http://www.avguide.com/blog/inside-the-yg-acoustics-factory?src=Playback

Our Az AV Club had a demo of these speakers earlier this week.

Curious what you folks thnk?

Interesting speakers from a design standpoint.

Hated it. IMHO, Ugly, Strident, Clinical (albeit based on an hour or so of listening only)

Heard them at CES. Wasn't amazed or anything, to be honest, I can't remember anything memorable, but I don't recalling them being horrible.

Do they change anything about the scan speak drivers they use?

Do they change anything about the scan speak drivers they use?
It says they do in the article.

"All the drivers are sourced from ScanSpeak and are either custom-made for YG or re-built by YG to its specifications. The tweeter in the Anat, for example, uses a diaphragm from Germany and ScanSpeak’s motor structure, with final assembly performed in YG’s factory."

Heard them in HK and Japan few times myself. Never quite got the appeal. Basically a Scan Speak 2-way kit with a sub. Nice cabinets an all but for the money, I would have expected much more.

Reviewed in the current issue of Stereophile.

I have a theater set up with them. They are the most musical speaker that I've ever heard.

I have the ANAT II Ref Professional for the fronts (which is all that I typically use for 2 channel listening), an ANAT II main module for the center, and Kipod Studios for the rears.

They really have to be heard to be believed, I've never heard anything that compares, and while I've not heard every speaker, their claims of the "best" works for me.

Kevin

Reviewed in the current issue of Stereophile.


http://www.bm.co.yu/YG%20Stereophile%20Review.pdf

I have to admit, that measures pretty amazingly well, pretty impressive. Maybe I wasn't all that amazed because I heard them in a small, crappy hotel room with music I don't know. OTOH, the vertical response kinds shows what I don't like about D'Appolitos and I didn't notice any special resolution capability from them like I enjoy, probably due to the relatively lossy driver complement.

.....or unbelievably high amount of distortion!

...measures pretty amazingly well?? As we told you before, there is more to loudspeaker then flat FR. Have a look at http://www.soundstagenetwork.com/measurements/speakers/yg_anat_ref_main_module/. Look at the unbelievably high THD these speakers possess. No wonder they always show less than ideal, no matter where I heard them.

Holy crap! :eek:

Steve,

Thanks for posting the article, I enjoyed reading it. I agree with Mr. Phillips' findings. The speakers are very musical, detailed and especially that every little change that you make can be heard. I compared to some of the speakers he mentioned and ultimately went with the YG's because of their ability to make you believe that the musicians are in the room with you.

Here is a picture of my system:

http://forum.audiogon.com/cgi-bin/fr.pl?vaslt&1214260813

Kevin

Heard them in HK and Japan few times myself. Never quite got the appeal. Basically a Scan Speak 2-way kit with a sub. Nice cabinets an all but for the money, I would have expected much more.

That's what I thought as well, I will give them a thumbs up for a cabinets but that is a big price to pay.

...measures pretty amazingly well?? As we told you before, there is more to loudspeaker then flat FR. Have a look at http://www.soundstagenetwork.com/mea...f_main_module/. Look at the unbelievably high THD these speakers possess. No wonder they always show less than ideal, no matter where I heard them.

Funny you should bring this up since a lot of high end equipment has a much higher THD and for some reason people like this sound. I guess it reminds them of the LP days.

Look at the unbelievably high THD these speakers possess. No wonder they always show less than ideal, no matter where I heard them.

there is a current discussion on diy about the audibility of distortion....

http://www.diyaudio.com/forums/showthread.php?s=&threadid=139046

it seems the jury is still out

Hi


Extremely interesting discussion on the DIY forums..Thaks for the link. I will bow out before I push this thread of topic. I have not heard the YG Acoustics so can't c. Reliance on graphs and single metrics is fraught with danger... I have had a look at the work of Dr Geddes and indeed his finding while unorthodox are quite interesting.
Back to the YG... I have NO experience My post was about the relative level of THD and how it related to our perception...

there is a current discussion on diy about the audibility of distortion....

http://www.diyaudio.com/forums/showthread.php?s=&threadid=139046

it seems the jury is still out

Debatable on levels of 0.05-1%. Not 30%. Imagine this was an airplane. Would you go on it?

Debatable on levels of 0.05-1%. Not 30%. Imagine this was an airplane. Would you go on it?

I presume you are asking me? Also presume the figure of 30% is from the soundstage graph? (I don't exactly know how to read the distortion figures on the graph, a quick explanation would be appreciated. At a stab tho, I'd say it relates to how many db down it is?)

If many audio products were airplanes I'd be catching trains, yeah.

I simply do not know enough to comment in this case, sorry. I simply posted the link cause it is an interesting discussion, and I personally am never too surprised when audiophile truisms don't stack up under investigation. Whether this example falls into that category I know not.

Re 'frequency response' not being the only arbiter, perhaps so. But, to me at least, it is a pretty good starting point. That seems to be strongly hinted at during blind comparisons/ratings of speakers.

http://www.aes.org/tmpFiles/elib/20090301/5270.pdf

Other factors can come into it, for example I too have 'gone off' D'Appolito alignments, so basic topologies etc need to be factored in.

But for armchair discussion good measurements are as good a place as any to come to a shortlist.

If I were a speaker builder, it would be in my own best interests to get as good a response as possible.

The data does seem to back it up.

My pleasure FrantzM.

Hwoo

I don't know if you have read Dr Geddes research.. His point is that there a conditions under which high level of THD are not audible.. I can poin tin my expereince to 2 amps one I have owned another I have heard in 3 occasions:
The Jadis Defy 7 (I owned it)
The Lamm Triode. The first time a while back in NY, I actually posted about it in this forum before even OB has heard of thesse... and the last time I heard it in OB's system. I have posited this a few times: We may not be measuring the right parameters... If you were to hear the Lamm You will know it sounds right.. It sounds scary right.. It plays music with a sense of realism that is rarely experienced.. 18 ridiculous watts and THD of 1% or more but once you listen the graphs seem to just disappear..
I will force myself not to post in the YG Acoustics thread.. We can take it elsewhere in this forum .. this is about the YG which I have not heard ... nor have any acquaintance...

Nothing says that one cannot like a system that gives a high level of distortion. But that is not the same as high fidelity system that the objective is to hear the source with so little distortion as possible.

I presume you are asking me? Also presume the figure of 30% is from the soundstage graph? (I don't exactly know how to read the distortion figures on the graph, a quick explanation would be appreciated. At a stab tho, I'd say it relates to how many db down it is?)

20dB down is 10% as I understand it, 40dB down is 1%, and so on. Distortion in low frequencies is very difficult to hear and is generally not noticed, but distortion over 1000Hz is increasingly annoying and obvious.

Nothing says that one cannot like a system that gives a high level of distortion. But that is not the same as high fidelity system that the objective is to hear the source with so little distortion as possible.

NIN

have you read his study. These papers were presented to the AES and there hasn't yet been any refutation. His points weren't about preferences but about perception... based on HIS studies high level of Harmonic distortion weren't as perceptible as previously thought...

Please read on

http://www.gedlee.com/distortion_perception.htm

This is OT but we can always open another thread in the DIY, Speakers and Subwoofer Forum which by the way Dr Geddes frequents... :).. What it means is that THD as we tend to measure it may not be what we would like to think it is, same as flat FR... which by the way his studies (and others) suggest as not been that important.. in passing nor is wide uniform dispersion...

Yes, but doesn't Dr Geddes like high distortion horn speakers? ;)

I mean, how do you scientifically isolate THD as a single factor when comparing loudspeakers? As in, how do you find two speakers that are identical other than distortion? Not possible. You could dump 10 or 20% distortion electrically into the same speaker and see if it's audible (and I'm sure it is).

And others have demonstrated that wide, smooth dispersion is important, or at least smooth dispersion is.

I do think that cone resonance is more noticeable than THD though. But that is seemingly hidden by most smoothed graphs.

IOW, I think Geddes is one of those people that doesn't seem to be connected to reality and has his own personal agenda that he doggedly pursues, at least from what I've seen from him.

Simply Prove him wrong else it is conjecture .. He presented his studies at the AES and they have not yet proven him wrong so...

