Neumann KH 80 DSP Monitor Review

[amirm]

That is true for consumer class loudspeakers like usually Harman compares (tests like Revel vs. Magnepan or B&W ) but when the level gets higher like with top of the line studio monitors and/or your use EQ to correct little imperfections then hearable problems like IMD play a bigger roller than expected.

I guess you don't consider the JBL M2 monitors that came as a result of the same research? Or the LSR series of monitors?

That aside, heaven help you if you are correcting "little imperfections" you see on frequency response graphs with a single microphone measurements. Heavy handed correction that way is a sure way to get bad sound, but pretty graphs. I don't listen with my eyes but if you do, then that may be good for you, but not me.

Having had the JBL M2s and played them with 2000 watt+ amplification power, they would scare you out of the room way before any audible distortion sets in. I will be measuring distortion after I research it more but be very careful in throwing those words around like IMD.

Recently Sean Olive and crew tried very hard to see if they could establish preference based on distortion in headphones but could not. Between two different transducers, the difference in frequency response is so large that it dwarfs other factors. Now, if you are designing a speaker and have a choice of low or high distortion drivers, then by all means choose lower distortion. That is a very different thing than selecting between two speakers with different frequency response.

 

[Juhazi]

Another possible explanation for the difference in bass spl is how NFS works vs. quasi-anechoic bass "estimation" that Neumann, Sound&Recording and others do.

In attachment is BDBS simulation of KH80 on-axis sound with ideal 4" woofer response. Sound with long wavelength wraps around the baffle which makes on-axis sound to drop (baffle loss). This is why off-axis measurements of NFS show same bass spl around the box! In normal size room s with reflective walls (unlike anechoic chambers) this wrapped sound energy gets reflected back to the listener and this gives back the baffle loss of spl! This is the main reason why room measurements shows higher bass level than direct on-axis.

With quasi-anechoic lwo bass is measured in near-field, reflex port output is added and the sum of those is lifted to simulate same distance/eq'd with theoretical baffle loss curve. This is always an estimation!

More about quasi here
http://audio.claub.net/software/FRD_Blender/White Paper - Accurate In-Room Frequency Response to 10Hz.pdf
and
http://www.aes.org/e-lib/browse.cfm?elib=14677

but we have other methods too, and articles
http://www.aes.org/e-lib/browse.cfm?elib=15593
http://www.prismsound.com/test_measure/support_subs/apps/loudspeaker_testing.php

p.s. Among loudspeaker designers, it is a matter of personal descision, how much of this on-axis baffle loss (or baffle step) is compensated in the design. The designer has an idea and often gives a recommendation of how to position the speaker regarding to boundaries - to get optimal bass level and tonal balance. Small speakers basically have the driver's cone area/excursion limiting bass spl and eq possibilities. It is alway a compromise, but the end user can use tone controls or eq within reasonable distortion limits. So, the relative spl level/slope of low bass is not a real problem per se and can be eq'd, but bass reflex tuning and distortion limits are always only the given. This is one reason why we want to see distortion measurements and to know the bass tuning!

 

[SDX-LV]

The system is not "newly released." It was introduced back in 2015 or nearly five years ago. It is specifically designed to correct for errors in anechoic chambers. It is used by who is who in the industry already. Here are published results from Harman using this system:
https://3.bp.blogspot.com/-OB4hm25d...GAs/s1600/Spin+-+Revel+Performa3Be+F228Be.png

Anyway, the difference is tiny and not material. If you don't trust it, that is cool. If you are asking me to run to Klippel, I won't do so since there is nothing actionable here. There is variation in anechoic chamber measurements as I have explained at length and quoted authorities on them.

I think you are abandoning caution in believing speaker measurements as flat as what you have reported. Such an animal does not exist and requires huge justification for any measurement that shows it such. Put your doubt there, not in my measurements.

Hi, are you sure Harman uses Klippel NFS system? To my knowledge the latest Spinoramas (gen3) from Harman come from a Klippel measurement system inside a proper anechoic chamber at Harman. Here is at least one souce: https://www.avsforum.com/forum/89-s...peaker-what-science-shows-8.html#post57359530

 

[daftcombo]

For example a close vertical surface like a table/desk creates a bump around 150 Hz

How to you call a table a vertical surface?

 

[daftcombo]

My guess is the Jbl sounds at least as good because of the more sloping in room response, and additional thump due to the larger woofer. I ve heard the KH120 it sounded good, flat but not very exciting to listen to (which is not really what you look for in a monitor anyway).

You are not alone to prefer the cheap JBL though:
https://www.gearslutz.com/board/so-...ime/1009704-neumann-kh120a-vs-jbl-lsr305.html
(FWIW)

 

[amirm]

Hi, are you sure Harman uses Klippel NFS system? To my knowledge the latest Spinoramas (gen3) from Harman come from a Klippel measurement system inside a proper anechoic chamber at Harman. Here is at least one souce: https://www.avsforum.com/forum/89-s...peaker-what-science-shows-8.html#post57359530

I do. They own one.

 

[aarons915]

I don't think any of the measurement results are invalid but I do agree with measuring every speaker with some sort of standard input, CTA 2034 specifies 2.83V in passive speakers and 79db @ 2m(averaged between 500-2000Hz) for active speakers.

 

[thewas]

Why would anyone interested in good sound be using the "nothing" option?

Why would someone do listening comparison tests on a desk without EQ when it is known that the peak around 150 Hz is dentrimental for neutral tonality?

Nothing I am doing here can rescue from bad bass if you don't invest in proper analysis and fixes for that region.

Bass modes are one thing, but at desktop placement the higher 150 Hz peak is the most serious tonal offender in my experience of such setups.

That aside, I hope you also listened to the correction to make sure it sounded better than no EQ rather than just following what Genelec said years ago.

