• 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 Monitor Review

Status
Not open for further replies.
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,376
Likes
234,530
Location
Seattle Area
If the goal is to just find speakers to avoid than the current procedure is fine. It might make sense to avoid any speakers like the Neumanns that have a large amount of quality measurements available and just focus on potential "offenders" or interesting speakers without a solid suite of data. If more is going to be asked of the data set maybe more needs to be done to refine the setup.
I don't trust any manufacturer or magazine measurements relative to capabilities we have. Using the identical measurement system allows us to much better compare one speaker to another. In addition, I perform listening tests to see if I can correlate measurements with my experiences.

In addition, the measurements provide a wealth of information about the off-axis performance of a speaker. That is very helpful in determining how to position them in your room and how to apply room treatments.
 

Mawclaw

Member
Forum Donor
Joined
Aug 20, 2019
Messages
69
Likes
133
Location
The ATL
Thanks for the reply. That makes a lot of sense and I appreciate your efforts.

On the SPL topic here is the breakdown of max SPL at different frequencies and distances from the Neumann website.

Sorry for the screenshot, I am on mobile.
 

Attachments

  • Screenshot_20200124-010009.png
    Screenshot_20200124-010009.png
    325 KB · Views: 154

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,442
Location
NYC
As I noted, I am standardizing on Soundstage loudness standard so that we can compare my results to them. They use 90 dB at 2 meters. That is 106 dB at 1/3 meter. They say that they lower the level if the speaker strains. I did not hear these strain. Higher volumes allow more exclusion of noise for even more accurate measurements.

This is the manufacturer spec:
View attachment 47152

I am assuming that is at one meter. At 1/3 of meter then, it should be able to produce 119 dB which is well in excess of the level I measured at.

If you want to standardize for an equivalent of 90 dB at Two meters, then that's totally fine and up to you, but compression issues aside, your original image does clearly say 106dB equals = 90dB @1m. Hence the confusion about which one was your intent.
Snag_15054af6.png



That doesn't matter. The Klippel system does not care what the speaker looks like. It wants the tweeter axis because it will have an easier time computing the spherical coefficients for the field expansion. Setting it to the wrong spot simply increases the complexity of the soundfield it has to compute. It performs a post measurement analysis and reports on its difficulty to do so. There was no problem at all in this case.

I also don't know how Neumann computed the acoustic center anyway. That could just be the axis they want you to listen at, not some type of accurate measurement of soundfield that resulted in that.

Well, it's where the speaker measures flattest from (including in your own measurements at -10v) and how one would presume most consumers should/would be setting it up, that's all. Again, Imo not a big deal at the end of the day, just caused some confusion to start. Some speakers do show big differences if the reference axis is off a by 5-10 degrees, so might be important to consider going forward.

Just might be worth clarifying these two things in the OP. At least the intended SPL conversion.
 
Last edited:

Thunderlips

Member
Joined
Jan 24, 2020
Messages
33
Likes
32
Had to sign up just to comment on these spectacular Neumann KH 80 DSP speakers and their measurements. These results are pretty much state of the art. As expected from a reputable European manufacturer. Add a KH 750 DSP subwoofer and it's a killer entry level into high end system. Excited to see how Genelecs compare.

In comparison, the Samsung 305P MKII's have a suckout of almost the entire midrange (250Hz-1500Hz), followed by sizable peaks at 1700Hz and 3300Hz. That's pretty audible.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,423
Likes
2,407
Location
Sweden
As I noted, I am standardizing on Soundstage loudness standard so that we can compare my results to them. They use 90 dB at 2 meters. That is 106 dB at 1/3 meter. They say that they lower the level if the speaker strains. I did not hear these strain. Higher volumes allow more exclusion of noise for even more accurate measurements..

Soundstage measure at 2 or 3 meters and standardize their graphs to 90 dB at 1 meter, not 2 meters. Sensitivity is given at 1 meter.

"As volume increases, all frequencies should rise at the same rate. However, as a speaker is stressed, compression will occur at certain frequencies. The stress may be mechanical, thermal or otherwise. This test shows those frequencies at which deviation occurs as a result of compression. Many speakers show slight deviations at 90dB. Most speakers start to show serious deviations at 95dB. Very few speakers can be tested at 100dB without damage."
 

