Neumann KH 80 DSP Speaker Measurements: Take Two

[hyperplanar]

FWIW, my own measurements of a KH120 I had years ago (now long lost) also showed a depression in the presence region, which actually filled out somewhat (but not completely) slightly off-axis (below axis IIRC), suggesting a time misalignment between the drivers.

Did you measure on the axis of the tweeter? The acoustic axis is the midpoint between the woofer and tweeter's centers. The tweeter is also recessed, presumably to time align it with the woofer. This way the main acoustic lobe would point straight forward.

 

[andreasmaaan]

Did you measure on the axis of the tweeter? The acoustic axis is the midpoint between the woofer and tweeter's centers. The tweeter is also recessed, presumably to time align it with the woofer. This way the main acoustic lobe would point straight forward.
View attachment 64189

I incorrectly measured on the tweeter axis. The measurement distance was 1.5m though, so even accepting that the correct axis is the midpoint, my measurement could only have been around 3.5° off-axis, whereas IIRC it was more like 10 or 15° below the tweeter axis that I obtained the flattest response.

As I said though, I unfortunately don't have the measurements any longer. I raised them only to point out that the Klippel measurements are not alone in failing to correlate with the Rockwell/S&R measurements.

Anyway, given my own measurements exist now only in my memory, they are not of any independent value ofc.

 

[hyperplanar]

I incorrectly measured on the tweeter axis. The measurement distance was 1.5m though, so even accepting that the correct axis is the midpoint, my measurement could only have been around 3.5° off-axis, whereas IIRC it was more like 10 or 15° below the tweeter axis that I obtained the flattest response.

As I said though, I unfortunately don't have the measurements any longer. I raised them only to point out that the Klippel measurements are not alone in failing to correlate with the Rockwell/S&R measurements.

Anyway, given my own measurements exist now only in my memory, they are not of any independent value ofc.

Ah, gotcha. Yeah, at 1.5m the difference would be very minimal. Measuring on the tweeter axis doesn't change the FR much for me either, past 60 cm or so. However, I don't get a dip in the crossover range unless I measure at some pretty wild angles...

I'm befuddled as to why Rockwell's, S&R's, and even my own in-room measurements show a linear response, while the one in the Klippel slides (and yours as well) has a huge dip in the upper midrange.

 

[andreasmaaan]

Ah, gotcha. Yeah, at 1.5m the difference would be very minimal. Measuring on the tweeter axis doesn't change the FR much for me either, past 60 cm or so. However, I don't get a dip in the crossover range unless I measure at some pretty wild angles...

I'm befuddled as to why Rockwell's, S&R's, and even my own in-room measurements show a linear response, while the one in the Klippel slides has a huge dip in the upper midrange.

Yeh, it's odd. I've spent years occasionally puzzling over why mine measured worse than most published measurements too.

FWIW, and ofc only based on recollection, my gated measurement at 1.5m on the tweeter axis looked extremely similar in both frequency, depth and Q to this on-axis Klippel measurement (as opposed to the previously posted screenshot which showed a broader, shallower dip):

However, I don't recall measuring a second little dip at around 7kHz, or a rising response above 10kHz, as apparently measured by Klippel.

 

[hyperplanar]

Yeh, it's odd. I've spent years occasionally puzzling over why mine measured worse than most published measurements too.

FWIW, and ofc only based on recollection, my gated measurement at 1.5m on the tweeter axis looked extremely similar in both frequency, depth and Q to this on-axis Klippel measurement (as opposed to the previously posted screenshot which showed a broader, shallower dip):

View attachment 64192

However, I don't recall measuring a second little dip at around 7kHz, or a rising response above 10kHz, as apparently measured by Klippel.

The plot thickens... Schrodinger's upper mids apparently? Simultaneously there and not there
Man, I wish one would get into Amir's hands already for measurements.

I do think there's something to the sound power dip though. I have to say the KH120 has a slightly "polite" character to it.

 

[andreasmaaan]

The plot thickens... Schrodinger's upper mids apparently? Simultaneously there and not there
Man, I wish one would get into Amir's hands already for measurements.

I do think there's something to the sound power dip though. I have to say the KH120 has a slightly "polite" character to it.

Haha, yeh that makes sense I think.

Actually, I'd describe the sound of mine as dull. For that reason, it was actually kind of a relief to measure them and find the dip there.

I also felt there was something a bit off about the bass alignment, the bass always sounded somewhat wrong to me too. I guess it may have been that midbass dip + port tuning peak.

OTOH, I must admit that my memory of how they sounded is almost certainly less reliable than my memory of how they measured. And they were used in a bedroom studio setup, so under far from ideal conditions.

