Neumann KH 80 DSP Speaker Measurements: Take Two

[amirm]

If it's feasible financially, what about raising funds to buy @amirm a KH 80 off the shelf for testing and having the unit measured by Klippel (they do offer measurement as a service)? No second thoughts about golden sample then and Klippel sure will be more interested in demonstrating their product produces reliable and repeatable data than Neumann.

Could also be done with any other speaker @amirm has already measured.

PS: If you need a guy in Germany to deal with Klippel formally I'd help.

Thanks. No need for contribution but I do appreciate it a lot. Just ordered a KH80. By the time we are done, it will be the most measured speaker in the world!

 

[amirm]

Sure, only Klippel can solve it, but they won't raise their little finger unless Amir contacs them expressing his concern of LF innacuracy, and it is hard for him to do that without measuring identical speaker on 2 NFS'es. So, as long as Amir doesn't have a solid case that LF part of the measurement is not ok things will stay as they are.

I did reach out to entire Klippel team (both US and Germany including the architect of the NFS) when my Revel F35 measurements did not match Harman's. I shared my full project file with them for analysis. They could not find any issue with it that would explain the discrepancy. They firmly stood behind my measurements being correct in that regard.

I have also repeatedly requested Harman to send me a sample speaker they have measured in their chamber to test here. Despite telling me they could do that, there has been no progress. This has been through three channels of communication with them including people who I considered long time friends and colleagues.

Seeing how there are two different measurements from Harman for some of their speakers from anechoic chamber, we have to accept that getting to the bottom of such things ultimately may not be easy or convenient for some.

Fortunately none of this matters in grand scheme of things as I have repeatedly said. And stipulated in CEA/CTA-2034 standard:

 

[amirm]

@amirm @GuyLayfield From my point of view the start was not the best. I hope that you will be talking with another.

There is no emotion in anything I do here. Look at how I test and highly recommend Schiit audio products after a very poor response from them to my early testing. I owe it to you all to not put any barriers between you and the truth about audio performance. It also would not be professional for me to hold grudges against the industry participants.

What I won't do is go begging beyond what I have done here. Company knows where I am, how to contact me and if they choose not to, that is that. I have plenty of other products to test including speakers from their competitors which have been offered by them for testing.

 

[Ron Texas]

It's all well and good, but the KH80 has a 4" mid/low which limits it's output even with a sub. Of course, it's a kick ass desktop or mixing console piece of gear.

 

[amirm]

It's all well and good, but the KH80 has a 4" mid/low which limits it's output even with a sub.

That's indeed a problem for someone like me who craves high SPL. No amount of fiddling with the measurements will fix that.

 

[infinitesymphony]

It's all well and good, but the KH80 has a 4" mid/low which limits it's output even with a sub. Of course, it's a kick ass desktop or mixing console piece of gear.

Exactly, these are near-fields meant for ~92 dB max without a sub, maybe a bit higher when you add a sub and cross them over. Not going to be enough for some people, but when you add the manufacturing consistency, flat frequency response, and the built-in DSP they represent a pretty good value to anyone looking for a neutral starting point. If you need more volume and can sacrifice a bit of pre-DSP frequency response accuracy, the KH120 has a much higher SPL bass output in comparison (+13 dB). Add a KH 750 DSP sub and you'll have DSP with the KH120s.

 

[napilopez]

@GuyLayfield, I hope you know your silence here is very concerning to your customers (like myself) and this community as a whole.

Please do not make me eat my words where I stated how much a Neumann fan I am, in response to your initial post (which contains very helpful data!) I had assumed you would be sending your test sample to @amirm for testing as well, since you already apparently went to great lengths to send it to Klippel.

You should know that if you refuse to send this sample to Amir, this changes the tone and impact of your measurements contribution drastically, in contrast to what I had assumed would be the start of a wonderful collaboration.

If you cut off ASR from this experiment, this would demonstrate that — while you are willing to go to great lengths to make Neumann look better, you refuse to put in an additional 1% of effort required to turn this experiment into something that benefits the community far more (by simply shipping the sample to Amir so he can work with Klippel to debug any discrepancies). I really hope this is not the case.

Cost is not a factor here; many of us here would gladly donate the cost of shipping. For the record, I would gladly do so myself, if shipping cost were somehow the only concern here.

