• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Neumann KH 80 DSP Speaker Measurements: Take Two

Putter

Senior Member
Forum Donor
Joined
Sep 23, 2019
Messages
493
Likes
771
Location
Albany, NY USA
I'm more than a bit in this over my head, but sometimes I feel like this is a 'How many angels can dance on the head of a pin.' What I get from the variation in measurements is that the Neumann is the best measuring speaker Amir has tested so far that is commercially available and that there are minor variations due to different measurement techniques that likely don't highlight any differences that would show up in a blind test.
 

briskly

Active Member
Joined
Jul 10, 2019
Messages
115
Likes
153
But why then does Klippel explicitly refer to the individual measurements for multi-way systems?
Source Superposition.png

The individual elements do not need to be isolated to identify a specific configuration of the loudspeaker. Again, the spherical expansion is perfectly capable of representing a sound field generated by multiple sources.
In the design of a loudspeaker, the transfer function of each emitter is not set. The superposition module is used to help identify optimal configurations of the array for the desired coverage pattern, whether by physical offset or by adjustment of electronic slopes. This is useful for line arrays, where the units are driven individually. as well as the design process of a single loudspeaker unit that incorporates multiple emitters.
 
Last edited:

Juhazi

Major Contributor
Joined
Sep 15, 2018
Messages
1,717
Likes
2,897
Location
Finland
I'm more than a bit in this over my head, but sometimes I feel like this is a 'How many angels can dance on the head of a pin.' What I get from the variation in measurements is that the Neumann is the best measuring speaker Amir has tested so far that is commercially available and that there are minor variations due to different measurement techniques that likely don't highlight any differences that would show up in a blind test.

Did you notice the "Science" in the forum's name?
 

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
It is not for nothing that there is the recommondation, for traditional style measurements, using a distance of at least ten times the largest dimension of the DUT, and preferably larger than 20 times. Otherwise you're not in the far field.

Logically speaking the "at least 10x the largest dimension" seems extremely far. If that is true then floorstanding loudspeakers are all used (and marketed) incorrectly. Most floorstanders are 3 feet (~ 0.9 m) tall. So, by this 10x method, one would need to be listening at least 30 feet (~ 9 m) away to achieve far-field conditions. That just seems ludicrous.
 

Putter

Senior Member
Forum Donor
Joined
Sep 23, 2019
Messages
493
Likes
771
Location
Albany, NY USA
I'm more than a bit in this over my head, but sometimes I feel like this is a 'How many angels can dance on the head of a pin.' What I get from the variation in measurements is that the Neumann is the best measuring speaker Amir has tested so far that is commercially available and that there are minor variations due to different measurement techniques that likely don't highlight any differences that would show up in a blind test.

Did you notice the "Science" in the forum's name?

Which is why I mentioned the blind test scenario. It's also important to not lose the 'forest' of speaker comparisons for the 'trees' of small differences due to differing measurements that are still reasonably similar.
 

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
Logically speaking the "at least 10x the largest dimension" seems extremely far. If that is true then floorstanding loudspeakers are all used (and marketed) incorrectly. Most floorstanders are 3 feet (~ 0.9 m) tall. So, by this 10x method, one would need to be listening at least 30 feet (~ 9 m) away to achieve far-field conditions. That just seems ludicrous.

Ok, I pulled out Testing Loudspeakers and looked it up. 3x the largest dimension is stated in here as the typical but up to 6x for loudspeakers that use odd-order crossovers. Maybe the consensus has changed since this version was published but I just have a hard time with 10x being the minimum recommended measuring (and therefore listening) distance.

IMG_8851.jpg
IMG_8852.jpg
IMG_8853.jpg
IMG_8854.jpg
 
Last edited:

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,442
Location
NYC
Logically speaking the "at least 10x the largest dimension" seems extremely far. If that is true then floorstanding loudspeakers are all used (and marketed) incorrectly. Most floorstanders are 3 feet (~ 0.9 m) tall. So, by this 10x method, one would need to be listening at least 30 feet (~ 9 m) away to achieve far-field conditions. That just seems ludicrous.

