Neumann KH 150 Monitor Review

[fuzzychaos]

estimated in room response is a far field prediction. You won't be listening to KH 150's from far field.

-While sitting very close to speakers(*close* in this context depends on speakers and how the drivers sum up), 2 factors determine the tonality & tonal balance:

Sum of early reflections PLUS direct sound.

As you get farther from speakers, sound power response of speakers start influencing tonal balance more and more. That's why Estimated in room response consists of
44% direct sound, 44% early reflections, 12% sound power. Estimated in room responses do not give much idea about how nearfield speakers perform unless you'll be sitting very far to speakers which is not what they are designed for! (and the magnitude of how much sound power influences the tonal balance depends on listeners distance to speakers, that 12% is going to be higher if you are sitting in a completely reverberant field where direct sound can't have an influence)

-To wrap up, if you are sitting in a reverberant field(far from speakers),

Sum of early reflections PLUS sound power PLUS direct sound determine tonal balance.

Depending on your distance to speakers the influence of sound power in perceived tonal balance changes. Far field estimation is basically a very rough estimation of how speakers should sound from afar.

I see people keep checking far field responses of nearfield speakers which they listened from nearfield. That is so pointless.

So why do Neumann speakers -usually- sound darker?

Because their Early reflections Directivity Index and Sound power Directivity Index is higher than Genelec speakers. Remember, we hear a combination of reflections and direct sound. And how reflections are formed inside a room can be predicted based on the directivity of speakers.
For example:

View attachment 254052

Even if we EQ KH 420's on-axis to 8361's on-axis response, 8361's will always sound a bit brighter than Neumann KH 420,s because their directivity index in higher frequencies is lower overall.

“You won't be listening to KH 150's from far field.” - Nuemann states they are useable up to 18’. However, 3’ to 8’ is recommended. I listen to them from about 11’ away and the sound great.

If I were using them to mix, sure, the closer the better to a point. I listen to them in my living room for pleasure.

 

[Zaireeka]

because EQing speakers above the room transition frequency(x3 above of it~) is not a good idea.

To what I understand from Toole's work, EQing the hills and dips above schroeder freq is not advised while setting a global tone curve (or a tilt EQ) is perfectly ok when needed. The man himself is also in favor of the return of tone controls on amplifiers (to compensate for too bright/dark sounding recordings)...
I set a -1.5dB high shelf filter on my speakers to tame the highs a bit, and I feel completely ok with doing it.

 

[phile2]

Hello,
Still very interesting thread, but...
I find a bit useless to talk & talk about Amir measurements about directivity & so on.
These measurements are interesting as such but they are relevant regarding "listening", "only" when you listen any speaker in near-field (1.5m max, without room effect).

Then, we have 2 kinds of people =>
1. the ones that listen near-field (to "produce" music for instance)
2. the ones that listen far-field for pleasure.

The very few ones that listen near-field for pleasure are... very few... (I'm talking about "in living-room", not on desk)

Based on that, why so much blabla about precision & directivity of these speakers, for daily use into a living room ? (do you all listen near-field ? waouh !?)
I did not listen to the KH150, but this "near-field" speaker, simply put them into a 40-50m2 living room, with 1 sub, and they will sound very very nicely.
"near-field" is a "sticker", it doesn't mean that it sounds good only at 1.5m because Neumann designed these speaker for "this" use... LOL..

Then, something important that has not been discussed (I'm talking about far-field use now, 3m min) =>
1. no one buy a KH or Genelec or Dyn without listen to them.
They all have their "sound". You listen to them, side by side, and you'll buy "these ones" because you "like" their sound.
Thus, here => "precision" & Amir's measurements etc... you don't care, the buying choice is not related to that.
I bought my Dynaudio Core59 after listening, knowing that after reading many pages about KH & Genelec I was sure I was going to listen & buy a KH or a Genelec (finally my ears said => forget your readings, the 59 are the one )

2. the implementation of these speakers.
These actives are bloody great stuff. But we should have in mind that "source first" is still valid with actives.
Thus, the KH420... are not fancy with their sole old-fashioned analogue input... Genelec's digital input enables to step up SQ.
The KH150 are great (on paper) because they have analogue & digital inputs (although I'm laughing about this crapy coax digital input... I don't get why Neumann did not put an AES instead...). Use the analogue input and you'll find the limit "A". Use the digital input with a proper digital interface, and you'll overcome "A" easily.
I'm saying that because I did the test (analogue vs digital) on the Core59 ; digital input beats analogue... wah ! By the way, the Core series has word clock inputs... another tool to even improve the SQ... no wordclock-in in genelec & KH... LOL (but I guess the AES67 option on the KH150 will be a really interesting feature !)
my 2 cents

 

[Rednaxela]

These measurements are interesting as such but they are relevant regarding "listening", "only" when you listen any speaker in near-field (1.5m max, without room effect).

Not true.

 

[GXAlan]

To what I understand from Toole's work, EQing the hills and dips above schroeder freq is not advised while setting a global tone curve (or a tilt EQ) is perfectly ok when needed. The man himself is also in favor of the return of tone controls on amplifiers (to compensate for too bright/dark sounding recordings)...
I set a -1.5dB high shelf filter on my speakers to tame the highs a bit, and I feel completely ok with doing it.