It would be hard to prove right or wrong because you can't build a speaker that is identical in every way except distortion or dispersion, etc, etc. Very difficult if you want to be scientific. Just because he presented it to AES doesn't make it the gospel. It also depends on your interpretation of the word 'important'. The meaning of the word 'is' is, etc.

has his own personal agenda that he doggedly pursues

Sounds like a certain prolific member of these forums.

Alimentall

Once again you persist... Have you read his papers? I am not sure you have... else you would not advance this

It would be hard to prove right or wrong because you can't build a speaker that is identical in every way except distortion or dispersion, etc, etc. Very difficult if you want to be scientific. Just because he presented it to AES doesn't make it the gospel. It also depends on your interpretation of the word 'important'. The meaning of the word 'is' is, etc.
__________________.. It has NOTHING to do with identical loudspeakers NOTHING.. It is a discussion on the validity of our usual distortion measurements protocols.. NOTHING to do with Identical speakers... and The AES could have contended his observations. They did not or maybe could not... You take to flippant statements when you have no argumentation.. kind of the kid at school who willsimply speak louder to cover the voice of his opponents without being able to advance anything.. I for one do not take the AES for the Gospel, yet it is quite interesting that this body of people who collectively know more about Audio than all of us here on AVS combined were not able to destroy his extremely unorthodox positions 6 years after... I suggest you read it, there is a fair amount of College level Maths in it ... then again according to you, you were very good at it so much so, you found it boring... thus it should be a piece of cake for you...

http://www.gedlee.com/downloads/Distortion_AES_I.pdf

I do not doubt for a second that you will delve into flippancy and come with a reply however irrelevant it might be.. I will leave you to your devices, this thread is about the YG Acoustics

IOW, I think Geddes is one of those people that doesn't seem to be connected to reality and has his own personal agenda that he doggedly pursues, at least from what I've seen from him.
His agenda is scientific research. The truth is sometimes a bitter pill to swallow.

the discussion is still ongoing in the thread I linked to......

this is one persons summation of one of these studies (have not read it myself)

Just as a quick followup, the Olive paper did attempt to correlate 2nd and 3rd order harmonic distortion at 90dB SPL from 20Hz-2kHz (I'm assuming that was the fundamental freq) with listener preferences (again, this was blinded). They did NOT find a correlation between mean and max % of 2nd or 3rd HD and listener preferences. My two criticisms are:
1) this was only done on a sample of 13 loudspeakers, and they were all relatively inexpensive
2) the studied may not have been powered to detect small correlations in distortion in this small sample. (i.e. there may have been a correlation, but the study was too small to find one).

I only mention that because from memory Sean Olive does frequent these parts?

Thanks John, thought the distortion readings were summat along those lines. I think frantz is right, I mean if anyone were looking for a discussion of the importance of distortion figures, they would hardly think to look within a thread on YG speakers.

From memory, the large distortion peak shown in the measurements earlier of the YG acoustic speakers was pretty closely correlated to just about the only sharp discontinuity in the FR, if so then maybe (assuming right now that there is indeed merit that distortion is not as important as usually thought) 'all we need' is accurate FR's as a pretty good basis to judge a speaker? (as very strongly suggested by the results in the toole paper also linked earlier)

When challenged, often the things that get defended most vigorously are those unexamined assumptions.

The point is that these assumptions HAVE been challenged and examined.

Maybe alimentall is right somehow and the flaw is in the procedure?? dunno, but rather than arguing against it here with those of us who have not done the studies it might be better to take those points up with the person who did do the studies.

Earle might be willing to engage in this discussion if it had it's own thread...preferably in this section (for me only, saves having to go to another section, I'm lazy)

Alimentall

Once again you persist... Have you read his papers? I am not sure you have... else you would not advance this

I've glanced through a few of them. But mostly I've just seen his posts about how incredible horn speakers are, so I just largely discount everything he says.

.. It has NOTHING to do with identical loudspeakers NOTHING.. It is a discussion on the validity of our usual distortion measurements protocols.. NOTHING to do with Identical speakers... and The AES could have contended his observations. They did not or maybe could not...

BS!!! Of course it does! How do you 'prove' whether the THD in a speaker is a problem or not? You have to isolate it completely. That's how science works. Otherwise, all you are doing is marking general observations. This reminds me of when liberal scientists determined that every form of food aside from a non-fat vegan diet, non-alcohol diet is deadly, one expensive study at a time. Then another study concluded that everything in moderation is safe or even good for you. It even reminds me of Bob Stuart's presentation where he came to the conclusion that any resolution above ~20/66 is inaudible. BUT......he forgets to mention "with current speaker technology" and/or "with these particular speakers". Who knows if things become more audible with more advanced speaker technology or speakers that are *gasp* more resolving than the Meridians most likely used?

You take to flippant statements when you have no argumentation..

I contend Geddes has no sufficient argumentation unless he has designed a speaker with variable THD.

kind of the kid at school who willsimply speak louder to cover the voice of his opponents without being able to advance anything.. I for one do not take the AES for the Gospel, yet it is quite interesting that this body of people who collectively know more about Audio than all of us here on AVS combined were not able to destroy his extremely unorthodox positions 6 years after...

With regards to speakers, I don't know how you could.

I suggest you read it, there is a fair amount of College level Maths in it ... then again according to you, you were very good at it so much so, you found it boring... thus it should be a piece of cake for you...

http://www.gedlee.com/downloads/Distortion_AES_I.pdf

I do not doubt for a second that you will delve into flippancy and come with a reply however irrelevant it might be.. I will leave you to your devices, this thread is about the YG Acoustics

Frantz, sometimes you remind me of a hysterical mother in law type. You can be so rational and composed at times and then you just lose it at others. I don't know why this is the case, but our personal conversations give me an idea.

Frantz, sometimes you remind me of a hysterical mother in law type. You can be so rational and composed at times and then you just lose it at others. I don't know why this is the case, but our personal conversations give me an idea.
I think it is largely a matter of with whom he is trying to have a rational conversation.

Just as a quick followup, the Olive paper did attempt to correlate 2nd and 3rd order harmonic distortion at 90dB SPL from 20Hz-2kHz (I'm assuming that was the fundamental freq) with listener preferences (again, this was blinded). They did NOT find a correlation between mean and max % of 2nd or 3rd HD and listener preferences. My two criticisms are:
1) this was only done on a sample of 13 loudspeakers, and they were all relatively inexpensive
2) the studied may not have been powered to detect small correlations in distortion in this small sample. (i.e. there may have been a correlation, but the study was too small to find one).[/I]

Well, also, this would be interesting *if* they could take a speaker with less than 1% distortion and simulate distortion at different frequencies and see how this correlates. There are a dozen different *general* and uncountable *specific* aspects to a speaker aside from THD.

Thanks John, thought the distortion readings were summat along those lines. I think frantz is right, I mean if anyone were looking for a discussion of the importance of distortion figures, they would hardly think to look within a thread on YG speakers.

Well, how would you know if the speaker already has 10% or higher distortion through much of its bandwidth ;)

From memory, the large distortion peak shown in the measurements earlier of the YG acoustic speakers was pretty closely correlated to just about the only sharp discontinuity in the FR, if so then maybe (assuming right now that there is indeed merit that distortion is not as important as usually thought) 'all we need' is accurate FR's as a pretty good basis to judge a speaker? (as very strongly suggested by the results in the toole paper also linked earlier)

Well, as I said, it really depends on the frequency of the distortion and even so, cone resonance is often more obvious. I've heard speakers that have essentially zero distortion with nearly perfect FR and yet they sounded harsh. I've heard other speakers that are just as flat, have marginally more distortion and yet sound far smoother and less fatiguing. So I would say that, short of extreme levels of distortion, Geddes is correct in that other things are more important. That does NOT mean that you can excuse a $107K speaker that is this rife with that level of distortion! That's all I'm saying ;)

terry J

I thank you again for this link. It is an eye-opener... I have been puzzled for a good while by the fact that some gears measured so wrong yet so right. By right I mean so good in recreating the impression of live performers in a venue... I took the example of the Lamm 2.1 which have heard enough to opine and those who have heard it too can attest, whether they are "subjectivists" or "objectivists" those 18 watts simply sound right!! I had the Jadis Defy 7 and yes it was that kind of magical too... Could it be they that THD AS MEASURED currently is not the end-it-all of measurements as Dr Geddes suggest?
To come back to the YG Acoustics, I have not heard it but would not simply dismiss it on this simple measurement, especially in light of Dr Geddes studies...

