[James Tanner]

Hi Folks,

Ran accross this the other day and thought it may be interesting to post and ask what percentage of a speakers measurements would you consider relevant when purchasing said speaker?

http://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=16&Itemid=18

james

[vegasdave]

Frequency Response and Sensitivity

[Sasha]

Distortion, axial and power response, spl.

[redbook]

 Proper build and materials of all elements ( transducers, crossovers and enclosures) If so I think many of the top brand designs will always perform well inspite of various signatures.   

[robb]

Frequency response,
Electrical phase

[95Dyna]

Frequency response, sensitivity and impedence.  Whether you have an amp(s) or are looking for one the latter two would be important.  A bad match would render all the other specs to be moot.

[vegasdave]

Frequency response, sensitivity and impedence.  Whether you have an amp(s) or are looking for one the latter two would be important.  A bad match would render all the other specs to be moot.

This is true...

[rollo]

Frequency Response and Sensitivity

  Thats really about it. The rest lies in ones personal taste. These specs get the speaker in the room. The rest up to the speaker. Size as well.


charles

[vegasdave]

  Thats really about it. The rest lies in ones personal taste. These specs get the speaker in the room. The rest up to the speaker. Size as well.


charles

Glad you agree.

[Phoenix]

These measurements should be released for every speaker:

Free-Field Response:



Group Delay:



Harmonic Distortion:



Max. SPL:



Cumulative Spectral Decay:



Horizontal Directivity Plot:



Vertical Directivitiy Plot:




Examples are from the new Neumann KH 120 A Studio-Monitor.

And now we should all think about companies like PMC and others who are not willing to show these measurements. I for myself decided not to buy a speaker if the measurements aren't disclosed.

A speaker with good measurements can (and in most cases will) sound good. A speaker with bad measurements will never deliver a good performance.

[Diamond Dog]

Phoenix : Which speaker manufacturers other than PMC are you aware of who are "not willing" to show the measurement criteria which you have defined as being a prerequisite to " good performance " ?

D.D.

[Danny Richie]

Things to keep in mind is that not one single measurement will tell you the whole story. Each is just one aspect. What none of the measurements will tell you (truly) is how a speaker sounds. Each measurement will allow you to see if there is a problem in that area or not.

Of coarse problems clearly seen in the measured responses will indicate noticeable subjective issues.

The most telling is the cumulative spectral decay. This shows the on axis response and stored energy over time. Stored energy, break up areas, or resonance peaks are easy to hear but are unseen in all other measurements. So the CSD is very important.

The on axis response is a parcel window of the speakers output. If we listened to a speaker in an anechoic chamber then this is all that we would hear if the speakers were aimed right at us. The good thing about this measurement is that it has an industry standard of 1 watt/1 meter. So some comparisons can be made. And it will tell you the speaker sensitivity (very important).  Be careful though as some throw in wide scale ranges from 0db's to 100db's to give the appearance of a smoother response.

One of the most overlooked is the vertical off axis. This is really more important than the horizontal off axis. For one, the ceiling is typically untreated where as side walls are. So the vertical off axis can often be the most disruptive to the final in room response. The vertical off axis will also tell you how well drivers are in phase in the on axis. Or how well they are in phase over a wide range. This is not an industry standard but should be.

Horizontal off axis is important as well.  This will give you an idea of what to expect in a neutral room. The room however does tend to dominate the total in room response more so than the true horizontal off axis of the speaker. So again think of the horizontal off axis as showing if there is a problem or not and don't get real nit picky in judging a speaker from it.

Impedance responses are a must also. This will also highlight any problems regarding the load it presents to the amplifier. This one can make or break a speakers consideration.

