Neumann KH 80 DSP Speaker Measurements: Take Two

[bobbooo]

Taking the average seems reasonable in this instance, though for the Harbeth I think the updated information is more primary.

The way I've dealt with this before was to have links in the speaker table to the precise post in the review thread where @MZKM gives this data. I think to avoid having him do more manual work than necessary, like the repost I asked for in this instance, I'll take out those links from now on and refer to his tables only like you suggested.

Just remember that I won't know of updates unless I'm notified or I happen to read the thread.

Yeah unlike the second measurement of a different Neumann unit, the new Harbeth score is based on a higher accuracy recalculation of the same speaker's measurements, so it makes sense to just use the new score. Thanks for all the work producing the interactive speaker and electronics tables by the way.

 

[BYRTT]

Hi @amirm or owner of sample 2, come to think about what if dome on tweeter by accident has been poked and diy repaired because sample 2 difference to sample 1 else than SPL was lowered and NFS points upped to 1000 for sample 2 is all about the highest octave sector, then had a zoomed look at KH 80 and could see it has grill also for tweeter so dome should be protected other than if grill is or has been poked in past there's a chance dome has hit grill and changed its original form or grill is out of its original form, absolut fine if you ignore post amirn and wan't to look forward get sweep production up in speed thanks, on the other hand it takes you or owner few minuttes to inspect and could release or not manufacture from variances in quality is reaching end user.

 

[amirm]

There was no physical damage that I could see. The angle of the speakers relative to mic though is very tricky. These speakers don't have a flat face to align them or anything. So it could still be some difference in the presented angle for on-axis.

 

[BYRTT]

There was no physical damage that I could see. The angle of the speakers relative to mic though is very tricky. These speakers don't have a flat face to align them or anything. So it could still be some difference in the presented angle for on-axis.

If speaker angle to microphone is very tricky because NFS is working so super close to that waveguided surface, then i share below because printed it couple of days ago as a fun and learning study to myself, focus on line chart spins 12kHz and up sample 1/2 differences and assess because you are the science specialist and acoustic educated if that super close NFS can be a problem sample to sample is declared quality variances, probably not but think no harm in a relook and rethink, thanks taking your time.

 

[edechamps]

Your visualizations are really interesting @BYRTT, thanks for taking the time to make those. I'm intrigued by the fact that, on sample 2, lowering the acoustic axis seems to make the speaker significantly flatter between 2 kHz and 10 kHz. However, if I remember correctly, NFS measurements have lower resolution off-axis than on-axis, so maybe what we're seeing is merely the response being smoothed.

 

[BYRTT]

Happy visualizations are interesting, think its a treasure of information/variables amirm share in those spins.

@edechamps thanks input and that NFS probably resolute bit higher on axis, about lowering the acoustics axis think in Spinorama graph 2-10kHz area is also helped smooth sample 1, in post 315 sample 1/2 is overlaid and think confirm 2-10kHz area is where both the 2 samples is probably most coherent but have to admit sample 2 there in post 315 is offset +0,3dB to normalize they were analyzed with 20 days offset and whatever this shift in days brought of experience for NFS operational use such as SPL level for sample 2 was lower and the shared spin data for sample 2 is not take 2 or 3 but think its take 4 you know the one upped to 1000 points : ) about 52mm lower manufacture suggested listening axis omit or filter out the tweeter overshoot mostly seen in ridges at 0/-10/+10deg is probably a design decision or compromise, think in there is DSP power build in they could also had designed for flat on axis on tweeter axis to omit that tweeter overshoot NFS reveal on tweeter axis but probably it gave some other drawbacks that they decided on that -52mm point in space.

 

[Pio2001]

Maybe during the measurement taken on-axis, a bug landed on the microphone's tip

 

[BYRTT]

Ha ha really good humor there from France thanks, honestly Neumann's microphone tip also seemed bit buggy when they dialed in that low end ported knee, a microphone sensitive to proximity effect could have a chance exactly realign that graph, but more realistic reason is probably it was coincidentally monday.

Could look there would be a little tendency ASR scanner will get old on all those retakes to confirm weird low end reach for ported designs.

 

[Pio2001]

But the Adam S2V measurement shows nothing wrong in the low frequency response.

 

[BYRTT]

Thanks since yesterday had some other duties to think about so probably 2 to 4 pages behind over there, can you brief me why nothing is wrong, will say even LS50 was not under same critic in low end it had that disapointing was it 4dB shelve from about 160Hz area down to ported corner.

 

[Pio2001]

can you brief me why nothing is wrong

People had been complaining that the measurement of the KH-80 was not flat in low frequencies while it should have been.
What I'm telling is that it is unlikely to be a problem with the measurement, since the Adam S2V has a flat measured frequency response in low frequencies.

 

[eliash]

We had a saying during my time in german RD&E:
Wer viel misst, misst Mist*...
...I wish I had that kind of freq. response in my listening room...
Anyway, a great experience participating in all those "audio science" findings!

*If you measure a lot, you measure crap

 

[BYRTT]

People had been complaining that the measurement of the KH-80 was not flat in low frequencies while it should have been.
What I'm telling is that it is unlikely to be a problem with the measurement, since the Adam S2V has a flat measured frequency response in low frequencies.

Ahhh..thanks man was much too slow to get it, yes you absolut right S2V is smooth up to that 500Hz surprise.