Well, as I said, it really depends on the frequency of the distortion and even so, cone resonance is often more obvious.

I see a big glitch in the FR at just over 500 hz, and the big distortion hump is just under 500 hz. The closeness makes me wonder if they are connected, and if so might make one or the other (admittedly only in this sample of one) redundant. Ie, there is enough data in one measurement to convey the needed conclusions?

terry J

I thank you again for this link. It is an eye-opener... is not the end-it-all of measurements as Dr Geddes suggest?

To come back to the YG Acoustics, I have not heard it but would not simply dismiss it on this simple measurement, especially in light of Dr Geddes studies...

I thank you again for the thanks heh heh. I think jeff fritz has mentioned somewhere that one of his references was/is a pair of YG acoustic speakers (could be totally wrong of course), prob not the same pair as linked to tho. In any case, maybe that adds weight to your suspicion regarding the relevance of that particular measure?

Just to clarify, I have no horse in this race, but find the topic interesting.

If I was correct that sean does not mind posting here, perhaps someone could ask him to give his experiences??

Alimentall

I am not surprised by your post.. I was actually expecting these... Again it is matter of speaking louder.. Do speak as loud as you please... I am certain you have convinced no one... I'll bow from this discussion.. You won.. be happy.. and when you have the time please present your well rounded arguments to the AES tha'll help them dismiss Dr Geddes papers...

P.S I can bet you have not read ONE word from the studies...So be it.. I will not waste much energy .. Most here know where you stand... I have abandoned all hope that you'll ever be intellectually honest...

Alimentall

I am not surprised by your post.. I was actually expecting these... Again it is matter of speaking louder.. Do speak as loud as you please... I am certain you have convinced no one... I'll bow from this discussion.. You won.. be happy.. and when you have the time please present your well rounded arguments to the AES tha'll help them dismiss Dr Geddes papers...

P.S I can bet you have not read ONE word from the studies...So be it.. I will not waste much energy .. Most here know where you stand... I have abandoned all hope that you'll ever be intellectually honest...

Frantz, you really need to stop being so hysterical. I generally agree with Geddes scientifically because masking is a reasonably well understood phenomenon. But how easily the distortion in a speaker will be heard could depend very well on the speaker itself, the material, the person, so it's not perfect science. Regardless, distortion measurements are a legitimate indicator of objective speaker performance, rather than a subjective one. But this isn't a big revelation, one way or another. Dispersion, distortion, FR, cabinet design, spectral decay etc are all *indicators* of performance rather than arbiters of it. To be honest, I'm more worried about the YGA's driver and driver layout choices, but the distortion measurements are startling to say the least.

As far as the intellectually honest comment, please look in the mirror before casually tossing that kind of language around.

Back to the subject. The THD on the YG are totally out of line. And that has nothing to do with its price point. This would not be acceptable to a speaker costing a $1000. From listening , it is clear that these, at moderate to high levels, do fall apart. Just put a good soprano on and run for cover. On orchestral pieces, even at moderate levels, the speakers are audibly stressed. What is interesting is that on other speakers with similar drivers, the THD are much lower.

What it means is that THD as we tend to measure it may not be what we would like to think it is, same as flat FR... which by the way his studies (and others) suggest as not been that important.. in passing nor is wide uniform dispersion...


What? That 1% THD distorting tube amp is impossible to hear?

1% has generally been the accepted threshold for whether THD is audible, AFAIK. It just comes down to whether the tube amp's distortion is minimal compared to its strengths or whether it's simply a matter of the amp being euphonic.

Tests have showned that one can hear a lot lower than that also. :)

Back to the subject. The THD on the YG are totally out of line. And that has nothing to do with its price point. This would not be acceptable to a speaker costing a $1000. From listening , it is clear that these, at moderate to high levels, do fall apart. Just put a good soprano on and run for cover. On orchestral pieces, even at moderate levels, the speakers are audibly stressed. What is interesting is that on other speakers with similar drivers, the THD are much lower.

I agree, why they would purposely design and produce a speaker that performes and measures this bad is beyond me. Then again the high end now days sells based on this for some reason.

1% has generally been the accepted threshold for whether THD is audible, AFAIK. It just comes down to whether the tube amp's distortion is minimal compared to its strengths or whether it's simply a matter of the amp being euphonic.

1% is the point of a speaker measuring poor and most speakers in the mid/high measure much less then that. If it is audible I do not know, I have never delt with equipment that measures that poor.

1% is the point of a speaker measuring poor and most speakers in the mid/high measure much less then that.
From whence does that little nugget of knowledge originate?

I certainly am not an expert on speaker design nor have I spent a lot of time looking at performance graphs, so I have no basis of comparison to the graphs on the link. However, I am a bit perplexed by this excerpt from the article:



YG Acoustics Anat Reference Main Module Loudspeakers: Measurements

PLEASE NOTE: Our standard is to provide the THD+N measurement at 90dB with a measuring distance of 2 meters (within the anechoic chamber). Since this speaker produced low distortion levels under those conditions, we have added a second measurement performed at 95dB to give an indication of performance under higher-output conditions.



The author is saying that these speakers prodoced low distortion levels, so they subjected them to a more rigourous testing protocol. It sounds (no pun intended) like a positive result to me.

Kevin

I do believe I've seen speakers that produced less distortion that didn't get a 95dB measurement. But, you know, it's like when Stereophile puts speakers that can't produce the specified 20Hz FR in Class A full range and puts speakers that actually can do 20Hz in Class A restricted LF. Sometimes, it's just done based on size or price. YGAs are big and expensive, therefore they get the 95dB test even if they probably shouldn't bother.

From whence does that little nugget of knowledge originate?

You can look at a number of speaker reviews where THD is tested, basic designs can achive far less then 1% at $100 or less. The stock scan speak drivers produce far less THD then the YG speakers.

One wonders if YG didn't so some silly things to try to achieve such a flat response.

BTW, Frantz, consider this - if measured THD isn't terribly significant even over 1000Hz (though I'd say it still is) because of masking, it may help you understand why I don't like speakers with high cone resonances (soft cones or rigid cones not crossed properly) and speakers with multiple drivers playing the same frequency ranges (line arrays, D'Appolitos, etc). My favorite speakers have low distortion, low cone resonance *and* little to no overlap of drivers (as well as flat FR).

You can look at a number of speaker reviews where THD is tested, basic designs can achive far less then 1% at $100 or less. The stock scan speak drivers produce far less THD then the YG speakers.
Okay, so if the test show you "can achieve far less than 1%" why is 1%THD the test instead of say .75%THD. I'm just trying to figure out where you came up with the claim that 1%THD is the standard.

My favorite speakers have low distortion, low cone resonance *and* little to no overlap of drivers (as well as flat FR).

Distortion comes in large variety of "flavors", so I don't think that the THD distortion figures posted in SoundStage come even remotely close to giving us enough information to determine whether that speaker will sound pleasing.

I've also heard speakers with less than perfect waterfall plots (driver resonances) that sound great and vice versa. Same goes for speakers that use 1st order crossovers vs. 4th order. It is all in the execution and the balancing of trade-offs. I really don't think you can tick off a list of conditions a speaker must meet on a piece of paper and expect it to sound good if those are met.

I've come to realize that flat FR is of relatively low importance to me. Distortion in its many forms (not just THD) is what makes or breaks it for me.

I agree and that's basically my point. I'd very much love to see more unsmoothed graphs, for instance, that would expose more of the cone resonances and it would be really cool to see some sort of TIM test developed for this. But any time you have a really lousy result in any one area, it's very likely to cause audible problems (except maybe time/phase), whether it's dispersion, FR (inaccurate or erratic FR really bothers me), THD, whatever. I had some flagship speakers that were *stunningly* low in THD, but they also had cone resonances and a D'Appolito midrange that I just couldn't enjoy, it made the speaker sound forward and even a bit harsh and unnatural. Great in other areas (dynamics, bass, accuracy, etc), but a truly great speaker will be good or great at everything. I can't deal with 5, let alone 6-figure speakers with one or more huge weaknesses. Isn't this supposedly the reason people spend that much? To get away from these weak points? Or one would think......