 

[hyperplanar]

Haha, yeh that makes sense I think.

Actually, I'd describe the sound of mine as dull. For that reason, it was actually kind of a relief to measure them and find the dip there.

I also felt there was something a bit off about the bass alignment, the bass always sounded somewhat wrong to me too. I guess it may have been that midbass dip + port tuning peak.

OTOH, I must admit that my memory of how they sounded is almost certainly less reliable than my memory of how they measured. And they were used in a bedroom studio setup, so under far from ideal conditions.

Was there a desk reflection involved? I push mine all the way to the front of the desk because they sound like crap with the desk reflection LOL...

I also agree on the bass, but I think it's usually mostly the room acoustics at play. I carried these little dudes through maybe three different apartments in college and the bass never really sounded right. But after I moved back home and was able to set up my desk in a very large open area, I think the bass actually sounds pretty decent unequalized! However, their response falls off a cliff below 45 Hz, and there's very little headroom, like I think you mentioned earlier. It's easy to activate the limiters when you crank up the volume on music with any sort of bass with these, and I can hear the IMD increasing well before that point.

These totally transform once you add a sub though! They can play at 100 dB SPL at all frequencies above 100 Hz at 1% distortion which is pretty nuts for a 5.25" woofer.

 

[thewas]

More third party KH120 measurements show at least a little low-mids hump:

(dark curve with LOWMID= 0 and Treble= +1 and red curve LOWMID= -3 and Treble= +1)
source: http://www.hifi-forum.de/viewthread-30-13717-630.html#31709

source: https://heissmann-acoustics.de/dxt-mon-vs-neumann-kh-120a/

 

[andreasmaaan]

Was there a desk reflection involved? I push mine all the way to the front of the desk because they sound like crap with the desk reflection LOL...

I also agree on the bass, but I think it's usually mostly the room acoustics at play. I carried these little dudes through maybe three different apartments in college and the bass never really sounded right. But after I moved back home and was able to set up my desk in a very large open area, I think the bass actually sounds pretty decent unequalized! However, their response falls off a cliff below 45 Hz, and there's very little headroom, like I think you mentioned earlier. It's easy to activate the limiters when you crank up the volume on music with any sort of bass with these, and I can hear the IMD increasing well before that point.

These totally transform once you add a sub though! They can play at 100 dB SPL at all frequencies above 100 Hz at 1% distortion which is pretty nuts for a 5.25" woofer.

View attachment 64194

You make good points.

I wasn't enamoured of the sound of these much at all, so I didn't persevere too much with setup, EQ etc, and got rid of them pretty quickly, especially given the apparent misalignment between the drivers and the very limited output in the bass before distortion/limiting. Another factor was that the enclosure or the port would rattle slightly at higher SPLs.

FWIW though, I did try setting them up on a desk, on stands, near the front wall and away from walls. I also tried EQing them to compensate for the upper-mid dip. Couldn't manage to get them to sound right to my ears, though as I said, I didn't persevere for too long.

 

[hyperplanar]

You make good points.

I wasn't enamoured of the sound of these much at all, so I didn't persevere too much with setup, EQ etc, and got rid of them pretty quickly, especially given the apparent misalignment between the drivers and the very limited output in the bass before distortion/limiting. Another factor was that the enclosure or the port would rattle slightly at higher SPLs.

FWIW though, I did try setting them up on a desk, on stands, near the front wall and away from walls. I also tried EQing them to compensate for the upper-mid dip. Couldn't manage to get them to sound right to my ears, though as I said, I didn't persevere for too long.

Yeah, that's totally reasonable. At the end of the day if a speaker isn't working out for you there's no reason to keep it! I'm curious about the rattling sound you heard, I haven't heard any rattling or port chuffing even once for the five years I've had these, with no sub and driven to limiting as well... In fact, the ports make for great fans in the summer without a sub LOL.

The KH120s are by no means perfect, and I think I would replace them with Genelec 8341s in the long run (saving up my money for a certain sports car first ). But I had a really hard time finding anything better in the $1000-1500 range five years ago. I wonder if there's something better in the price range now.

 

[andreasmaaan]

Yeah, that's totally reasonable. At the end of the day if a speaker isn't working out for you there's no reason to keep it! I'm curious about the rattling sound you heard, I haven't heard any rattling or port chuffing even once for the five years I've had these, with no sub and driven to limiting as well... In fact, the ports make for great fans in the summer without a sub LOL.

The KH120s are by no means perfect, and I think I would replace them with Genelec 8341s in the long run (saving up my money for a certain sports car first ). But I had a really hard time finding anything better in the $1000-1500 range five years ago. I wonder if there's something better in the price range now.

