distortion on the subjective experience. I am more than half way through Floyd Toole's new book on reproduction and the topic has not even come up.

Quote:
Finally, listening to a speaker at a show can be a world different from the home experience. But even here, you should have been impressed in recent cycles. Many others were....

I do not wish to piss off the horn fans so, yes, I have been impressed by some aspects of their sound but not the overall presentation. At least, so far. But, again, this is a red herring since you are presuming that distortion is a parameter of what you perceive as superior performance.

Kal

Quote:

"JV": Nice theory, but there is a reason why CD players and preamps cannot have 3% distortion, literally. And just how do loudspeakers produce a "musical" variety of distortion ?

And why are folks reporting massive gains in sound quality when comparing a direct-radiator to a modern compression-driver horn ?

You infer that I like distortion or find it acceptable. How did you come to that conclusion?

Ignoring what is posted is not an effective debate style. Read my post again, this time with an eye toward what is being said not just what you wish to see.

You seem to have fallen into the basic trap of this one "thing" will make reproduced music accurate. That would be, sorry to say, dumb. Horns have advantages and disadvantages just as everything else I've encountered in audio over the last four decades has had those same sorts of trade offs. Every one of us picks and chooses what we find acceptable and what we find not acceptable. To tout horns as the be all and end all of audio reproduction is to ignore such wisdom.

Quote:
What I *do* know is that a typical high-end system is a world away from live, acoustical music. You cannot deny this.

What is your idea of a "high end system"? I think that would serve this discussion well, to know what you consider to be "typical".

Quote:
Most audio researchers say that the loudspeaker is *by far* the most corrupting element in the playback chain.

Where did you come up with that bit of information?

Most audio researchers I know of say that the source is "by far" the most corrupting element in the playback chain.

Garbage in = garbage out.

Just inserting your own "facts" into the debate is equally as false as only knowing how to interpret one fact.

Quote:
Kal:

You must not be listening to a high-end system for any length of time. Maybe as background music or maybe to impress your friends, I don't know. What I *do* know is that a typical high-end system is a world away from live, acoustical music. You cannot deny this.

Most audio researchers say that the loudspeaker is *by far* the most corrupting element in the playback chain. And if we're not even *close* to live music at home, what do you think is blocking us toward achieving our goal ?

If you're sitting there at home - day in, day out - *thinking* your speakers are accurate (with those distortion figures, frequency response deviations and sharp reduction in SPL's), then may God help you...

Finally, I wouldn't trust Toole. Here's a guy who says that "all components sound the same", if level-matched. Do you believe this ?

JPH

I have not gone away but I find nothing of substance in this post, just opinion and innuendo.

Kal

> When I said "most" loudspeakers, naturally I was referring to audiophile types - which is not the type of product you
> linked us to.

Manufacturers of audiophile loudspeakers do not provide plots of distortion against frequency as part of their specification whereas K&H do. The plots of distortion at constant 95dB SPL and the plots of SPL at constant 1% and 3% distortion were intended to give you some idea of how distortion varies for a competent cones-in-a-box loudspeaker. There is nothing magical about K&H loudspeakers although they obviously come from the top end in terms of quality.

> According to the "consensus" on Wikipedia and the measuring gurus at the National Research Council, speakers of our
> (audiophile) type do indeed distort at 2-3%.

All loudspeakers can be driven with a signal that results in them distorting at 2-3%. It is a meaningless statement without qualification about the input and the output.

> As a disgruntled audiophile, I now realize that our loudspeakers were not accurate after all - with regards
> distortion, frequency response deviation and SPL peaks at typical listening distances.

You did not notice that loudspeakers actually did sound different from each other?

> That said, I think you're a little misguided as to horns - they *are* much lower in distortion. The air in the cavity
> (temporarily) creates a higher level, yes, but this is not what we hear in the end.

I think we are agreed that the high compression at the throat causes high distortion. Are you stating that this distortion then disappears?

> The "geocities" link I provided above talks more about this.

This is an honest attempt to present the pros and cons of horns?

> And I don't think it's true that we have to "increase the area driving the sound" to lower distortion.

So much for physics.

> Horn theorist say we have to do just the opposite !! They say that horns are the "most natural concept" towards
> acoustic transduction.

What is a horn theorist? How horns work has been widely known for a very long time and they were widely used in the early days of sound reproduction.

> The latest listening impressions clearly back this claim up...and I am not alone in this assessment....