Yep, as mentioned earlier, no bookshelf speaker I've measured shows a significant difference at 2m vs 1m. The iLoud Micromonitors, the smallest speaker I've measured, is 7 inches or 18cm at its largest dimension. I measured it at 50cm(because the manufacturer refers to this distance for max SPL), 1m, and 2m. No meaningful difference. Maybe some of the ripples are slightly different, but the overall measurement is for all intents and purposes identical. Any difference is more likely to come from respositioning my microphone and differences in resolution from different gating.

So while I don't doubt differences exist in the "true farfield", these differences appear to be vanishingly small past a certain distance. Otherwise, as you say, tower speakers wouldn't sound right at typical listening distances of 3-5 meters

I'm more than a bit in this over my head, but sometimes I feel like this is a 'How many angels can dance on the head of a pin.' What I get from the variation in measurements is that the Neumann is the best measuring speaker Amir has tested so far that is commercially available and that there are minor variations due to different measurement techniques that likely don't highlight any differences that would show up in a blind test.

Don't take this the wrong way. The Neumann is good, we all knew that coming into this review. But there are clear inconsistencies in Amir's measurements. With all due respect, from measuring speakers, it's clear that measurements should easily be replicable.

Also the differences described so far would 100 percent be audible in a blind test. Go ahead and apply a 12dB rolloff to your speakers from 7KHz to 20KHz =]

If you're happy with the results, that's awesome. But for those of us interested in the science of how the klippel works and to what degree its results are accurate and reliable, there are some inconsistencies that are worth questioning that are pretty obvious if you've ever measured speakers. This isn't to say the klippel isn't useful or an amazing contribution, but I want to know what it's differences, limits, and advantages are compared to traditional anechoic and quasi-anechoic measurements.
 

LeftCoastTim

Senior Member
Forum Donor
Joined
Apr 15, 2019
Messages
375
Likes
757
I think this thread illustrates the state of audiophoolery in all its glory.

There are so many crack-pot ideas in here, that one cannot tell the what's real and what's crack-pot.

So for me, I'm sticking with Occam's razor and CEA2034 measurements and ignore the rest, unless Amir gets a sample from Neumann that THEY'VE MEASURED to directly compare with Amir's.
 

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,442
Location
NYC
I think this thread illustrates the state of audiophoolery in all its glory.

There are so many crack-pot ideas in here, that one cannot tell the what's real and what's crack-pot.

So for me, I'm sticking with Occam's razor and CEA2034 measurements and ignore the rest, unless Amir gets a sample from Neumann that THEY'VE MEASURED to directly compare with Amir's.

To be clear, no one is really doubting Amir's measurement at the tweeter axis. That seems correct. Maybe a few interference artifacts from the measurement rig (as amir notes) but fundamentally correct for a measurement made at the tweeter axis.

In the previous thread, people complained about the choice of axis and SPL levels. So Amir very kindly remeasured the speaker at the "proper" axis and more reasonable SPL levels. The bass measurements hold up.

The treble response, however, simply doesn't make sense in the reference axis remeasurement So we're trying to figure out why that is. As noted earlier in the thread, even if you original spinorama and move 10 degrees below it, you don't get the response amir gets.
 

briskly

Active Member
Joined
Jul 10, 2019
Messages
115
Likes
153
Basic module without MSS must use measurement distance that represents "free field conditions" for the DUT.

This is the problem/question I have been asking all the time...
Next question is "How do you dertermine the distance needed for FFC?"
Multiple Source Separation is a module distinct from the Sound Field Separation. The latter is required to isolate low-frequency behavior in an acoustically reflective room. The requirement of the anechoic space is for users without the SFS module.

Logically speaking the "at least 10x the largest dimension" seems extremely far. If that is true then floorstanding loudspeakers are all used (and marketed) incorrectly. Most floorstanders are 3 feet (~ 0.9 m) tall. So, by this 10x method, one would need to be listening at least 30 feet (~ 9 m) away to achieve far-field conditions. That just seems ludicrous.
The far-field response is an asymptotic limit, so the question is about the accuracy penalties in assuming far-field behavior. Here is one example showing the axial effects of proximity on a large* box driven by a small pistonic disc.
Loudspeaker Cabinet Diffraction (Jeff Candy).png


*(L x, L y, Lz) = (0.94m, 1.24m, 0.64m). Driver located at (x, y) = (0.47m, 0.76m) on Sz
Taken from Jeff Candy's AES publication concerning problems with certain diffraction simulators, available here. Comments on the article's AES page may be of interest.
 