I think I got a dud pair TBH. The port rattle was only in one but not the other, which suggests it was a bad sample. In retrospect, I could have probably asked for a replacement under warranty, but my attention was mostly on other things at the time...

Anyway, I'm sure your KH120s perform better than mine did. It was always a bit of an enigma to me why mine measured/sounded worse than others seemed to. Given the apparent similarity to the Klippel measurement that I've only just seen for the first time now, I wonder if maybe they changed something at some point to improve the crossover alignment and that both Klippel and I had earlier, more poorly-performing iterations. Or perhaps there are just some consistency issues with this speaker in general.

BTW, Genelec 8341 would be an excellent upgrade I would think

 

[hyperplanar]

Anyway, I'm sure your KH120s perform better than mine did. It was always a bit of an enigma to me why mine measured/sounded worse than others seemed to. Given the apparent similarity to the Klippel measurement that I've only just seen for the first time now, I wonder if maybe they changed something at some point to improve the crossover alignment and that both Klippel and I had earlier, more poorly-performing iterations. Or perhaps there are just some consistency issues with this speaker in general.

This is anecdotal, but one of my KH120s is a random used unit from Guitar Center, and the other is a Sennheiser refurbished unit. The two pretty much measure the same, so I figured that meant that the production consistency is really good. Their serial numbers are very far apart too. I guess it's possible I got lucky, but FWIW Neumann actually has a consistency spec on their website:

 

[andreasmaaan]

This is anecdotal, but one of my KH120s is a random used unit from Guitar Center, and the other is a Sennheiser refurbished unit. The two pretty much measure the same, so I figured that meant that the production consistency is really good. Their serial numbers are very far apart too. I guess it's possible I got lucky, but FWIW Neumann actually has a consistency spec on their website:

View attachment 64200

Very interesting. I also find it hard to imagine Neumann has consistency issues.

 

[LeftCoastTim]

Very interesting. I also find it hard to imagine Neumann has consistency issues.

My thought is that Neumann, who seems to really care about measurements and consistency, have not-so-minor a gap between "what they say (marketing)" and "what is observed in the field (independent Klippel 2034)". If that's true of Neumann, I have zero faith in other manufacturers.

Furthermore any review of professional studio monitors that doesn't include a Klippel 2034 is useless to me. I would even go further and say that I will never buy a set of speakers that I could not find a reliable 2034 measurement.

 

[andreasmaaan]

This is interesting. I just searched the forum for my own posts on this topic as I remembered mentioning it in another thread a couple of years ago. This is what I said back then:

@ping I had a pair of the Neumann KH120 at one point and was not impressed. They measured worse than the graphs provided by Neumann, with a 4-5dB dip at the crossover point that was mitigated only if the mic was placed (IIRC) about 15° vertically off-axis, suggesting the crossover was poorly implemented. Also, the cabinets were quite resonant (rattly even, on one of them). I may have got a bad pair, but that was my experience.

That was still ofc many years after I actually owned/measured them, but at least my recollection seems to have remained consistent.

Furthermore any review of professional studio monitors that doesn't include a Klippel 2034 is useless to me. I would even go further and say that I will never buy a set of speakers that I could not find a reliable 2034 measurement.

I'd generally be happy with any reasonably comprehensive set of reliable measurements, although ofc it's not easy without either Klippel or a good anechoic chamber. But otherwise, I totally agree.

 

[LTig]

I think it'd be wise to move all those KH120 postings into a separate thread. People may overlook this and think you're talking about the KH80.

 

[napilopez]

From the many measurements I've seen of the KH120 @andreasmaaan I think either you got a bad pair, or the KH120 are one of those speakers with a super-sensititive vertical listening window, like the buchardt S400. That said, they do seem to have a dip in the power response so that could've been what you were hearing too

More third party KH120 measurements show at least a little low-mids hump:

View attachment 64197
(dark curve with LOWMID= 0 and Treble= +1 and red curve LOWMID= -3 and Treble= +1)
source: http://www.hifi-forum.de/viewthread-30-13717-630.html#31709

View attachment 64198
source: https://heissmann-acoustics.de/dxt-mon-vs-neumann-kh-120a/

Yeah, Ken Rockwell also mentioned a bit of a mid-bass hump and that he preferred them with the Low Mid setting at -2.5dB.

With low-mid set to -2.5dB (note taller vertical scaling):

My thought is that Neumann, who seems to really care about measurements and consistency, have not-so-minor a gap between "what they say (marketing)" and "what is observed in the field (independent Klippel 2034)". If that's true of Neumann, I have zero faith in other manufacturers.

Furthermore any review of professional studio monitors that doesn't include a Klippel 2034 is useless to me. I would even go further and say that I will never buy a set of speakers that I could not find a reliable 2034 measurement.