Distortion comes in large variety of "flavors", so I don't think that the THD distortion figures posted in SoundStage come even remotely close to giving us enough information to determine whether that speaker will sound pleasing.
For what it's worth, Stereophile doesn't even bother testing for THD, and they generally have the most comprehensive and credible set of measurements.

For what it's worth, Stereophile doesn't even bother testing for THD, and they generally have the most comprehensive and credible set of measurements.

That doesn't mean that THD doesn't matter. Stereophile measures THD for amplifiers, etc. Why not for speakers too? If THD matters for one class of device, it probably does for another...especially when typically 10x or more higher, as is the case for speakers as compared to SS amps.

That doesn't mean that THD doesn't matter. Stereophile measures THD for amplifiers, etc. Why not for speakers too?
You'll have to address that question to Stereophile.

I think that the measurements in Stereophile and Soundstage provide some tangible objective data to size up and compare the performance of speakers, but my point is that I've heard bothersome qualities in several speakers that don't correlate with the current repertoire of graphs provided by these magazines. So for me, the measurements are purely academic and have little value in helping me determine whether or not a particular speaker will sound good. Sure, I'm probably not going to bother auditioning a speaker with a 20dB dip in the midrange, but on the flipside, a speaker with +/- 3dB is certainly in the running.

Case in point: My current Wilson Benesch speaker have pretty wavy FR when compared to something like the Revels Studio2 that I owned before them. I like the WB's much more than the Revels even though they aren't as technically perfect in what is commonly considered a key part of speaker performance.

I like the WB's much more than the Revels even though they aren't as technically perfect in what is commonly considered a key part of speaker performance.

So what? That does not make them any better or worse than anything else. That is why, measurements are anything but academics. A good set of measurement, is the only way to actually tell if a speaker is at least technically sound. If you personally like it or not, is totally irrelevant to anyone else but you.

BTW, SP does not do THD measurements because they can’t. It takes a very quiet environment to do so (Anechoic chamber) and JA does not have these conditions, IMO.

So what? That does not make them any better or worse than anything else. That is why, measurements are anything but academics. A good set of measurement, is the only way to actually tell if a speaker is at least technically sound. If you personally like it or not, is totally irrelevant to anyone else but you.

BTW, SP does not do THD measurements because they can’t. It takes a very quiet environment to do so (Anechoic chamber) and JA does not have these conditions, IMO.

"A good set of measurements"

Thats the crux of the matter. What is a "good set of measurements"?

I think that there is more to the make-up of a speaker's sound than FR, dispersion, spectral decay, and THD. The measurements provided in the mags are only part of the picture. If this is the case, and the measurements found in mags cannot tell you if a speaker will sound good or bad, then IMO they are academic to the consumer.

"A good set of measurements"

Thats the crux of the matter. What is a "good set of measurements"?

I think that there is more to the make-up of a speaker's sound than FR, dispersion, spectral decay, and THD. The measurements provided in the mags are only part of the picture. If this is the case, and the measurements found in mags cannot tell you if a speaker will sound good or bad, then IMO they are academic to the consumer.

Not really. Between all these measurements you mentioned, at different output levels, plus impedance you can get a pretty good idea about the speaker viability . Of course you need to know what you are looking at. Also, a separate distortions analysis would help (2nd, 3rd, etc). A combination of what SS and SP are doing will be pretty telling. It is a pity non are doing all the require sets, IMO.

I think that there is more to the make-up of a speaker's sound than FR, dispersion, spectral decay, and THD. The measurements provided in the mags are only part of the picture. If this is the case, and the measurements found in mags cannot tell you if a speaker will sound good or bad, then IMO they are academic to the consumer.

Does "academic" mean "irrelevant" to you, or what? Plenty of consumers might not like a speaker that measures well in most/all respects, because plenty of consumers would prefer a hyped treble or hyped (non-flat) midbass or whatever, or because the consumer's room has a "sound" contribution of its own, and a well-measuring speaker might not sound good in it, or because the available measurements aren't capturing everything that is relevant about the sound of a speaker.

But to me at least, that doesn't mean the measurements are "academic." If I am going to pour alot of money into speakers, I'd like it to measure well, and looking for well-measuring speakers will help narrow down the list of audition candidates. I am not going to audition every available speaker in my price range, after all.

... or because the available measurements aren't capturing everything that is relevant about the sound of a speaker.

This is my main point. The commonly available measurements in audiophile mags don't capture everything that is important. Not to these ears anyway.

But to me at least, that doesn't mean the measurements are "academic." If I am going to pour alot of money into speakers, I'd like it to measure well, and looking for well-measuring speakers will help narrow down the list of audition candidates. I am not going to audition every available speaker in my price range, after all.

When I use the term academic, I mean interesting to study, but of limited practical use. But maybe academic isn't the best word, because I do believe that the commonly published measurements can tell us valuable information about a speaker's performance, but I've also found that there can be things just as or more important to sound quality than what is shown in the commonly published measurements. Things I hear that don't correlate with the pusblished graphs or my own measurements. (I use Room EQ Wizard and PEQ to flatten the in-room response of my system.)

So while I used to narrow down my choices by using published measurements, today I find it more effective to listen to every speaker in my price range that I have access to, and refer to the measurements just to be sure that they are doing anything too wonky. Even if a speaker doesn't measure well, rather than rule it out, I would still audition it and listen for problems in the areas that the graphs point to.

The bottom line is that I've heard too many speakers that measure well, but don't sound "right" to rely on the measurements as a selection tool.

For instance, I've owned Aerial 7B's and Revel Studio2's. Both excel in the measurements department, yet I found the Aerials sounded dull and "uninvolving" and the Revels sounded somewhat "clinical" and bright. Could these qualities have been the result of how the speaker interacted with my room, my electronics and the recordings I listen to? Certainly, but that's part of the problem. There are so many variables in the real-world use of a set of speakers that anechoic measurements published in a magazine play only a small elementary role in our understanding of how those speakers are going to sound in your home. And again, I believe there are things that contribute to the sound of speaker that are not shown in magazine measurements currently.

This is my main point. The commonly available measurements in audiophile mags don't capture everything that is important. Not to these ears anyway.

That's fine, but it doesn't mean that those things that are measureable are not potentially useful.

That's fine, but it doesn't mean that those things that are measureable are not potentially useful.

Agreed. I just used to place too much importance on the commonly available measurements.

For the record, I am a science-minded person, I just think more data and investigation is needed.

I think there is a general problem with measurements and their practical use in speaker selection. Given the increasing decline in dealers and the geographic and competitive product limitations imposed by speaker manufacturers on dealers, what is the likelihood that you will be able to audition two speakers side by side that you are interested in purchasing that have been reviewed and measured so you can compare their sound and measurements in real time? I think the likelihood of that occurring approaches zero. So while looking at measurements and comparing them while sitting in you easy chair may be interesting, they are really of very little use in the real world in determining which speakers to audition and purchase. Read the reviews, track down the speakers you're interested in, listen to them, and if you like the way they sound, buy them. Seriously, if someone likes the Magico V3s or the YG Acoustics Anats, is he really going to care one way or the other if they measure high in THD or have a .5db bump at a certain frequency? Let's face it, no two speakers are going to measure identically so you are going to have to let your ears be the final arbiter anyway. Having some intellectual problem with a measurement override what your hearing is telling you is a recipe for long term dissatisfaction.

HWoo & Syswei -

Allow me to turn the table for a moment and ask, how much practical / applied experience do you have with matching the measured performance of a speaker to their subjective performance? Have you auditioned a number of good-measuring speakers and found that they always sound good and likewise, can you reliably hear faults in poor-measuring speakers that correlate with their measured performance?

I've been trying to do this over the past 8 years and while some things line up some of the time, more often they don't. Not because measurements are inherently useless, but they are too limited in scope and they only apply if the exact conditions of the test are recreated in your home or at the dealer. Not likely.

Paul Barton would tell you that, with his work at the NRC in Canada, he can predict which speakers a group of blind listeners will prefer by measurements alone. He is also apparently big on extremely low distortion as shown by his latest offerings.

Paul Barton would tell you that, with his work at the NRC in Canada, he can predict which speakers a group of blind listeners will prefer by measurements alone. He is also apparently big on extremely low distortion as shown by his latest offerings.
So we should just forget about auditioning and buy some PSB speakers from you? Besides, Dave Wilson, Jim Thiel, etc. would tell you the same thing about their speakers. That's why they make them the way that they do.