The most useless and most misunderstood measurements are distortion measurements. There are several reasons that this is of no value. 1) You can't tell the difference between 1% and 2% of anything with music playback as there can easily be that much or more in the recording from open mics, etc. Take the recording on an electric guitar for instance. They are adding high levels of distortion to the signal already. So if a speaker adds 1 or 2% in a given area, it  could still be less than 10% of the total distortion in the signal. 2) You can't accurately measure distortion outside of an anechoic chamber because the distortion is below the room noise level and it is impossible to separate it from room noise. I have made distortion measurements and seen more difference between measurements due to a refrigerator running or not 30 feet away in another room than between two drivers. Just the noise from a standard desk top computer that I use for taking my measurements will be 4 to 5 times louder than any measured distortion from a speaker. 3) There is no standard for measuring this.

Here are the measurements that I took on a group of speakers for a small speaker shoot out several years ago. http://www.stereomojo.com/Small%20Speaker%20Shootout%202007/StereomojoSmallSpeakerShootout2007Measurements.htm

Because they were all made in exactly the same way, it is easy to make good comparisons. It is really hard to make good comparisons sometimes when speakers are measured differently or the software presents the graphs differently. So be careful in making the comparisons and remember that the measurements are for showing problems or not and they don't tell you how a speaker will sound.

[Phoenix]

Phoenix : Which speaker manufacturers other than PMC are you aware of who are "not willing" to show the measurement criteria which you have defined as being a prerequisite to " good performance " ?

D.D.

Hi D.D.!

Sorry about the little rant about PMC - it's not that I don't like PMC. The new fact8 actually is amazing and was one of the speakers at the High-End Show in Munich that impressed me the most.

Anyway, with "good" I mean that a speaker should be technically good in all scientifically verifiable facts that impact sound quality. A good speaker has to have good measurements. So why this secretiveness? Bryston is showing their specs too - they know that they are good.
And so is my question: Why don't speaker manufacturers disclose all their measurements? Are they afraid of something?

Good specs represent good engineering. And still there is room for the manufacturer to do things differently from others - even studio monitors have their own signature or character - but most of them (not all) are doing the basic things right which is not often the case with some so called "High-End"-Gear.

And this is really annoying because that undermines the trust of the costumer. And so I sometimes get angry when a company says: "We're building the best speakers...bla, bla....but we're not going to show you any (or only a few) measurements."
Mostly we get only the on axis response but, as Danny pointed out, this is only a small spot in a anechoic chamber. If you want to tell something about a speaker you need more information than this.
And if you have the measurements you can at least tell: "This speaker will eventually sound a bit brighter than neutral - but I like that!"

So, I'm always sceptical when there is little to no data available (and PMC for one has none).

Look at the Neumann site:

http://www.neumann-kh-line.com/neumann-kh/home_en.nsf/root/prof-monitoring_studio-monitors_nearfield-monitors_KH120A

You surely agree that this kind of information is much more ensuring than any marketing babble. However, there is still the chance that you don't like those kind of speakers (which would be personal taste). But in terms of hearing it right and as it was intended there is (IMO) now way going without extensive measurements.


@ Danny: Thanks for the additional input! 

So again think of the horizontal off axis as showing if there is a problem or not and don't get real nit picky in judging a speaker from it.

I don't entirely agree. The more uniform the off axis response is the more you get an even response from the room. Most important, there is no chance that undesired effects can add up. Imagine you have a room which damps quite strong at say 2000 Hz and additionally your speaker has a narrow directivity at these frequency too. These effects add up and will colour the sound.

[Danny Richie]

I don't entirely agree.

Okay, I'll elaborate.

The more uniform the off axis response is the more you get an even response from the room.

Yep, I agree 100%.

But then some guy sets his speakers two feet from the wall and there is now an out of phase cancellation at one frequency from the reflection. Or maybe it is three feet from the wall and the out of phase cancellation is lower in frequency and not as bad. But then on top of that the reflected area that is now in phase with the on axis wave is causing a peak. And the speakers off axis response could have been great.

See what I mean?

In an ideal treated room and speaker placement and with a speaker that has a balanced off axis response then the total in room response can be great (baring no other problems).

However, the total in room response of a set of speakers with a mediocre off axis response and a really well treated room might be twice as good as a speaker with a great off axis response and horrible room effects.