Last edited:

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
Far field starts at 3 times of the drivers diameter. It hasn't nothing to do with box size.

Loudspeakers and Drivers are measured differently. Look at my above pictures for far-field definition of loudspeakers.

Below are excerpts from the same book but wrt measuring drive units.

IMG_8856.jpg

IMG_8857.jpg

IMG_8859.jpg

IMG_8858.jpg
 

NTK

Major Contributor
Forum Donor
Joined
Aug 11, 2019
Messages
2,660
Likes
5,820
Location
US East
I see that there is some fundamental misunderstanding of the mathematical technique the Klippel NFS uses when the nearfield measurements are compared to the far field.

When far field measurements are taken the traditional way, the measurements are the reported results (but usually with distance adjustments). That means, if I measured a certain SPL at a certain azimuthal angle and elevation angle at a certain distance, that would be the reported result.

The NFS process works differently. All of the nearfield measurements are further processed to create a mathematical model of the sound pressure field in 3D (using spherical wave expansion functions). That's why the technique used is called nearfield acoustical "holography" (NAH) -- recreation the 3D field from 2D measurements. If you need to report the SPL at a certain location given by its azimuthal angle, elevation angle, and distance, these coordinates are fed back into the model and the SPL at that location is calculated and reported. The way NAH technique works is fundamentally different from the traditional (far field) methods.
 

DS23MAN

Active Member
Joined
Sep 18, 2019
Messages
161
Likes
190
Location
Hypex doorstep

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
Just play around in Edge. Put a 25cm woofer in a 30x30 cm baffle. Change the mic distance between 1m and 10m. No change in the respons in the normal used fr range for this size driver.

I use Edge to simulate and obtain a 'calibration' for my driver measurements. So I am familiar with it. With Edge, you're measuring the diffraction of the baffle. Not the high-frequency driver effects.

The Edge is and will probably remain a pure baffle diffraction simulator.
 

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,161
Location
Tarragona (Spain)
..Also, I wouldn’t take Danny Ritchie too seriously. He’s trying to sell (often useless) tweaks, not advance the state of reproduction.

I disagree. Some of the things he says I have been checking, including something as doubtful, a priori, as the loudspeaker connectors. Or bypass the tweeter capacitor, change sand resistance for another Mills MRA. Of course he wants to sell, he lives on it, but without cheating!

- End off topic -

Regarding measuring distance, John Atkinson (Stereophile): 50", aka 127 cm.

The last stand loudspeaker measured: https://www.stereophile.com/content/q-acoustics-concept-300-loudspeaker-measurements

120Q300fig06.jpg


Fig.6 Q Acoustics Concept 300, anechoic response on tweeter axis at 50", averaged across 30° horizontal window and corrected for microphone response, with the complex sum of the nearfield woofer and port responses plotted below 300Hz.
Well, that, that there is too much variety when choosing the measurement distance, which makes comparison difficult.
 
Last edited:

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,423
Likes
2,407
Location
Sweden
I see that there is some fundamental misunderstanding of the mathematical technique the Klippel NFS uses when the nearfield measurements are compared to the far field.

When far field measurements are taken the traditional way, the measurements are the reported results (but usually with distance adjustments). That means, if I measured a certain SPL at a certain azimuthal angle and elevation angle at a certain distance, that would be the reported result.

The NFS process works differently. All of the nearfield measurements are further processed to create a mathematical model of the sound pressure field in 3D (using spherical wave expansion functions). That's why the technique used is called nearfield acoustical "holography" (NAH) -- recreation the 3D field from 2D measurements. If you need to report the SPL at a certain location given by its azimuthal angle, elevation angle, and distance, these coordinates are fed back into the model and the SPL at that location is calculated and reported. The way NAH technique works is fundamentally different from the traditional (far field) methods.

One question:

Why do you have to define where the acoustic axis beforehand, since it should be possible to calculate the frequency response in silico at any axis in the far field?
 

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,728
Location
Prague
It is a width of the front baffle that counts. That floorstanding Klipsch that had a deep dip in 1m measurement (and not in 2m measurement) just tells about wrong speaker design. Either distance of drivers with respect to crossover frequency or crossover circuit.

I measure this from 70cm distance on a 2-way speaker with 95 x 20 x 28 cm dimensions. 3ms window to cut reflections.
70cm_FR2.jpg
ns.
 
Top Bottom