Like most with an interest in sound I have heard a few horn loudspeakers over the years. "Natural concept" did not strike me.

If horn loading is superior why is it rarely used by those that earn their living from listening closely to music reproduced by loudspeakers?

You haven't made this easy. Easy is answering the questions you've been asked. That is the basis of a discussion. What you are doing is forming the sticking points of an argument. I don't care to argue, I came here to discuss.

I'm going along with Kal at this point, you have something you need to prove so you can show how smart you are. What anyone else says is irrelevant to you being smart. "Smart" people argue, intelligent people discuss.

Let's work on some things - and you'll have to pay attention and answer what you're asked, not ignore it and not change the subject.

Quote:
can a CD player or preamp have distortion of 2-3 %.

What sort of distortion component makes up that 2-3%? I can't answer your question when you so poorly define what you are referring to.

Quote:
You have not provided evidence on how a speaker produces "musical" distortion.

I have, you have just ignored it. Do this, describe "euphonic" distortions. I've mentioned them in this thread, what did I say?

Quote:
And I did not say that horns were a perfect cure.

Neither did I. I said everything is a compromise.

Quote:
This latter point is important if we are trying to achieve concert-class sound at home. Do you care about these things ?

I care about the music that is the most interesting to me. I don't sit and concern myself with what measurements are coming from my components. That's foolish.

So, what do you care about?

Quote:
JV: Let's make it easy (again) - do you think playback systems sound like real music ? I hope not...

Why on Earth would you not hope my system provides me pleasures similar to what I experience when I hear live music? I do not wish such ill favor for you and your system.

Quote:
And I don't know who you're listening to concerning sound reproduction, but everyone from John Eargle to Sigefried Linkwitz says ...

So, you are allowed your "experts" but I am not allowed my own? My experts say "garbage in = garbage out". They have for nearly forty years. How can anyone disagree with that logic? Do your experts agree with that logic? I would think so. If they do, why do you not agree with your experts?

Quote:
... that loudspeakers were the primary source of audio corruption - by an order of magnitude. Isn't it true that loudspeakers vary wildly in sound ? That means that they're the weakest link !!

Then they would be wrong by amile. Each speaker will sound different in each room and in each position in any one room. That would make the room the weakest link in any audio reproduction system. I know of no one who claims to know what they are talking about other than you who would disagree with that concept. "The room" is anywhere from 50-90% of what you hear.

Quote:
CD players don't sound *that* much different - although I will admit that the best digital system I ever heard (codeless memory drive w/ German Trinity DAC) was a huge leap, but this was exceptional.

I am not interested in name dropping, it doesn't impress me in the least. But with that statement I would say you are not listening to what I am hearing.

Therefore I repeat the request, please describe what you consider to be a "typical" high end system.

I'm with Kal. There's nothing here.

-The input is usually standardized when measuring loudspeakers, isn't it ?

I am not sure what you mean by standardized. Your 2% distortion has little meaning without mentioning the type of signal and the SPL.

> So the NRC numbers are not "meaningless".

It was your statement that was effectively meaningless. The people at NRC will have given the type of signal and the SPL.

> -Horn "distortion" must disappear if they measure (and sound) much lower in distortion in the end.

Indeed this would be true if they did measure (and sound) much lower in distortion. Do you know where this distortion went? If it disappeared in the horn how would whatever was running off with it know, for example, what was 2nd harmonics from the non-linear compression (to be removed) and what was 2nd harmonics from the musical instruments (not to be removed)?

> But if you reject the "geocities" link, then I guess I can't force the issue.

I did not reject it. I asked if you considered it an honest report and you did not respond. I know how horns work and so enjoyed skimming the presentation.

> -Increasing the driver area only increases the chance that the unit will not operate as perfect-piston.

That is good to know. So we can use small $1 tweeters in our subwoofers and forget about having to shift all that air?

> What physics do *you* read ?

I read the technical press as part of my 9-5 job.

> -Don't know what a "horn theorist" is ? Someone who makes horns.

Makes? That does not sound particularly theoretical.

> Of course, we have to check their claims with measurements and listening...and both of these are checking out pretty
> good these days.

Can you provide a link to the equivalent distortion plots as the K&H ones for your horn speakers?

> Have you heard (at length) a recently designed rear-loaded horn ?