My understanding is this is because you are more prone to running into measurement errors, but an Eq based off anechoic data that is full frequency is OK?

 

[LTig]

I find a bit useless to talk & talk about Amir measurements about directivity & so on.
These measurements are interesting as such but they are relevant regarding "listening", "only" when you listen any speaker in near-field (1.5m max, without room effect).

No, it's just as important in near field (when the slightest movement of your head puts it away from the sweet spot) as in mid/far field (when the room reflections dominate).

 

[Sancus]

because EQing speakers above the room transition frequency(x3 above of it~) is not a good idea.

This isn't true at all. EQing speakers with good off axis response works great. It's just a form of tone control and we already know tone controls are critical to compensating for poor recordings and for personal preferences in bass and treble.

 

[Blockader]

“You won't be listening to KH 150's from far field.” - Nuemann states they are useable up to 18’. However, 3’ to 8’ is recommended. I listen to them from about 11’ away and the sound great.

If I were using them to mix, sure, the closer the better to a point. I listen to them in my living room for pleasure.

Then It's better to buy speakers which have better far field performance. KH150's are designed to perform best in direct sound dominated sound fields.

 

[danbei]

What precisely is not good in KH150 far field performances?

 

[changer]

Then It's better to buy speakers which have better far field performance. KH150's are designed to perform best in direct sound dominated sound fields.

Tell me why.

 

[Blockader]

This isn't true at all. EQing speakers with good off axis response works great. It's just a form of tone control and we already know tone controls are critical to compensating for poor recordings and for personal preferences in bass and treble.

Wrong. Please read the chapter where Toole explains fusion interval and precedence effect.

The reasons are:

1- Direct sound is perceptually dominant. Our auditory system fuses direct sound and reflections to each other(for a certain time window for each frequency), however if tonal balance of the direct sound is uneven, we'll perceive the outcome as uneven, no matter what the timbre of reflections are. Human hearing is not linear like microphones have. Microphones simply sum direct sound with reflections. In short, EQing on-axis response of flat speakers to fix the problems in total sum of reflections and direct sound is not a great idea. High shelf/low shelf filters are better than notch or high Q filters to, I agree, however why change a variable which is dominant and correct(flat on axis response of KH 150s for example)?

2- Eqing speakers above 700hz(or room transition frequency x2/x3~) is not a good idea. Because above that frequency human auditory system starts discerning direct sound from reflections. Due to non linearities in sound power response of speakers, EQing on-axis +4db may result a +5db tonal change in reflections. Above room transition frequency, EQing on-axis response of flat speakers do not only make the tonal balance of on-axis sound(which is perceptually dominant) uneven, it also widens the tonal difference between on-axis sound and reflections speakers.(assuming speakers do not have perfect off-axis response) Only very few speakers have smooth enough ERDI and SPDI(for example Genelec S360 in high frequencies) to be EQ worthy in my opinion(you might disagree with that, I can understand that).

Also then comes the question, if constant-change-in-DI speakers or constant DI speakers are more EQable?

 

[Blockader]

Tell me why.

Please read my post where I explained what estimated in room response is, I'll repeat briefly

Far field estimation =estimated in room response = (44% direct sound + 44% early reflections + 12% sound power). Sound power response tells us how speakers sound in a completely reverberant field. The thing is, KH 150/KH 310 are designed to be listened in a direct sound dominated field.

The most interesting example of speakers which have incredible sum of early reflections and uneven sound power is Neumann KH 310. They are designed for nearfield use, while the listener sitting very close to speakers, the sound field is dominated by the direct sound hence the prominent factors in determining sound quality in such close distances are only sum of early reflections and direct sound. Neumann designed KH 150s and KH 310s waveguide and driver configuration in such way that they have incredible sum of early reflections but clearly uneven sound power response. Simply because, sound power response of these speakers do not matter for their use cases much.(again, I repeat, they are designed for nearfield use) As you get farther from those speakers they sound starting wonkier compared to, say Genelec 8361 because, again, their sound power response is not as balanced as far field speakers have. KH 420 for example, designed in such way the drivers and waveguides are configured to perform great in both direct sound dominated field and reverberant field hence they have incredibly ERDI and SPDI at the same time. In layman terms, KH 150s drivers and their radiation pattern sum up pretty well in short distances but not so well in long distances.

 

[Robbo99999]

Wrong. Please read the chapter where Toole explains fusion interval and precedence effect.

The reasons are:

1- Direct sound is perceptually dominant. Our auditory system fuses direct sound and reflections to each other(in a certain time window for each frequency), however if tonal balance of the direct sound is uneven, we'll perceive the outcome as uneven, no matter what the timbre of reflections are. Human hearing is not linear like microphones have. Microphones simply sum direct sound with reflections. In short, EQing on-axis response of flat speakers to fix the problems in total sum of reflections and direct sound is not a great idea. High shelf/low shelf filters are better than notch or high Q filters to, I agree, however why change a variable which is dominant and correct(flat on axis response of KH 150s for example)?