Given the increasing decline in dealers and the geographic and competitive product limitations imposed by speaker manufacturers on dealers, what is the likelihood that you will be able to audition two speakers side by side

That's actually a good reason to consider measurements as part of the evaluation process. If my candidate speakers aren't available at a single dealer, then how can I know how much of what I hear at dealer X is the room? Measurements can help abstract from this problem, and are useful partly for that reason.

Allow me to turn the table for a moment and ask, how much practical / applied experience do you have with matching the measured performance of a speaker to their subjective performance?

None because I haven't started auditioning for my to-be-built room.

can you reliably hear faults in poor-measuring speakers that correlate with their measured performance

If I can't, that doesn't mean that I don't care. If one speaker measures much higher distortion than another but I can't tell the difference in listening at different times, perhaps an hour or more apart, at 2 different dealers, I still might be able to tell if I owned both long term, listening in my own room and possibly benefiting from training my own ears. If I don't have the luxury of long-term dealer loans (or purchasing both sets) then I would sooner buy the lower-distortion set.

syswei -

You sound like you are at the place I was about 5 years ago. I used to pour over the measurement sections of countless reviews looking for the most technically capable speakers within my budget. I'd even go to the extent of looking for measurements posted in foreign audiophile magazines.

I would cross reference the measurements looking for common traits, as well as look for commonalities between the subjective reviews.

Confident that I could use all of this information to make an informed choice, a few years ago a bought a couple of speakers without auditioning them at all (KEF Reference and Aerial Acoustics). I did so because neither were available locally for audition and I *thought* that they would outperform what was available. Both had commendably flat FR, low distortion, etc., but ultimately I didn't like either of them enough to keep them. Maybe your journey go differently, and you will find that technical perfection is just what you want, but I still recommend that you listen to everything in your price range that you can. You may find a diamond in what looks like a piece of coal.

If my candidate speakers aren't available at a single dealer, then how can I know how much of what I hear at dealer X is the room?
Unless you have measurements for the speakers AND measurements for the dealer's listening room, then you're not going to be able to tell very much from the measurements of the speakers alone. Most dealers don't provide measurements of their listening rooms. Now, if you had measurements for two pairs of speakers and you listened to them in the same spot in the same room, even if you did not have measurements for the room, you might be able to correlate the difference in sound to the difference in measurements - if you were trained to listen for such differences. Otherwise, the speaker measurements are not going to do you much good, except perhaps to tell you something about low frequency roll-off or a tipped up high-end.

I think it's just a matter of having the right measurements and knowing how to weight them. Still, Tim, I'd have DEQXed the Studio2s. I almost did, but I needed the money more than I needed to spend even more of it, given the economy.

I think it's just a matter of having the right measurements and knowing how to weight them. Still, Tim, I'd have DEQXed the Studio2s. I almost did, but I needed the money more than I needed to spend even more of it, given the economy.

What would be the theoretical *audible* benefits of DEQXing the Studio2's?

I'm not confident about this belief, but I'm not so sure that I like the sound of Beryllium. I have listened to three sets of speakers with tweeters made from this material (Focal, Revel, Usher) and they all sound detailed, but in an etched, unnatural sounding way. Revel was the least offensive, but the quality was still there. Your thoughts on being able to hear driver materials?

What would be the theoretical *audible* benefits of DEQXing the Studio2's?

I think you could smooth them out, improve the deep bass and make them more effortless, as well as improving the soundstaging.

I'm not confident about this belief, but I'm not so sure that I like the sound of Beryllium. I have listened to three sets of speakers with tweeters made from this material (Focal, Revel, Usher) and they all sound detailed, but in an etched, unnatural sounding way. Revel was the least offensive, but the quality was still there. Your thoughts on being able to hear driver materials?

I think I agree with you on Be. I haven't heard the Ushers, but I feel the same way, to the point of preferring good old cheap aluminum over Be. Not sure why and not sure if the spectral decay is a Be problem or what. I have to admit being a bit disappointed in that, the more I listened to them.

Hi


All I can say is that common measurements correlate VERY poorly to speaker performances... YMMV

Okay, so if the test show you "can achieve far less than 1%" why is 1%THD the test instead of say .75%THD. I'm just trying to figure out where you came up with the claim that 1%THD is the standard.

1% is an easy number to go by and also indicates clipping for amps.

So what? That does not make them any better or worse than anything else. That is why, measurements are anything but academics. A good set of measurement, is the only way to actually tell if a speaker is at least technically sound. If you personally like it or not, is totally irrelevant to anyone else but you.

I agree, measurements prove everything but this high end does not like to use any measurements because they are usually poor and do not reflect the cost.

some of the 'confusion' I think is that yes, FR does not tell the whole story, but yes fr does tell the whole story.

Two 'identical' fr graphs from two very different speakers (on axis I guess, they would rapidly diverge off axis?). They will sound different of course. They have different driver choices, layouts, box construction and shapes etc etc etc.

BUT, within a given set of constraints ( drivers, layout, yada yada yada) THEN the FR becomes a measure of 'competence' in putting it all together.

In my understanding, referring back to the works of toole etc, that is when the 'ability to predict' preferences comes into play, and that seems to be flat accurate response with smooth off axis performance. that is where flat measured response as published becomes important. (same drivers, same topology, then it is a better comparison between coloured and uncoloured sound)

And also explains why preferences can vary, ' I do not like Be tweeters, or I don't like D'Appolito or a myriad of other personal preferences, even to the stage of 'preferring non-flat (my choice of speaker given all the above variables) over non-flat'. But I think the erronous conclusion to draw is 'hence FR is no real indicator'.

Other side of the same coin? "I prefer speaker xyz even tho it has a poor FR response'. Given my hypothesis above, then it might be said that 'then you would enjoy it even better if the response was made more accurate' (which is what the blinded studies seem to show).

And like all things from human studies, there is always a bell curve. No matter what, there will always be someone who prefers one of the extreme ends of the spectrum.

I think you could smooth them out, improve the deep bass and make them more effortless, as well as improving the soundstaging.

While those are admirable improvements, there would have to be a change to the naturalness of tone, and possibly the harmonic structure for me to change my mind about the Revels. Which in part comes back to the sound of the Be tweet-- I think.

I think I agree with you on Be. I haven't heard the Ushers, but I feel the same way, to the point of preferring good old cheap aluminum over Be. Not sure why and not sure if the spectral decay is a Be problem or what. I have to admit being a bit disappointed in that, the more I listened to them.

Intersting isn't it? I don't know *why* either, and I don't see any problems in the spectral decay graph in the Stereophile measurements, yet there does seem to be a sound signature to Be that doesn't sound natural to me. There is a little low level hash in the SP measurment, but I've seen the same sort of hash in the measurment of a softdome that I've listened to and they certainly didn't sound like like the Revel in the HF.

I thought that maybe I just don't like the sound of ultra stiff drivers, but I listened to the Avalon Acoustic Indra speakers the other day at a dealer, which use Accuton Ceramic tweeter and midrange, and they sounded fantastic. Gobs of detail and resoltion, but not etched or tizzy sounding at all. I've also heard aluminum tweets that I like more than Be.

some of the 'confusion' I think is that yes, FR does not tell the whole story, but yes fr does tell the whole story.

Sure, if you extend FR to 3D and in the time dimension, you could get the whole picture, really.

And also explains why preferences can vary, ' I do not like Be tweeters, or I don't like D'Appolito or a myriad of other personal preferences, even to the stage of 'preferring non-flat (my choice of speaker given all the above variables) over non-flat'. But I think the erronous conclusion to draw is 'hence FR is no real indicator'.

I agree. In fact, inaccuracy is a way of covering for other weaknesses. When I hear a speaker, I can generally picture the FR in my head, often with a spooky accuracy that is revealed when the speakers are measured by a magazine (and I don't consider myself to be 'golden-eared', just persistently self-trained). So, inaccurate speakers bug me. When you think of it, +/-3dB is hugely inaccurate, so that kind of FR aberration shouldn't be anything less than obvious. And a lot of 'high-end' speakers can't even fit within that.

Other side of the same coin? "I prefer speaker xyz even tho it has a poor FR response'. Given my hypothesis above, then it might be said that 'then you would enjoy it even better if the response was made more accurate' (which is what the blinded studies seem to show).