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,758
Likes
16,222
I had yesterday a short talk with the designer of the KH80 and he also considers the the too high level (95,5 dB at one meter) as a very possible reason of the bass drop as it is in the region where the limiter can be active. Another reason of deviations above is as said the wrong vertical reference axis.

I agree with all above comments that because of that and because the KH80 is one of the most linear loudspeakers currently in the market on- and off-axis it would be really worth to rerun its measurement with these 2 corrections to have it as kind of a "golden standard" (the deviations of each KH production model is less then 0,5 dB so I am very sure its not a production tolerance) and another confirmation outside from the Harman group for the NFS system.

I am also sure most here would rather have one unit/day delay of the test pipeline but higher confidence for future tests. Also I think doing the frequency response measurements with a level equivalent to 90 dB in 1 meter in the future should be considered as 95 dB is for many compact floorstand/desktop/monitoring speakers too high. On-axis distortion measurement sweeps can be of course very quickly done in several levels like for example the German (Audio & Stereoplay) press does, 80, 85, 90, 95, 100, 105 dB etc. till significant distortion is reached.

We are just on the beginning of something great, so we rather carefully adapt and further perfect the procedure now before we regret it later after having measured dozens of loudspeakers.
 
Last edited:

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,423
Likes
2,407
Location
Sweden
This is the manufacturer spec:
View attachment 47152

I am assuming that is at one meter. At 1/3 of meter then, it should be able to produce 119 dB which is well in excess of the level I measured at.

Their graph says otherwise 100 Hz-10 kHz, so the above max SPL must be an average over the frequency range.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,423
Likes
2,407
Location
Sweden
And assuming a an Sd of 50 cm2 and xmax 5 mm , max SPL at 1 meter and 100 Hz should be in the range of 90-91 dB according to my calculations.
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
This argument during last few pages makes me want to remind, that there are other measurements too, spinorama/CEA doesn't tell everything about a loudspeaker! It is like arranging Miss Universe competition and judging the girls by their fingernails!

Armi-Kuusela%20(21).jpg

IIRC the reason why other parameters (time domain stuff, THD, etc.) didn't make it to CEA standard is that in the listening tests listener didn't put much weight on them. Following your analogy that would imply that CEA reveals pretty much most of the "strategic" parts of your Miss Universe and that fingernails are actually the ones been hidden. :D
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
Well, I do actually have to, you know, work sometimes :). Audio isn't even my focus at my job. I also don't have the 305Ps. But it would be happy to try it sometime down the road :)

Oh, so you're saying you have a life outside of this audio world, like having a partner, job and so on? :D

Btw, any 2 monitors with similar bass extension will do, otherwise the one with lower LF extension will have an advantage.

Just be sure to EQ them properly as Mitch did.
 
Last edited:

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,923
Likes
7,616
Location
Canada
As I noted, I am standardizing on Soundstage loudness standard so that we can compare my results to them. They use 90 dB at 2 meters. That is 106 dB at 1/3 meter. They say that they lower the level if the speaker strains. I did not hear these strain. Higher volumes allow more exclusion of noise for even more accurate measurements.

This is the manufacturer spec:
View attachment 47152

The spec you pasted is for 100hz - 10khz. I pasted the bass spec earlier in the thread. It's 92.1db @ 1m averaged between 50-100hz in half space at 3% thd. Much, much lower. The fact that it's half space lead me to believe you may have exceeded it during this test. Not sure, but it seems within the range of possibility. Someone else pasted a graph from the german reviewer which indicates the KH80 will let you go up to at least 10% THD below 250hz, though, so I'm not really sure. But it is possible Neumann changed the limiter behaviour at some point. I would expect these types of speakers go through many software revisions during their lifespan.
 
Last edited:

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
That experience is one of a few that convince me that frequency response is the over-riding factor in most things we hear in stereo playback of music.

Very true. Yet not many people are realising that and willing to do proper room EQ to get the frequency response right.
 

BYRTT

Addicted to Fun and Learning
Forum Donor
Joined
Nov 2, 2018
Messages
956
Likes
2,452
Location
Denmark (Jutland)
JBL 305P Mark ii was prefered in the "Informal Listening Tests/Conclusions" and made some discussions here and there :) myself have no of these two beautiful speakers or heard them but sense many here that prefer KH 80 focus on smooth on axis as possible, so tried virtual EQ in VituixCAD using the 72 steps amirm shared to get JBL 305P Mark ii as smooth as possible on axis plus another EQ transfer function to same curve as KH 80 was scanned, also later intend do the same exercise for KH 80.