Don't worry about me, I listen most of the time from my desktop setup where I do my own measurement based EQ but nevertheless showed before that even the pre-fixed "dumb" old Genelec filter would give a better tonal balance than without anything.

 

[Juhazi]

I do. They own one.

Since when? Not in Toole's days!

 

[thewas]

I guess you don't consider the JBL M2 monitors that came as a result of the same research? Or the LSR series of monitors?

???
I think you misunderstood my quote there, I said that tonality is the most important to preference when it comes to blind testing a la Harman, but when both "A" and "B" are there almost perfect than other important factors like IMD become the decisive factor. That's why a LSR will never be the same as a M2.

That aside, heaven help you if you are correcting "little imperfections" you see on frequency response graphs with a single microphone measurements. Heavy handed correction that way is a sure way to get bad sound, but pretty graphs. I don't listen with my eyes but if you do, then that may be good for you, but not me.

If you had read some other posts of me here like this one
https://www.audiosciencereview.com/...ut-room-curve-targets-room-eq-and-more.10950/
I am the one who started a thread with quotes of Toole that EQ above the transition region should be based on direct sound (anechoic measurements) and not listening position measurements.

Having had the JBL M2s and played them with 2000 watt+ amplification power, they would scare you out of the room way before any audible distortion sets in. I will be measuring distortion after I research it more but be very careful in throwing those words around like IMD.

??
I didn't mention the M2 as an example of high distortions or IMD anywhere, did I??

Recently Sean Olive and crew tried very hard to see if they could establish preference based on distortion in headphones but could not. Between two different transducers, the difference in frequency response is so large that it dwarfs other factors. Now, if you are designing a speaker and have a choice of low or high distortion drivers, then by all means choose lower distortion. That is a very different thing than selecting between two speakers with different frequency response.

I didn't say anything different, the Revel will always beat xxx .... due to better tonal neutrality, but when you compare a tonality class AAA+++ speaker X with another one of the same highest class, then suddenly IMD and other secondary effects become the deciding factor.

 

[thewas]

How to you call a table a vertical surface?

Sorry, meant a horizontal surface of course, my mind was again at the reflections.

 

[carlob]

I have the JBL and the Neumann side by side on stands. Level matched (not blind) and without EQ from my listening point at around 2 meters the Neumann sound vastly better. Also it's amazing if you consider the volume/size of the Neumann compared to the JBL.

 

[Juhazi]

I have the JBL and the Neumann side by side on stands. Level matched (not blind) and without EQ from my listening point at around 2 meters the Neumann sound vastly better. Also it's amazing if you consider the volume/size of the Neumann compared to the JBL.

Please open up, what are the differences? Vastly better doesn't tell anything to me...

 

[carlob]

Please open up, what are the differences? Vastly better doesn't tell anything to me...

It's a subjective opinion, even if level matched, but to me and in my room the neumann sound neutral and the JBL very bright and fatiguing. I am returning them btw.

 

[SDX-LV]

Here is the KH80 Spinorama plot scaled to match SpeakerData2034 Spins.

 

[beefkabob]

They are just too flawed IMO, and that of every engine designer I know (though most of those are only designing racing engines).

Wankels are not flawed. They just have different tradeoffs versus other ICEs. If I were to build a small, light sports car, a Wankel would be a great choice since it's so small, light, and powerful.

And therein lies some of the issues with speaker measurements. Is it designed to be placed against a wall or on a desk or on stands? With a sub or without? Does it work better in this room or that? Near field, mid field, or far field? Perfectly flat or Harman flat? Can a DSP make up for all of its flaws? How loud can it go? How well will it integrate with a sub? So many things play into it all.

 

[carlob]

I really don't care for now as speakers measurement is a much different beast from electronics measurements and I think we are in a kind of beta test for now, but if you put the speakers on a desk and if they have a freaking switch that you are instructed by the manufacturer to use when you put them on a desk I think you should measure them that way. In addition measurement wise the neumann measure better compared to JBL (and are very very very small too!).

Anyways the JBL are going back and I have a pair of Genelec incoming.

 

[Xyrium]

What about the difference in frequency response ? I thought that only the 305P mkII had the treble boost, and that the original LSR 305 was more neutral.

Unless there was a legit site performing measurements using professional gear, and knew how to use it, who then measured the two of them within the same control group, I doubt it.

Unless you're referring to anecdotal subjectivism from the likes of some "other" sites we know of. Because, I love when an individual that's never heard an instrument and only listens to electronica on a daily, makes claims that the "air around the instruments" was great with x speaker vs y speaker, even though the speakers were in different rooms, using different sources, and generally, completely unreliable controls.

 

[folzag]

if they have a freaking switch that you are instructed by the manufacturer to use when you put them on a desk I think you should measure them that way.

Both ways! But I agree. It's the analog of testing both USB & SPDIF inputs on a DAC... one doesn't necessarily imply performance of the other. Plus it would be interesting to characterize the effect of the switch.

 

[BYRTT]

Why would someone do listening comparison tests on a desk without EQ when it is known that the peak around 150 Hz is dentrimental for neutral tonality?


Bass modes are one thing, but at desktop placement the higher 150 Hz peak is the most serious tonal offender in my experience of such setups.


Don't worry about me, I listen most of the time from my desktop setup where I do my own measurement based EQ but nevertheless showed before that even the pre-fixed "dumb" old Genelec filter would give a better tonal balance than without anything.

Seems both DUT had same enviroment conditions was it the subjective or objective tests so why bother, else catch amirm's words into post #159 below.

...Anyway, this getting frustrating on both listening side and measurements. Listening tests are informal and just shared for what it is. And measurements have to be taking in larger context, and not focusing on half a dB difference here, and there. It is the big picture that matters.