Definitely agree. Even if you find the FR impressive, accuracy grows on you. That's what I like about DEQX, you can take an inaccurate speaker with other qualities and make it accurate, even without tri-amping it. And tweak from there.

All I can say is that common measurements correlate VERY poorly to speaker performances... YMMV

I disagree. I think the more astute a listener you are, the more they correlate. I often have to point out speaker flaws to 'audiophiles' who simply don't listen as carefully as they think. The average person listens to the overall sound and makes a judgement. This is why many people go on and on about how they like or dislike a particular speaker, but can't really explain why in any kind of sensible terms. I listen to a dozen different aspects, one at a time, then mentally step back and put the whole picture together. I think that method of listening gives you a much better idea of whether a speaker is good or not and with a much higher correlation to how it measures.

Alimentall -

I find that variations in FR, unless drastic, have little bearing on how well I will like the sound of a speaker.

I used to put a lot of value into flat FR, mainly because it made sense that you wouldn’t want a speaker boosting or cutting the original signal, and because of the Harman whitepapers on the appeal of flat FR. But then I did some homebrewed experimentation that changed my mind on the subject.

For a couple of days, I used a graphic EQ to monkey with the FR of my system, boosting and cutting by 3dB across the range. Surprising to me, none of the EQ adjustments made the sound offensive in any way, even a boost to the delicate presence region by +3dB (leaving the rest flat). The changes to FR that I made through EQ usually just offered a different perspective on the recording, and whether I liked the change varied by recording. I should also mention that master volume also affected my preferences for flat or non-flat.

It was also interesting to discover how quickly my ear/brain adapted to the new sound of each EQ filter change and how sometimes going back to flat FR sounded less pleasing.

While I’m using a manual PEQ within my prepro to flatten the in-room response of my speakers (measured using Room EQ Wizard), I chose not to go for ruler flat FR after listening to it this way. Leaving some of the speaker/room character in the response ended up sounding better balanced and natural to me.

Ah, see, I think of 6dB fluctuations in a speaker to be annoying with lesser ones being less annoying. And simple tone controls, even good ones just don't make things any better for me, even in small amounts (because it usually doesn't fix the problem with sufficient precision).

Also, I think you'd find naturalness and harmonic structure come alone for the ride with the DEQX.

BTW, loved the thread on the acoustic cloud!

I agree with your point about 'personal taste'. My view on it is slightly more complicated, and that is that 'salting to taste' with eq (for example) as you describe is best done from a clean slate, ie a smooth accurate native response.

It is fine to give a gentle boost in the 'presence' region (which is what btw??), but preferably without having that smooth boost 'corrupted' by sharp discontinuities and resonances of the basic speaker coming thru.

Of course, even better is the ability to salt to taste according to the recording! What works on disc one may be totally inappropriate on disc two.

So we've gone from:
1% is the point of a speaker measuring poor and most speakers in the mid/high measure much less then that. If it is audible I do not know, I have never delt with equipment that measures that poor.
to
1% is an easy number to go by and also indicates clipping for amps.
Admit it. You're just making up 1% as being the standard for a speaker measuring poorly, and you don't even know if it's audible. Sheesh. :rolleyes:

Ah, see, I think of 6dB fluctuations in a speaker to be annoying with lesser ones being less annoying. And simple tone controls, even good ones just don't make things any better for me, even in small amounts (because it usually doesn't fix the problem with sufficient precision).

Also, I think you'd find naturalness and harmonic structure come alone for the ride with the DEQX.

I agree that if a 3dB peak is followed by a 3dB dip, you could run into problems with the 6dB deviation from flat (though this is probably less of an issue if this happens off-axis IME). So a general +/- 3dB spec doesn't tell me enough, it depends on where the variations lie.

BTW, loved the thread on the acoustic cloud!

I agree with your point about 'personal taste'. My view on it is slightly more complicated, and that is that 'salting to taste' with eq (for example) as you describe is best done from a clean slate, ie a smooth accurate native response.

It is fine to give a gentle boost in the 'presence' region (which is what btw??), but preferably without having that smooth boost 'corrupted' by sharp discontinuities and resonances of the basic speaker coming thru.

Of course, even better is the ability to salt to taste according to the recording! What works on disc one may be totally inappropriate on disc two.

Yeah, the acoustic cloud worked out well. Gives me improved room sound without intruding on the look of the room. Still I would rather have a dedicated sound room one day.

I agree that flat FR is a good goal to aim for when performing room EQ, but I also found that a textbook approach didn't get me all the way there. My goal certainly was not a colored sound, but for whatever reason, leaving some of the irregularities in the measured room response sounded more "accurate". I think it has to do with our perception of direct vs reflected sound. In effect the microphone sees problems that my ear does not and vice versa.

BTW - the presence region is 2-4KHz-- the most sensitive range of human hearing.

Seriously, if someone likes the Magico V3s or the YG Acoustics Anats, is he really going to care one way or the other if they measure high in THD or have a .5db bump at a certain frequency?


I really don't understand why anyone wants to pay for high THD. :confused:
I rather have speakers that don't distort like easily.

Admit it. You're just making up 1% as being the standard for a speaker measuring poorly, and you don't even know if it's audible. Sheesh.

No, 1% is used for amps and most of the speaker reviews have a THD under 1%. The problem with YG is that they take a stock driver and some how make it perform worse and ramp up the THD.

http://www.stereophile.com/floorloudspeakers/yg_acoustics_anat_reference_ii_professional_loudspeaker/

Steve you started this thread and I think you would be interested in this from YG.

http://www.stereophile.com/floorloudspeakers/yg_acoustics_anat_reference_ii_professional_loudspeaker/

One last thing: Geva claims that his loudspeakers are not voiced, or otherwise created with any "human bias." Seeing my shocked expression, he hastened to add that he verifies by human experience, but establishes the speaker's performance solely through measurements. Does that mean that if he produced a speaker that measured right but sounded wrong, he wouldn't change it? "I would question the measurement in that case and refine my methodology."

I just started this thread as our club had a demo of these speakers. The fact that I didn't give my own review of what I heard, take my silence for whatever its worth.
When it comes to audio, I personally think that both objective measurements and subjective hearing should be involved to fine tune the product. The fact that YG brags that its all objective with no subjective, which was discussed at our club meeting,
is a legit concern for many of us. Though I would think Doug would extoll objective vs any subjective any day of the week.

I just started this thread as our club had a demo of these speakers. The fact that I didn't give my own review of what I heard, take my silence for whatever its worth.
When it comes to audio, I personally think that both objective measurements and subjective hearing should be involved to fine tune the product. The fact that YG brags that its all objective with no subjective, which was discussed at our club meeting,
is a legit concern for many of us. Though I would think Doug would extoll objective vs any subjective any day of the week.I kind of doubt that YG pays no attention to the subjective characteristics. WP, who reviewed them in Stereophile and was invited by YG for the factory tour, seems like a very strange choice for an invitee. This is a man who seems quite comfortable with the purely subjective crossing happily over to the delusional.
Any setup tweaks you might normally perform will pay huge dividends. Ayre Myrtle wood blocks under the Ayre KX-R preamplifier? Huge. Cable dressing? Ditto.
And cables themselves? Puh-leeze—let's not even go there. One afternoon, VTL's Luke Manley, my friend Jeff Wong, and I were playing silly audiophile tricks. We substituted Shunyata Research Python Helix Alpha AC cables for the stock pair feeding the Anat Professional's powered subwoofers. I expected to hear little difference, possibly none at all. Au contraire—the bass was audibly better integrated with the rest of the audioband. And another of my certainties bit the dust.

WP would have been much more comfortable and ladled on more effusive hyperbole had YG played up to WP's idiosyncracies (dare I say mental illnesses?) by going on about how he'd chosen this configuration, that particular speaker surround material composition, such and such capacitors and inductors, based on extensive listening evaluations where he evaluated such things as tonality, dimensionality, delicateness of presentation, etc. In essence, WP's bringing up of matters such as those quoted above was his way of saying if he'd had a hand in the design and manufacture of the YG speaker, it would even be a far better speaker that would respond positively with anything you'd throw at it.