Rarely are conditions ideal and I very often see the person most concerned with the off axis response place their speakers in a very less than ideal condition (go figure).

Don't get me wrong. I shoot for smooth off axis responses. But more often than not the variations caused by the room over shadow them.

Again the measurements should show you if there is a problem or not, but one can't draw a lot of conclusions of quality if one speaker is marginally better in that regard than another.

[PRELUDE]

It is not that easy to look at any measurement and buy the speaker and if this is not true then,why everybody want to hear it before buy it? We can ask for the measurements and order the speaker.
How about if the measurements are done in an anechoic chamber and now you have to play them in your living room? Or,what percentage of us have a treated ceiling and can alter the room in way we want it.
And  after all I do agree with Danny that measurments should show you if there is a problem or not but has nothing to do that how the speaker would perform.

[mvwhiting_83]

Phoenix : Which speaker manufacturers other than PMC are you aware of who are "not willing" to show the measurement criteria which you have defined as being a prerequisite to " good performance " ?

D.D.

Measurements in an anechoic chamber will differ vastly from that of a real world listening room. Sensitivity & impedance are really the only measurements that should matter... Listen in your space to what you know, and see what it does for you... If you happen to like it and be drawn into it like never before, what is a chart going to do to further your enjoyment?. Music isn't about charts and graphs.... At least to me it isn't.

[Sasha]

Measurements in an anechoic chamber will differ vastly from that of a real world listening room. Sensitivity & impedance are really the only measurements that should matter... Listen in your space to what you know, and see what it does for you... If you happen to like it and be drawn into it like never before, what is a chart going to do to further your enjoyment?. Music isn't about charts and graphs.... At least to me it isn't.

I measured some of PMC speakers, and PMC speakers are very good, however a company that sells professional products and claims to base their products on sound engineering should refrain from such statements, as they only serve speculations. Measurements are extremely important and they tell a lot to prospective buyers, professionals who know what they are looking for.
In professional world sensitivity & impedance are the least important speaker parameters and are completely irrelevant in active speaker design which many top professional companies base their top models on, and far more important are the axial and power response, distortion and SPL, essentially all that companies like K&H show and PMC does not.
Regardless of how high opinion I have about PMC performance there is no excuse for refusal to provide measurements.
Besides, there are some measurements of PMC speakers available, and those can tell a lot about performance and shortcomings of PMC speakers, so again there is no point in hiding it.

[James Tanner]

The significance of what the NRC listening tests did was to show that there was a correlation between a specific set of measurements and the preferred sound in a double blind listening test setup.

I sat in on many of these tests as well and here is how it went.  In each test there would be 5 speakers at a time and you would switch at your will between the 5 and pick the one you preferred from 1 to 5. You would rate them in a number of areas – if I remember correctly about 10 different categories. The speakers were behind a transparent screen so you had no idea of size or style of speaker.

What was concluded after thousands of tests from experienced listeners to man off the street listeners was the loudspeakers which had excellent ‘on and off’ axis flat response were the clear winners. So a speaker that had a flat frequency response on axis at 0 degrees, and at 15 degrees off axis – and at 30 degrees of axis and at 60 degrees off axis where preferred in the listening tests by the vast majority of subjects.

Here is an example of a good measurement of the above on a real speaker:



The bottom 4 curves are o – 15 – 30 and 60 degrees off axis.

A reasonably flat impedance curve is also desirable and here is the curve for the same speaker above.




The theory is that speakers which exhibit these measurements sound better in most rooms because the POWER response is even through-out the room. All, rooms have serious reflection and room node issues and a speaker that has a good on and off axis response will sound better because any reflected energy has a tonal balance close to the on axis frequency response.

That’s the theory – time to listen!

james

[Sasha]

I agree with what James said above in regard to power response, but would like to add that many smooth out the measurements what essentially hides problems, makes them invisible in graphs while they are clearly audible.

[Danny Richie]

I think you guys are right on target. It is all about the power response. That is why the room is so much of an integral part of the system.