While measurement data contributions are always helpful for consumers in a broad sense, it is not helpful to this community (to which you posted) if you perform this experiment with Klippel, and cut ASR out of the equation. What ASR is doing here is a vastly more important contribution to the audiophile and pro audio world than any one set of measurements for any one manufacturer could ever be.

So please understand that while your contribution of measurements individually is much appreciated, posting here on ASR does add some implicit responsibilities on your part to contribute constructively to this community, if you want to maintain a positive relationship with your customers (like myself).

While I get where you're coming from, do keep in mind that it's only been a few days. I don't know exactly how neumann operates, but the company has been around for nearly a hundred years and is owned by Sennheiser, which has thousands of employees, so I wouldn't be surprised if there's at least some red tape involved. I'd imagine that's why the initial response took so long in the first place, in addition to, you know, global lockdowns.

Personally, I'm glad Amir bought his own - imo otherwise doubts over golden samples would continue.

I also think it's worth reminding everyone that at this point the only meaningful difference is the bass and maybe the highest highs. Most of the differences were differences in reference axis, microphone cage, and mic calibration.

I don't know if you'd seen this @echopraxia, but a while back @BYRTT had used VituixCAD to apply corrections for those issues to the measurements of the KH80 in the OP of this thread. He got this:

Not sure of the cause of the drooping highs, but overlaid and scaled over the new Klippel measurements, it's close enough that the only issue that remains is the bass.

Between 100hz and 10khz, I'd say that's less than +/- 1db of difference overall.

(Note that VituixCAD uses the early reflections fix, so its ER naturally looks a little difference).

Now, the reason I care about the bass is that that the NFS is supposed to be more accurate than an anechoic chamber in this regard. So seeing the discrepancy between Klippel and Amir's results is interesting.

When Amir published his NFS measurements and got the same results twice, I assumed I screwed up my nearfield bass summations and then spent some time trying to get my bass measurements to look the same (different splice points, tweaking summations, etc), but I couldn't.

So I assumed it was a limitation of the technique. Then I was very surprised when Neumann posted their and klippel's results with basically the same bass contour as my own. I'm glad we're getting to the bottom of this.

 

[QMuse]

It's all well and good, but the KH80 has a 4" mid/low which limits it's output even with a sub. Of course, it's a kick ass desktop or mixing console piece of gear.

Here is what they say in spec, not bad at all for such small speaker:
Max. SPL in full space at 3% THD at 1m (averaged between 100 Hz and 6 kHz): 102,8 / 108,8 dB SPL

Amir's THD measurement confirms that distortion is low down to 80Hz but it would probably be better to cross it with sub at 100Hz.

 

[BYRTT]

Thanks. No need for contribution but I do appreciate it a lot. Just ordered a KH80. By the time we are done, it will be the most measured speaker in the world!

A big hug and thanks alot add a sample four into the story... agree napilopez for finale it looks the most interesting part left is the low end contour as seen in below animation where there is added some missing known calibrations to published objective analyze of those three different samples of a KH-80 speaker, as pointed out via simulation over in F208 thread my own guess is still than somewhere into Amir's Klippel interface there's one setting that needs to be changed so computation use minimum phase instead of linear phase when software compute the sum of a inverted output as is case for ports to non inverted output from the direct radiators and we can see in below that the Klippel service Neumann used in Germany can get the low end reach hotter and smooth but okay lets see what happens using a new sample four measured a few hundreds times from all over the world ...

 

[BYRTT]

Way cool. I might try this one of these days although I don't have much open space where I live. How far away does the closest surface (apart from the ground) need to be for accurate results to say 200Hz?

That is some constructive help so thanks alot, also it move the work burden a bit not to always lay on Amir's shoulders

 

[Ron Texas]

Here is what they say in spec, not bad at all for such small speaker:
Max. SPL in full space at 3% THD at 1m (averaged between 100 Hz and 6 kHz): 102,8 / 108,8 dB SPL

Amir's THD measurement confirms that distortion is low down to 80Hz but it would probably be better to cross it with sub at 100Hz.

I cross my LS50's at 100 hz to a pair of subs. At 100 hz you need two or more subs to avoid locating the subs. The KH80 likely needs 120 hz xover. Like it or not, KH80's do not play loud and that's a law of physics. IMO, it was designed to sit on a mixing console rather than typical hi fi use.