It seems Steve that WP and the members of your AZ audio club would've been more comfortable and receptive of the YG had it taken an approach more like Dave Wilson as in for instance when he talks about how he's got a guy whose job it is to precisely twist cable just the right amount, or his visits to a special concert hall. IOW, YG would probably financially benefit from more talk about subjective matters so as to not alienate that segment of the marketplace.

I can respect YG Acoustics goal of creating the "worlds best measuring speaker", but something doesn't add up. I was just perusing the Stereophile website, and came across the review of the Anat Reference 2.

One thing that has struck me as odd about this speaker is just how tall it appears to be with all three of the optional modules in place, and how high up the mid / tweeter module is when fully assembled into the "Reference" model.

I thought that maybe the speaker just looked larger than it really is, so I went to the YG Acoustics web site and the total height is 69" with all three modules.

Subtract 10" to get to the center of the tweeter and you are looking at 59", which according to Stereophile is the prime listening height and the only height in which there aren't significant suck-outs in the upper midrange FR.

Considering that the average seated listening height is about 40", that puts the tweeter at least 19" or close to 2 feet above the listening position.

Am I missing something here or doesn't it seem obviously wrong to have the mids/tweets that high up? Especially coming from a company that is so obviously concerned with the objective / measurable performance of their speakers?

It seems Steve that WP and the members of your AZ audio club would've been more comfortable and receptive of the YG had it taken an approach more like Dave Wilson as in for instance when he talks about how he's got a guy whose job it is to precisely twist cable just the right amount, or his visits to a special concert hall. IOW, YG would probably financially benefit from more talk about subjective matters so as to not alienate that segment of the marketplace.

Club members overall seemed quite receptive to the demo. Let me know when you want to fly to Az and attend a club demo so you can accurately talk to and observe what club members think and say rather than assume what you have only your mind as a basis for.

I can respect YG Acoustics goal of creating the "worlds best measuring speaker", but something doesn't add up. I was just perusing the Stereophile website, and came across the review of the Anat Reference 2.

One thing that has struck me as odd about this speaker is just how tall it appears to be with all three of the optional modules in place, and how high up the mid / tweeter module is when fully assembled into the "Reference" model.

I thought that maybe the speaker just looked larger than it really is, so I went to the YG Acoustics web site and the total height is 69" with all three modules.

Subtract 10" to get to the center of the tweeter and you are looking at 59", which according to Stereophile is the prime listening height and the only height in which there aren't significant suck-outs in the upper midrange FR.

Considering that the average seated listening height is about 40", that puts the tweeter at least 19" or close to 2 feet above the listening position.

Am I missing something here or doesn't it seem obviously wrong to have the mids/tweets that high up? Especially coming from a company that is so obviously concerned with the objective / measurable performance of their speakers?


Our demo has the woofer module with the midrange-tweeter module on top. But the Sales Mgr mentioned that they also configure it with two mid-range tweeter modules with the speaker somewhat towed in towards the listener.
I do not disagree with the concern you raise about that sort of configuration.
Makes one pause for thought!

Only if you can score good Kosher food Steve :D

Only if you can score good Kosher food Steve :D

For you, my wife will cook homemeade matza ball soup and the best cabbage rolls you've ever had!!!!

Steve

What was YOUR impression of this speaker?

Steve

What was YOUR impression of this speaker?



Mixed. But they were admittedly new out of the box, just starting to break in. And a club member brought his Class D amps. So I don't know its a fair demo of how they sound properly matched and broken in.

But they were admittedly new out of the box, just starting to break in. And a club member brought his Class D amps. So I don't know its a fair demo of how they sound properly matched and broken in.

Still with the break in?

Steve,

It sounds like you actually listened to the Kipods (only one mid-range)?

I found the speakers to be very sensitive to positioning, including tilt which can be done with adjustable spikes. They are very revealing and every little change to the system can be heard (good and bad).


Kevin

Still with the break in?

Speakers have physical drivers which move. HELLO! Even most folks who disagree on component burn-in sounding better agree that speakers require burn-in.

Woops! Doug knows more than the designers at Aerial, Wilson, Revel, etc.
What else is knew.

Doug, I coin you "The Mouse That Roared"!!!

Steve, most true engineers will tell you that there is somewhere between zero break in and maybe 24 hours (to be generous) of break in. 90-95% of all break in is you breaking in to the speaker. But that's what companies count on happening, otherwise, you'd immediately return most speakers. So, I would agree with DW on this one.

Steve, most true engineers will tell you that there is somewhere between zero break in and maybe 24 hours (to be generous) of break in. 90-95% of all break in is you breaking in to the speaker. But that's what companies count on happening, otherwise, you'd immediately return most speakers. So, I would agree with DW on this one.

No, you agree with me on this one. I told you that the speakers were brand new out of box and had not been played at all prior to our demo. HELLO?
Is anybody home???:D

Hi

On the subject of loudspeaker break-in I have an open mind. I had back in the days a MG 3.6 and with the same amps, it flapped on hard bass notes... Literally a big FLAPPP sound... it was new, a fellow owner recommended me to play some heavy bass music on it continuously.. I did for a few days, when out working, not really listening to the speaker...The flap sound disappeared... same amps, same room ,same everything else... never to return.. I do not believe in cable breakin but on speakers.. to a certain extent, yes.

Steve, that doesn't mean you will hear a change because many speaker company's drivers and crossovers are burned in before going into the cabinet for QC.

Hi

On the subject of loudspeaker break-in I have an open mind. I had back in the days a MG 3.6 and with the same amps, it flapped on hard bass notes... Literally a big FLAPPP sound... it was new, a fellow owner recommended me to play some heavy bass music on it continuously.. I did for a few days, when out working, not really listening to the speaker...The flap sound disappeared... same amps, same room ,same everything else... never to return.. I do not believe in cable breakin but on speakers.. to a certain extent, yes.

Something similar to that also happened to me. It was with a center channel speaker. It was a driver design with a stationary bullet in the middle. The driver cone was ever so slightly tapping against the bullet in the center of it causing a faint but distracting clanking noise.

After a week of trying to play bass passages through it, the driver must've literally straightened itself out a little bit because the sound disappeared, never to return.

thread cleanup

See an inappropriate or problematic post? Please do not respond to it: use the 'Report Post' button

Thanks

I noticed the Scan Speak components in the YGs at CES a couple of years ago. I thought it might be a fun DIY project to build something similar. Brady our resident speaker builder and I came up with a plan. Not metal,but a well built MDF cabinet. The waveguide is more or less a guess and the drivers are factory standard Scan Speaks. I am using an electronic crossover set at 1750hz with 24db per octave roll off. Higher frequencies definitely are rolling off as a result of the probably not exact duplication of the waveguide. I did use my Audio Control RTA to measure and Equalize the final product. Added a woofer for a full "Triamped" version.
I am very happy with the results. It seemed the closer I got to "flat" response on the RTA the smoother they got. It produces a great image with very lifelike reproduction. An old Sheffield recording using 2 mikes sounds like the musicians are in the room. I did compensate for the rolloff with the eq somewhat. Pictures attached

Below we can find a measurement conducted by Mr. Ishii, an independent audio engineer, who was hired in 2005 by Stereo Sound magazine in Japan, to conduct a comparative measurement test between market-leading loudspeakers in the range above $20k.
The measurements were held at Fostex labs of Japan, who have one of the best measurement labs in the audio industry. The measurements were published in Stereo Sound issue #156 (fall of 2005). The model measured was the previous-generation Anat Reference Main Module, just like the one tested by "SoundStage".

We can see in this graph, that the breakdown of the distortion products to 2nd harmonic and 3rd harmonic. The engineer who conducted the measurement and commented on it has followed the principle that 2nd harmonic distortion is basically inaudible below 5%, whereas 3rd harmonic is definitely audible as low as 1%. The written evaluation for each speaker was given based on this criterion. The dotted line is the 2nd harmonic, and the solid line below it is the 3rd harmonic. As you can see, while the Anat Reference Main Module’s 2nd harmonic reaches -36 dB at 7.5 kHz (the difference from SoundStage’s measurement is due to the fact that theirs was conducted at 95 dB, which is a higher volume than used here), i.e. 1.5%, the 3rd harmonic is very low throughout the graph, and never comes even close to 1%.
We can see that a speaker is never required to output 95 dB at 7 kHz, since there is very little energy in this high frequency range (mostly harmonics). A speaker would typically only play 95 dB at 7 kHz when asked to play music with an overall volume of over 125 dB, which no domestic speaker can handle anyway. Therefore, our discussion here is mostly theoretical, and does not bear any realistic implications for a home audio system. Nonetheless, it is an interesting discussion.