JBL 305P Mark ii out of box as scanned:
Vcad_EQ_Power+DI.png



JBL 305P Mark ii EQed smooth on axis filter function included:
Vcad_EQ_Power+DI_1.png



JBL 305P Mark ii transfer EQ to KH 80 out of box response filter function included:
Vcad_EQ_Power+DI._305P_transfer_to_KH80png.png
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
JBL 305P Mark ii was prefered in the "Informal Listening Tests/Conclusions" and made some discussions here and there :) myself have no of these two beautiful speakers or heard them but sense many here that prefer KH 80 focus on smooth on axis as possible, so tried virtual EQ in VituixCAD using the 72 steps amirm shared to get JBL 305P Mark ii as smooth as possible on axis plus another EQ transfer function to same curve as KH 80 was scanned, also later intend do the same exercise for KH 80.

JBL 305P Mark ii out of box as scanned:
View attachment 47167


JBL 305P Mark ii EQed smooth on axis filter function included:
View attachment 47168


JBL 305P Mark ii transfer EQ to KH 80 out of box response filter function included:
View attachment 47169

Did you with this 3rd curve applied on-axis filter to make JBl to have the same curve as KH80? :D
 

edechamps

Addicted to Fun and Learning
Forum Donor
Joined
Nov 21, 2018
Messages
910
Likes
3,620
Location
London, United Kingdom
You guys are honestly debating one or two dB here and there??? I reported that based on my measurements, this speaker has very flat on-axis response. That is all we need to know. Not, "oh wait, why is there a 1 dB dip at 200 Hz relative to this other person's measurements." Please learn to allow variances here.

I understand that you can't measure speakers as accurately as you can measure electronics. However, I feel like you're being too pessimistic about the variance one has to expect with these measurements.

@napilopez put it very well when he posted a comparison of the various measurements of the KH 80:

index.php


The bottom 3 measurements look very similar, only yours stands out. There is variance in the bottom 3 of course, but not nearly as much as you seem to suggest is "par for the course". This suggests the differences in your measurement can't be explained by random variance alone. Especially since the differences are relatively low-Q, i.e. they affect the overall trend of the response, it's not just local variations.

People here have suggested likely explanations for these differences. The drop in low bass likely comes from compression. The rise in the treble likely comes from measuring using the tweeter as the acoustical axis as opposed to the one that the manufacturer specifies (centred on the Neumann logo). Assuming these explanations are correct, these are very easy to fix: just re-measure at a lower level with the proper acoustical axis. Then, hopefully, you'll get a graph that matches the other 3, most of the "variance" will be gone, everyone will be happy, and doubts about your measurement process can be put to rest.

Quite frankly I'm a bit surprised that, in the case of electronics, you're more than willing to go to extreme lengths to investigate things like the ESS hump or to push the boundaries on 120 dB+ SINADs; but when it comes to doing speaker measurements with laboratory equipment that costs the price of a luxury car, you're willing to dismiss very suspicious, broad, low-Q frequency response anomalies as just "variance", even though measurements other people made don't show such variance.

Don't you want to know what your system is truly capable of? Why not try to extract its full potential instead of rushing through reviews? Why accept such "variance" from the most advanced, state-of-the-art, accurate speaker measurement system on this planet when there are clear signs such variance should not be there and can be explained in other ways?
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,244
Likes
11,484
Location
Land O’ Lakes, FL
I understand that you can't measure speakers as accurately as you can measure electronics. However, I feel like you're being too pessimistic about the variance one has to expect with these measurements.

@napilopez put it very well when he posted a comparison of the various measurements of the KH 80:

index.php


The bottom 3 measurements look very similar, only yours stands out. There is variance in the bottom 3 of course, but not nearly as much as you seem to suggest is "par for the course". This suggests the differences in your measurement can't be explained by random variance alone. Especially since the differences are relatively low-Q, i.e. they affect the overall trend of the response, it's not just local variations.