 

[q3cpma]

I cross my LS50's at 100 hz to a pair of subs. At 100 hz you need two or more subs to avoid locating the subs. The KH80 likely needs 120 hz xover. Like it or not, KH80's do not play loud and that's a law of physics. IMO, it was designed to sit on a mixing console rather than typical hi fi use.

Yes, the KH80DSP definitely isn't the same "beast" as the KH120. Both were made for nearfield use, but the midbass of the KH120A is miraculously loud compared to its smaller sibling. You can definitely get most out of it when using the KH750DSP, that brings a fullrange of 95-100 dB; still kind of reserved for nearfield.
Is this me or the "variable crossover frequency" feature got removed from the KH750DSP's marketing?

SPL at 1%, 3% and 10% THD - KH120A, KH80DSP and KH750DSP:

 

[ernestcarl]

I cross my LS50's at 100 hz to a pair of subs. At 100 hz you need two or more subs to avoid locating the subs. The KH80 likely needs 120 hz xover. Like it or not, KH80's do not play loud and that's a law of physics. IMO, it was designed to sit on a mixing console rather than typical hi fi use.

I xo KH120 at 120Hz myself. Prefer it that way... and this completely eliminates the problem of blasting air in your face/neck at close listening distance due to the front ported design. If at all possible, putting the sub between your L&R speakers* can fix the sub localization problem.

*presuming one has a highly symmetrical room & setup.

 

[QMuse]

I cross my LS50's at 100 hz to a pair of subs. At 100 hz you need two or more subs to avoid locating the subs. The KH80 likely needs 120 hz xover. Like it or not, KH80's do not play loud and that's a law of physics. IMO, it was designed to sit on a mixing console rather than typical hi fi use.

Assuming specs are telling the truth 108.8 dB is loud enough for me and even most of the large floorstanders don't go but a few dB over that. Btw, max SPL spec for LS50 wireless is 106dB.

 

[Ron Texas]

I xo KH120 at 120Hz myself. Prefer it that way... and this completely eliminates the problem of blasting air in your face/neck at close listening distance due to the front ported design. If at all possible, putting the sub between your L&R speakers* can fix the sub localization problem.

*presuming one has a highly symmetrical room & setup.

Just an observation, but you are the 5th to quote my post. The subject of sub crossover points for small speakers is a worthy one though. I have tried all the way from 80 hz to 120 hz with LS50's. Like the KH120 it has a 5" mid/low driver. 100hz seems to work best, but I have not tried higher since I moved everything out from the wall 2 weeks ago. It does sound good the way it is.

 

[Ron Texas]

Assumeing specs are telling the truth 108.8 dB is loud enough for me. Btw, max SPL spec for LS50 wireless is 106dB.

108 is peak. The frequency range is 100 to 6k and I don't know what KEF is using for it's measurement. I doubt that it is comparable and 106 from KEF is definitely not peak.

 

[QMuse]

108 is peak. The frequency range is 100 to 6k and I don't know what KEF is using for it's measurement. I doubt that it is comparable and 106 from KEF is definitely not peak.

108.8dB is not peak but averaged max SPL level in the 100-6000Hz range with THD lower than 3%. I have no reason not to trust either Neumann nor KEF specs so I'd say LS50 max SPL is 106dB in that same range as they certainly can't play louder than that below 100Hz.

Btw, Neumann max SPL spec is much more specific/detailed than KEF's as KEF doesn't specify neither the frequency range in which max SP is achived nor it specifies the distortion level at max SPL.

 

[ernestcarl]

Just an observation, but you are the 5th to quote my post. The subject of sub crossover points for small speakers is a worthy one though. I have tried all the way from 80 hz to 120 hz with LS50's. Like the KH120 it has a 5" mid/low driver. 100hz seems to work best, but I have not tried higher since I moved everything out from the wall 2 weeks ago. It does sound good the way it is.

Really? LOL. I have not counted so had no idea until you pointed it out. But, yes, a higher crossover does help the “mid-woofer” — this is esp. noticeable at higher listening levels. The recommendation has always been 80Hz even with Neumann’s own literature though — and they did give some reasons why...

 

[Habu]

Is this me or the "variable crossover frequency" feature got removed from the KH750DSP's marketing?

NEUMANN KH750 DSP : Crossover frequency: 80 Hz fixed or 60 - 100 Hz via Neumann.Control app
Maybe legal requirement for those who don’t use the app.

 

[ernestcarl]

... legal requirement for those who don’t use the app.

What an irritating limitation for a sub that costs that much.