Analysis of Measurements of the previous-generation Anat Reference Main Module’s by Mr. Shinichiro Ishii: “The YG Acoustics Anat Reference Main Module is a high-tech speaker system with two 15cm woofers and a ring-radiator type soft-dome tweeter. It is enclosed in a special-alloy fully-sealed enclosure. A model with a subwoofer is also available, but the one measured at this time is the Main Module only. Bass bandwidth in the anechoic chamber was down to 50Hz. From that point the response was perfectly flat… In the ultra-high frequency range, too, the response was smooth all the way to 50 kHz. The usable bandwidth is up to 100 kHz! The horizontal dispersion is flat, while the vertical axis is extremely well-controlled. This ensures a result listeners would find very pleasant. I had an impression that this is a speaker with excellent overall system design. As for freedom-from-distortion, it is superb with an extremely low third harmonic. What requires attention is a considerable impedance peak near 2 kHz, as well as wide impedance-vs-frequency fluctuation with a minimum of 2.8 Ω and maximum of 12.7 Ω. While this does not have any negative influence on quality, it does mandate matching this speaker with an amplifier that has a high damping factor"...

Just my two cents.

Who is interpreting this? You, tomersin or YG ? Looking at the SS 156 issue with these comparative measurement tests, I can tell you that it is not Mr. Ishii interpretation. Regardless, among all speakers tested, YG distortion level, both 2nd and 3rd were on the high side. And, with all due respect to Mr. Ishii, measuring distortion levels at only 87db SPL is not very telling. It is hardly 3 volts of input on the YG…

This is a really facinating thread! I've enjoyed all the discourse on THD and measurements in general. But back to the Anats..........

I cut my teeth, so to speak, on the Dunlavy's of old. So it was with great curiosity that I started noticing YG's adverts and all their fluff on measuring. I finally got a chance to hear them at CES a few months ago and was totally disappointed. There is no way I would buy these speakers, not even at 1% of the cost. They happened to be playing them with a track that I was very familiar with and with electronics that I know well so I considered it, for me, a fairly useful experience. Useful in that I wont ever bother listening to them again. These are definitely in the "ear bleeding" catagory of high end.

Anyway I read something in YG's literature that completely baffled me until I came across the THD measurements talked about here in this thread.
YG makes those cabinets as rigid as anything out there but there isnt a shred of dampening material inside! They are essentially a very rigid and sealed echo chamber! The intensity is so great that it actually INCREASES the impedance spike at the woofers resonant frequency. Why on earth they do this is beyond me. Well, actually, they explain their reasons but it makes zero logical sense to me.
Its no wonder the THD goes crazy so fast as the volume increases. Imagine every bit of the SPL that you hear out in the room trapped inside that little box and smacking the heck out of the back side of the driver. Not to mention the limitations it puts on the user in choosing the right amplifier.

If this helps them achieve a slightly better FR it comes at an unacceptable price. (to me)

I just switched to a Marchand XM9 electronic crossover from the Rane I had been experimented with. The drivers in my DIY speakers are very similar if not exact duplicates of the Anat's Scan Speak drivers.Even though YG notes Scan Speak makes a custom version for them.
The Rane had been picked for its price and adjustable crossover point, not for it's sound quality. The Marchand took the sound to a new level.Much more detail. I ordered it with a fixed 1750hz 24db per octave crossover.
With some small adjustments to the crossover levels I am getting sound I know anyone would like. No giant coils and capacitors in this set up!
Then I added "Audyssey" Equalization to the two channel set up. Wow! NAD has a target curve provided by Paul Barton of PSB.Very Nice!
This may be the future of loudspeaker design. Simple designs with low distortion components and use high performance DSP to smooth the results. Even corrects for time alignment!
Audyssey is just the beginning of DSP in loudspeaker design. In 10 years I bet more and more products using DSP will dominate.

DSP crossovers are a cool new thing but I doubt they will totally take over. They can account for timing differences but not phase. Electronic crossovers using steep slopes still introduce phase distortions and various other issues.

I find it really hard to understand why this speaker cost so much.
100K for this?
http://www.soundstagenetwork.com/measurements/speakers/yg_anat_ref_main_module/thd_95db.gif

When a supersmall ( 11.7"W x 9.8"H x 9"D) speaker that cost around 1/50 of it can do this:

http://www.soundstagenetwork.com/measurements/speakers/guru_proaudio_qm10/thd_95db.gif

And it probably have even lower distortion because of the placement in the room.

DSP crossovers are a cool new thing but I doubt they will totally take over. They can account for timing differences but not phase. Electronic crossovers using steep slopes still introduce phase distortions and various other issues.

Actually, they can account for phase, DEQX does exactly that. Not that it's that important. Probably dead last in the scheme of things. But why not if you can do it? That's the beauty of DSP, the ability to solve many problems with math, leaving a fewer number of problems to be dealt with acoustically.

I just switched to a Marchand XM9 electronic crossover from the Rane I had been experimented with. The drivers in my DIY speakers are very similar if not exact duplicates of the Anat's Scan Speak drivers.Even though YG notes Scan Speak makes a custom version for them.
The Rane had been picked for its price and adjustable crossover point, not for it's sound quality. The Marchand took the sound to a new level.Much more detail. I ordered it with a fixed 1750hz 24db per octave crossover.
With some small adjustments to the crossover levels I am getting sound I know anyone would like. No giant coils and capacitors in this set up!
Then I added "Audyssey" Equalization to the two channel set up. Wow! NAD has a target curve provided by Paul Barton of PSB.Very Nice!
This may be the future of loudspeaker design. Simple designs with low distortion components and use high performance DSP to smooth the results. Even corrects for time alignment!
Audyssey is just the beginning of DSP in loudspeaker design. In 10 years I bet more and more products using DSP will dominate.

That's the nice thing. Even if DEQX is too expensive, you could take 3 or 4 or 5 drivers with more simple and cost effective active crossovers, run Audyssey on it and end up with a very high end results, measuring and correcting for any major FR errors that you didn't try to solve with passive crossovers. Audyssey can do amazing things for speakers that have major flaws. Less useful for a very well tweaked system though.

Besides, most 'custom' drivers have only very minor tweaks to the motor design to suit a particular passive crossover design. Take out the passive crossover and these things essentially don't matter.

That's the nice thing. Even if DEQX is too expensive, you could take 3 or 4 or 5 drivers with more simple and cost effective active crossovers, run Audyssey on it and end up with a very high end results, measuring and correcting for any major FR errors that you didn't try to solve with passive crossovers. Audyssey can do amazing things for speakers that have major flaws. Less useful for a very well tweaked system though.

Besides, most 'custom' drivers have only very minor tweaks to the motor design to suit a particular passive crossover design. Take out the passive crossover and these things essentially don't matter.

That's the direction I am going! I actually found if I checked response with my trusty Audio Control RTA after completing Audyssey, I can "tweak" the response even further with their "Design" section of the software.
We have installed a few Scan Speak combined with Audyssey systems, with great results.

That's the nice thing. Even if DEQX is too expensive, you could take 3 or 4 or 5 drivers with more simple and cost effective active crossovers, run Audyssey on it and end up with a very high end results, measuring and correcting for any major FR errors that you didn't try to solve with passive crossovers. Audyssey can do amazing things for speakers that have major flaws. Less useful for a very well tweaked system though.

Besides, most 'custom' drivers have only very minor tweaks to the motor design to suit a particular passive crossover design. Take out the passive crossover and these things essentially don't matter.

I am having fun with my custom Scan Speak MTM! I was recently experimenting with different amps and accidentally increased the output of the midwoofer,while leaving the original Audyessy settings intact.This set a downward slope above the 1750hz crossover.
Really good sound. Musical Instruments in my room kind of results.
I attached a picture of pink noise playing through the right channel with the microphone in the prime listening seat.Displayed on an RTA The downward slope result is about as "just right" as I have ever had in my many years of messing around with Hifi.
Now I have to figure out how to duplicate it.
You can see why YG used these components and configuration.

yes, you should have a downward slope at the listening position. It seems inaccurate, but it's really not. Have fun!