People here have suggested likely explanations for these differences. The drop in low bass likely comes from compression. The rise in the treble likely comes from measuring using the tweeter as the acoustical axis as opposed to the one that the manufacturer specifies (centred on the Neumann logo). Assuming these explanations are correct, these are very easy to fix: just re-measure at a lower level with the proper acoustical axis. Then, hopefully, you'll get a graph that matches the other 3, most of the "variance" will be gone, everyone will be happy, and doubts about your measurement process can be put to rest.

Quite frankly I'm a bit surprised that, in the case of electronics, you're more than willing to go to extreme lengths to investigate things like the ESS hump or to push the boundaries on 120 dB+ SINADs; but when it comes to doing speaker measurements with laboratory equipment that costs the price of a luxury car, you're willing to dismiss very suspicious, broad, low-Q frequency response anomalies as just "variance", even though measurements other people made don't show such variance.

Don't you want to know what your system is truly capable of? Why not try to extract its full potential instead of rushing through reviews? Why accept such "variance" from the most advanced, state-of-the-art, accurate speaker measurement system on this planet when there are clear signs such variance should not be there and can be explained in other ways?
If you look at the vertical directivity graph I posted, the hot upper treble is tamed at -10° with no negatives elsewhere.
However, I highly doubt this would have any real impact on perceived sound quality or the preference score, it would be just validating the measurements.
 

BYRTT

Addicted to Fun and Learning
Forum Donor
Joined
Nov 2, 2018
Messages
956
Likes
2,452
Location
Denmark (Jutland)
For responses could be amirm is right regarding variances is not so big a deal and maybe the various scales used confuse a bit, for example below black (NFS) verse purple (anechoic) was C52 and the one that turned opinion of skeptic to approve results from NFS, now look at black (NFS) KH 80 verse red third party KH 80, of course bragging rights for the red curve is lost a bit but are we too sensitive here ...:)

deviations.png
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
For responses could be amirm is right regarding variances is not so big a deal and maybe the various scales used confuse a bit, for example below black (NFS) verse purple (anechoic) was C52 and the one that turned opinion of skeptic to approve results from NFS, now look at black (NFS) KH 80 verse red third party KH 80, of course bragging rights for the red curve is lost a bit but are we too sensitive here ...:)

View attachment 47170

True. But repating the measurements with lower level and using "official" axis, as @edechamps suggested, can't hurt.
 

edechamps

Addicted to Fun and Learning
Forum Donor
Joined
Nov 21, 2018
Messages
910
Likes
3,620
Location
London, United Kingdom
look at black (NFS) KH 80 verse red third party KH 80, of course bragging rights for the red curve is lost a bit but are we too sensitive here ...:)

Thank your for the graph, it presents things quite clearly indeed.

I don't think we are being "too sensitive" here. The KH 80 black curve on your graph (i.e. @amirm measurement) shows somewhat of a tilt which is abnormal. That kind of frequency response distortion is extremely audible - research shows we can perceive broad, low-Q tilts like these even when they're only 0.1 dB in magnitude. It's really important to get this right. (This is different from just the graph being "squiggly", i.e. small high-Q variations, which are much less audible, and are normal and expected due to differences in smoothing and actual measurement "variance".)

Amir's measurement shows bass around 100 Hz and below to be about ~3 dB (eyeballing the graph here) lower than it should be, and a slight excess of treble starting from 4 kHz all the way to the end of the spectrum. That's low-Q and, as far as research shows, quite audible. I don't think that can be explained by random "variance", IMHO. Even a -3 dB frequency response range spec for the speaker would look quite different depending on which measurement you use: if you use the red graph, the -3 dB point is about 60 Hz; if you use the black graph, it's 100 Hz! That's a very significant difference.

It doesn't have to be that way. Many people in this thread suggested that we might be able to eliminate these anomalies through a few tweaks to the measurement setup (output level, for the bass issue, and acoustical axis, for the treble issue). Better do that now than having that same debate over and over again on every review that shows abnormally low bass or abnormally bright treble. Why settle for low accuracy and measurement anomalies when there are clear indications we can do better?
 
Last edited:

dukanvadet

Member
Forum Donor
Joined
Oct 18, 2019
Messages
72
Likes
114
And assuming a an Sd of 50 cm2 and xmax 5 mm , max SPL at 1 meter and 100 Hz should be in the range of 90-91 dB according to my calculations.
It looks like a ported design so i would expect a little more.
 
Status
Not open for further replies.
Top Bottom