This seems a bit unfair. Neumann speakers are some of the most measured out there and so far the klippel measurements have been the odd one out... This could be for any variety of reasons and I'm not doubting utility of the system, but I do not consider them valid over multiple anechoic measurements and well-done DIY measurements, just another valuable set of data.

. That would suggest the Klippel is capturing something that can't be captured under anechoic conditions, in which case there are a whole lot of speakers that need to be redesigned...

That said, I wish Neumann had responded to this thread.

EDIT: also don't know if you saw this @LeftCoastTim but when @BYRTT corrected Amir's measurements for Mic calibration, mic cage deviation(this review was before Amir fixed that), and reference axis, he got a result that was much closer to the promised response, basically +/- 1dB from 120Hz to 16kHz:

The bass remains an issue but I'm okay with that as it is room dependent.

Thanks for the link. Huh. I'm intrigued! I'm curious why the dip here is narrower and nearly 5 dB deep, while the other graph has a much broader 2.5 dB dip.
View attachment 64183

Not sure if you watched the webinar or just looked at the slides . This measurement is not meant to represent the response at a normal listening distance, it is the response as measured at 10cm. In this section of the webinar (circa 35 mins here), he is making the point that even measured at 10cm you still get room influence (the green line), which is why the klippel's direct sound separation tech is important. He wasn't demonstrating the klippel's ability to extrapolate far-field response from nearfield measurements. We don't even know if the speaker set to its 'flat' tuning as that was not the purpose of the demonstration here.

Edit: For those who can't/ don't want to watch:

"Here you see the result in one measurement point... We still take benefit of the nearfield measurement, though the nearfield sound pressure is 20dB larger than in 1 meter distance, but still, you see that's not sufficient. We end up with this green line measured in 10cm distance, we have still some deviations of +/- 5dB."

 

[andreasmaaan]

From the many measurements I've seen of the KH120 @andreasmaaan I think either you got a bad pair, or the KH120 are one of those speakers with a super-sensititive vertical listening window, like the buchardt S400. That said, they do seem to have a dip in the power response so that could've been what you were hearing too

Could be. In my case, the outputs did sum correctly around 15° off-axis however, which is why I don't think it was as simple as me just having got the measurement axis wrong by a few cm. I wish I could measure the speaker again now.

In fact, I'm now starting to wonder whether maybe the factory had wired one of the drivers in reverse. Given the C2C distance on the KH120 and the XO frequency of approximately 2.5kHz IIRC, 15° off-axis would be about where you'd expect the drivers to sum (closest to) correctly if the crossover were wired in reverse. Hmm....

Anyway, gonna shut up about the KH120 now. Sorry all.

 

[Andreas007]

I‘ve never measured my KH120.
But in two different rooms I pulled down the bass by -2.5 dB, the mids by -4.5 dB and increased treble by +1 dB.

This gives me the most „natural“ sound. Otherwise they sound a bit dull/fat and unexciting. However, I abuse them listening in 2.5 m to 3.0 m distance.

Some guys here measured them (also with greater distance) and basically came to the same conclusion:

https://www.hifi-selbstbau.de/hifi-selbstbau-mainmenu-32/forentreffen/469-der-kleinlautsprecher-test

They also mention the 2 kHz dip which seems to make them a bit too polite in the far field.

 

[napilopez]

Personally I think this is a useful discussion - not like there's much else to be said about the KH80 at this point that hasn't been said across two separate threads, and it's interesting to see how the KH120 might measure =]

@TimVG Where in the webinars did you see it was the KH120 for the CEA2034 image you posted? I mean it certainly seems like it, but they don't identify it as the KH120 during the session. I'd be a little sketpical of this graph as, once again, it doesn't seem like they were presenting it as an accurate representation of the speaker's response, but rather demonstrating the NFS' features. Not to mention the slide showed several wrong points for the early reflections line lol.

I‘ve never measured my KH120.
But in two different rooms I pulled down the bass by -2.5 dB, the mids by -4.5 dB and increased treble by +1 dB.

This gives me the most „natural“ sound. Otherwise they sound a bit dull and unexciting. However, I abuse them listening in 2.5 m to 3.0 m distance.

Some guys here measured them (also with greater distance) and basically came to the same conclusion:

https://www.hifi-selbstbau.de/hifi-selbstbau-mainmenu-32/forentreffen/469-der-kleinlautsprecher-test

They also mention the 2 kHz dip which seems to make them a bit too polite in the far field.

Thanks; I've come to realize I really do not like 2kish dips in the in-room response. It's also just about where the interaural crosstalk dip will happen if you listen in a typical living room setup (slightly 'longer' than an equilateral triangle).