About 2 years ago I listened at reasonable length to a pair of very large concrete horns and they were OK. Not like a monitor but one could work out what might be attractive. I was also demonstrated a single "full range" drive unit in a cabinet loaded by a folded horn. I could perceive nothing attractive despite the enthusiasm of my host. While I struggled to be diplomatic he shot off and brought back sets of measurements which I would have kept hidden.

> -Why don't more co. make horns ? They are. But compression-driver based horns are still an emerging field. The
> reasons being that they're costlier to build and it takes more work (R&D) to produce results, due to so little
> previously published info.

Compression drivers have been used almost since the beginning of loudspeakers. Loudspeakers are straightforward, well developed low technology devices. There is certainly development centred around advances in materials and drops in the costs of previously overly expensive materials/approaches but very little that could be classified as proper research except, of course, in marketing material for audiophiles.

> The good news - at least here - is that quality matters. And the more innovative and forward (not backward) thinking
> engineers in high-end are making horns. But most audiophiles still don't know about this - in time they will.

Perhaps. Audiophile quality is quite different from technical sound quality and does not seem to be constrained by the laws of physics.

> At least you admitted that audiophile loudspeakers don't emphasize distortion reduction and are unlistenable.

I did not say that audiophile loudspeakers are unlistenable.

> Whether closed or dipole in execution, horns are the future of high-end audio.....

It seems unlikely to me given the evolution over the 30 year life of the sector but I readily admit to having no better idea of what audiophiles will find attractive in the future.

Quote:
Don't you read AES papers ?

Another "don't you"? What is it with everyone and their "don't you" questions? Why all this arguing from some supposed authority everyone has if someone hasn't done what the other fellow has?

No, I no longer read AES journals. That doesn't make me wrong and you right.

I've never walked on the moon. What does that make me - not a moon walker? Somehow because you subscribe to a periodical you are now the expert?

I really can't begin to tell you how many "experts" I've known who don't have the common sense to balance their own checkbook. So, please, let's not go there. OK?

Quote:
I've named some experts who stress the loudspeaker as the main culprit - please provide your people who say it's the room's fault.

Why am I doing this?

We are talking in circles. The room and the position of the speakers and listener within that room have the largest influence on the final sound. Do you agree?

I use point source dipoles, thank you.

Quote:
-Finally, thanks for agreeing with me that loudspeakers would be "wrong by a mile" if they sound very different.

I keep asking people this same question on this forum. How did you arrive at that conclusion after reading what I posted? My post was not so obscure that you shouldn't have been able to comprehend what was said.

As per andy ...

Quote:
I asked if you considered it an honest report and you did not respond.

You've noticed that trait too, I see.

Lots of good stuff, and a fascinating topic!

1) I think 2% is pretty darned good, but I'll go along with ragging on it as 'high' dstortion for the sake of the thread. However, consider: Perhaps more important to our ears is the other stuff a speaker does besides the distortion.

2) As volume increases, our own ears start to distort more, too. This may light people on fire here, but there are many studies on this issue. The increase in speaker distortion with increasing volume may actually increase at a rate that is less than our own ears do the same thing. In other words, rising distortion with rising speaker volume may not matter all that much, since we starts distorting more than the speaker does!

Here's a cool link:

Higher volume can mask perception of distortion products

3) When we talk about flat out 'distortion,' we are using language that is too simplistic. There are many kinds of speaker 'distortions,' not just THD.

Getting worked up over one speaker's THD being 2% and another's 50% higher at 3% may be a terrible way to compare speakers.

4) There is a difference between accuracy and precision. A speaker with 2% distortion may still be very good at allowing us to evaluate upstream changes if it is the same speaker every time, so to speak.

5) I don't wholly agree with Jan, but upstream 'distortions' are much like hitting a golfball.

At one microsecond after contact, the path of the ball may be only 1/1000th on a degree off true, but by the time the ball travels 250 yards over 4 million microseconds, the importance of that original 'error/distortion' becomes blatantly obvious. I hope Jan would approve of this analogy, actually.

So, lots more to it than measuring a distortion product and calling it done.

Anyone else remember Velodyne's adventure outside the realm of subwoofing?

They produced a loudpseaker with the lowest distortion ever measured, and it pretty much sucked.

Can I get a witness?

First, Kal has been writing for Stereophile forever (Sam Tellig and Larry Greenhil might have him beat but that's about it.) So he may well have heard more high-end systems in a week (at CES) than even an active hobbyist would get to hear in a year. When you project that out over many, many years and he's heard more gear that any of us could hope to experience.