Dirty quick blues of bass reach mini overviev, thanks amirm:

001-a_JBL_305P_Mark_ii
001-b_JBL_Control_1_Pro (has not a released zip folder with spin data)
002_NHT_Pro_M-00 (has not a released zip folder with spin data)
003_Kali_Audio_IN-8
004_Revel_C52 (has not a released zip folder with spin data)
005_Yamaha_HS5
006_Neumann_KH_80
007_JBL_One_Series_104
008_Harbeth_Monitor_30
009_KEF_LS50
010_Emotiva_Airmotiv_6s
011_Selah_Audio_RC3R
012_Micca_RB42
013_Realistic_MC-1000
014_Pioneer_SP-BS22-LR
015_Dayton_Audio_B652-AIR
016_ADAM_S2V_Studio_Monitor

 

[Amateur2020]

Hello,
There is a speaker WOXTER DL-410 which had many good critics in some threads in spain. It is also 4" but it costs about 50 €. It would be great to compare it with this KH80 even if it is no DSP driven.

 

[BYRTT]

Hi @amirm, @MZKM,

Have a appeal you look below two investigations of the spins so far, thanks in advance for spending your time and hope you agree arguments below show some sign its worth reassess few data.

First investigation is there looks be a pattern in spins for the highest octave when points is lower or a manual recompute will benefit what probably was a automatic assessment of data in Klippel software, pattern is better focused up span to 80dB range from the usual 50 number. Compare upper two graphs of Harbeth take 1 where it looks clear how the original at left have a vice versa U the more off axis we get in highest octave, when recomputed amirm's manual settings we see at right side that area improve be better covered/resoluted to what we probably can call kind of a diffuse pattern in 500 point category, diffuse because it looks that for other 1000 point scans curves gathers a bit better so less diffuse to look at. If you agree appeal is by occasion KH 80 sample 1 and KEF LS50 shall be manual recomputed because their spin data used for preference rating have that vice versa U as seen in row 2 and 3, other spins in the 15 times shared data was okay only ones missing in my analyze is the hor/ver txt-files for JBL Control 1 Pro/NHT Pro M-00/Revel C52 that look never had a download icon.

Second investigation is probably bit irritating hear about again in that amirm had told he is not going to re adjust center axis for computed spin to whatever axis manufacture specification is relative to being that dead center on tweeter, myself was okay that statement because it adds more work in process and initial thought it probably not means so much as below new investiagation show, on the other hand well could be wrong but had feel amirn like precision and cooperate some standards and then jump over re adjust recommended point for compute spin data is the other way arround, also because myself speculate maybe Klippel or Neumann have been in contact or wonder on subject so think below input should be of good enough value to assess if that offset number is important to be in line with manufacture data.

In past was able to improve on axis curve for KH 80 sample 1 adjust a offset in software for microhone Y-axis -52mm relative to spin or set transducer to Y-axis+52mm, KH 80 sample 2 was also helped using the same offset setting but much less because it had that weird big such out in highest octave, then yesterday come to think about that because i use a button in software's option called "CTA-2034-A" then few setting adjust to that standard including listening distance is set to 2000mm and well that is not point where MK 255 microphone is positioned so i set it to 200mm and sample 2 with the high quality 1000 point scan and manual optimal setting for fitting now get mostly repaired for that such out in highest octave and on axis in general begin look smooth as in Neumann specification, also KH 80 sample 1 benefit set listening distance to 200mm but in will mean it needs a manual recompute of spin data to improve highest octave omit upload animation curve, but one can see sample 1 improvement in overlay below animation where my thought is they start get closer that interpolated official print of KH 80 on axis.

There was no physical damage that I could see. The angle of the speakers relative to mic though is very tricky. These speakers don't have a flat face to align them or anything. So it could still be some difference in the presented angle for on-axis.

Also quote above together coincidence setting that short 200mm distance to same as where NFS microphone really is positioned makes get curve in coherence with what was supposed.

In below IIRC sample 1 is a 500 point scan that with reference to graphs above animation will benefit manual recompute especially in highest octave, and sample 2 is a higher 1000 point scan that include optimal manual adjustment before spin was computed, both below have microphone position in space or listening distance set to a short 200mm and Y-axis for microphone is offset on Y-axis -52mm.

Thanks listening, below i couldn't stop share because it looks a fantastic smooth response based on above 200mm adjustment for microphone distance, first is not bad raw out of box resonse, second is EQ filter prediction exactly for X-axis -20 deg step where listening window is a close overlay to on axis curve, target for reference angle axis is set flat using stopbands BW 7th @57Hz and BW 2nd @25kHz.

 

[stevenswall]

This is a bit sad, as I was thinking the KH80 would be a poor man's Genelec when it came to flatness and even dispersion. Didn't have the terrible blotches from lobing the others have when you move vertically, just a bit of a pinch around the crossover based on some measurements.

If it's this picky though, guess I'll condemn it like all of the other non-coincident speakers I've heard.

 

[q3cpma]

Hope I'm not blind, nobody from Neumann came in this thread, right? I wrote them an email and even got an encouraging answer, but that's all.

 

[Sancus]

Hope I'm not blind, nobody from Neumann came in this thread, right? I wrote them an email and even got an encouraging answer, but that's all.

Nope, don't think we've seen anything from them. It would be interesting to see their response on the bass dip, since it seems to be real, albeit minor since it's obliterated by room modes in all real-world situations.

Unclear if the 13khz dip on #2 is an out-of-spec sample or something else, and the 5-10khz "comb filtering" seems to be present in every speaker measured thus far, so it's looking increasingly like a systematic issue.

Will be interesting to see what happens with the Genelec 8341a measurement since that is another allegedly close-to-ruler-flat speaker.

 

[thewas]

Neumann are taking these matters very seriously so they want to do several tests (also with the Klippel NFS) before replying, it just takes some time.

 

[q3cpma]

Neumann are taking these matters very seriously so they want to do several tests (also with the Klippel NFS) before replying, it just takes some time.

I see, thanks for the info.