2- Eqing speakers above 700hz(or room transition frequency x2/x3~) is not a good idea. Because above that frequency human auditory system starts discerning direct sound from reflections. Due to non linearities in sound power response of speakers, EQing on-axis +4db may result a +5db tonal change in reflections. Above room transition frequency, EQing on-axis response of flat speakers do not only make the tonal balance of on-axis sound(which is perceptually dominant) uneven, it also widens the tonal difference between on-axis sound and reflections speakers.(assuming speakers do not have perfect off-axis response) Only very few speakers have smooth enough ERDI and SPDI(for example Genelec S360 in high frequencies) to be EQ worthy in my opinion(you might disagree with that, I can understand that).

Also then comes the question, if constant-change-in-DI speakers or constant DI speakers are more EQable?

Probably a good reason for doing an Anechoic Listening Window EQ from Amir's spinorama. I do that for my 308p's that were measured by Amir - Anechoic EQ of Listening Window for a smooth Listening window whilst at the same time trying to strike a tonally Anechoic Flat on-axis response.

 

[Sancus]

Wrong.

It is important for consumers to realize that it is not a crime to use tone controls. Instead, it is an intelligent and practical way to compensate for inevitable variations in recordings, that is, to “revoice” the reproduction if and when necessary. At the present time, no loudspeaker can sound perfectly balanced for all recordings.

Eqing speakers above 700hz(or room transition frequency x2/x3~) is not a good idea.

No one is talking about "fixing" broken speakers here. Everything you listen to has been EQed several times between the performance and your system. Did that break the direct/reflected sound balance too? No, of course not. It is completely normal to EQ the steady state room curve to taste, not only is it strongly encouraged throughout Toole's book, but there is literally an image from a study showing variance in that taste(Figure 12.7, "Subjectively preferred steady-state room curve targets")

In short, EQing on-axis response of flat speakers to fix the problems in total sum of reflections and direct sound is not a great idea.

So every single active speaker, including this one, is wrong? Because they're *literally all EQed above the transition frequency to accomplish that*.

This isn't rocket science. If a well-designed speaker sounds too bright *on the recording you're playing* then shelve the treble down a bit. Or use tone controls if your electronics have those in a place that's easy to access. Or change the target curve of your room correction. Or whatever. None of these things are wrong or bad to use, in fact they're basic requirements of a good playback system. As directly stated in the book.

P.S. No matter how many times people repeat this, there is still no magical difference between "nearfield", "midfield", and "farfield" speakers other than output. There is absolutely no design characteristic in this speaker that makes the tonality designed for the nearfield.

 

[MrSoul4470]

Thus, the KH420... are not fancy with their sole old-fashioned analogue input... Genelec's digital input enables to step up SQ.

I doubt that digital inputs really make a significant audible difference. At least not to my tin ears.

 

[Blockader]

P.S. No matter how many times people repeat this, there is still no magical difference between "nearfield", "midfield", and "farfield" speakers other than output. There is absolutely no design characteristic in this speaker that makes the tonality designed for the nearfield.

I am talking about sound power response of speakers. Some nearfield speakers are designed in such way that they perform great from nearfield but not so great from in longer distances, their sound power responses tell that story too. Check Neumann KH 310 for example.

I am not talking about nearfield, midfield, farfield speakers here. Again. What I am talking about is sound power response of speakers. Checking far field estimations of nearfield speakers and building bold opinions about their nearfield performance is wrong. People check measurements but usually they do not know what those measurements are about. I am pointing out that few speaker designers make compromises on sound power response of their nearfield speakers. Sound power is not entirely relevant to determine sound quality of nearfield speakers while listening to them from nearfield distances. That's what I am saying.

As for what Toole's comment, yes we all are victims of circle of confusion. I was not talking about fixing tonal errors of recordings though. I was talking about fixing tonal issues of speakers above 700hz based on the measurements taken by microphones which have linear *hearing* unlike humans do.

 

[Rednaxela]

I was talking about fixing tonal issues of speakers above 700hz based on the measurements taken by microphones which have linear *hearing* unlike humans do.

In response to whom or what?

 

[Sokel]

I don't know much about designing speakers,unlike most of of the people here I never diyed one,but I know enough to have faith to the recommendations of their designers.
Sure can work over the recommended distances or in bigger rooms than the recommended volume.

And it would be their benefit (considering marketing) to stretch these numbers but let's hope they don't.
What you can't escape though is SPL.

And I was myself surprised when I saw peaks of 110db(Z) in normal 75-80db(A) listening (peaks not so evident when I switched amps for weaker ones just to test).
I think their recommendations has also this in mind.

I will be happy to be wrong but practice says otherwise.

 

[Rednaxela]

View attachment 254145
[…]

So your point is basically if you fix this you compromise its on-axis ruler-flatness, and because it’s the on-axis response that matters most, doing so won’t lead to an improvement. On the contrary.

Correct?

 

[Blockader]

So your point is basically if you fix this you compromise its on-axis ruler-flatness, and because it’s the on-axis response that matters most, doing so won’t lead to an improvement. On the contrary.

Correct?

Yes.