Second, as Kal said this isn't news. The fact that speakers have more distortion than any other component (unless you count "the room" which I do)is well established and is common knowledge. That's why stereophile doesn't talk about it. A loudspeaker is a device that takes electrical energy and converts it into mechanical energy. It's MUCH, MUCH, MUCH more difficult to get a mechanical device of any type (not just audio) to act as lineally as an electrical device. I'd bet that the same is true for phono cartridges too.

Quote:
First, Kal has been writing for Stereophile forever (Sam Tellig and Larry Greenhil might have him beat but that's about it.) So he may well have heard more high-end systems in a week (at CES) than even an active hobbyist would get to hear in a year. When you project that out over many, many years and he's heard more gear that any of us could hope to experience.

Second, as Kal said this isn't news. The fact that speakers have more distortion than any other component (unless you count "the room" which I do)is well established and is common knowledge. That's why stereophile doesn't talk about it. A loudspeaker is a device that takes electrical energy and converts it into mechanical energy. It's MUCH, MUCH, MUCH more difficult to get a mechanical device of any type (not just audio) to act as lineally as an electrical device. I'd bet that the same is true for phono cartridges too.

Good one.

Personally, I vacillate between being critical and being astounded that the damn thing works at all!

Great reply, man.

Quote:
Can I get a witness?

Certainly, being accurate does not always ensure being interesting. The two sides of this debate have been going at it for decades. Is it better to have an "accurate" reproduction that is uninteresting and only serves to point up all the flaws of most modern recordings to the point off turning the music off after a short time or inducing headahces in the listener? Or is it better to have a "beautiful" system that leaves you wanting more?

Having heard both, I'll take the latter. That doesn't mean I don't value things with accuracy attached to them. There is a place for Wilson Watts.

Quote:
There are many kinds of speaker 'distortions,' not just THD.

Someone here still has not said they understand the ability of distortion components to add a musical or euphonic tinge to the sound. If you do not understand how that can occur, then you do not understand distortion IMO, no matter how many AES journals you've read.

Quote:
I don't wholly agree with Jan, but upstream 'distortions' are much like hitting a golfball.

At one microsecond after contact, the path of the ball may be only 1/1000th on a degree off true, but by the time the ball travels 250 yards over 4 million microseconds, the importance of that original 'error/distortion' becomes blatantly obvious. I hope Jan would approve of this analogy, actually.

I understand the analogy, I don't understand why that golf ball had to travel 250 yards in "4 million microseconds".

Sounds like techno-babble to me.

And does it matter how far it travelled if it was headed off course?

But then I suppose someone will ask, "Aren't you a golfer?"

However, in your analogy, you speak of distortion which is not exactly what I'm referring to when I say "garbage in = garbage out". Distortions are only a small part of why that saying is almost universally true.

Quote:
-And you said it clear as day: speakers are "wrong by a mile". Do you even read your own posts ??!!

You must have the attention span of a chipmunk on speed. Read the whole post.

I'll try to make this easy for you since you like "easy".

I quoted your sentence ...

Quote:
And I don't know who you're listening to concerning sound reproduction, but everyone from John Eargle to Sigefried Linkwitz says (or said) that loudspeakers were the primary source of audio corruption - by an order of magnitude.

I split the sentence between "everyone from John Eargle to Sigefried Linkwitz says (or said)" and "that". Read your entire sentence and figure out who "they" refers to in my post.

Hint, it's not the loudspeakers. Loudspeakers don't "say" anything.

Does anyone else have a problem with this?

Quote:
What we're trying to do is get speakers up to 110-115 db (peak) at the listening seat ...

Why?

You are arguing that I must damage my hearing on a daily basis.

You don't own Whispers and VanAlstine, do you?

Go back to fourth grade and find out what "..." means.

This ain't rocket science, guy.

It looks like this branch may be winding down. You are not interested in how horns work which would be one basis for discussion. You cannot provide a set of information like the K&H specification to support your view which would provide another. You also do not seem to have understood the content of the link that you do provide which is not a balanced view and is advocating increasing the area of the driver beyond what is required in order to reduce the distortion but the negatives associated with this are not discussed.

You would seem to have decided to adopt a belief that horn loudspeakers are magical which is fair enough since this is what audiophiles do. Unfortunately this is not a belief that the audiophile industry is currently geared up to support in the mainstream. Perhaps it will in the future but you do not seem to be making much headway convincing the audiophiles here.