Neumann KH 150 Monitor Review

[Tangband]

Does anyone know what the sampling rate capability is on this unit?

Not clear, but the crossover slopes seems to be 48 dB/oct .
The monitor can probably take 24 bit 192 kHz digital input, but the internal processing in the dsp crossover is probably 48 kHz or 96 kHz , 24-32 bit .

 

[Pearljam5000]

Not clear, but the crossover slopes seems to be 48 dB/oct .
The monitor can probably take 24 bit 192 kHz digital input, but the internal processing in the dsp crossover is probably 48 kHz or 96 kHz , 24-32 bit .

 

[Tangband]

What a beast. Horizontal dispersion looks coaxial. Makes non-coaxial genelecs look flawed.

Theres always a tradeoff when using waveguides - with a one inch dometweeter and a big waveguide combined with a low crossoverpoint such as in kh150, at 1,7 kHz , the directivity can be very smooth . The tradeoff is that a 0,75 inch dometweeter in a waveguide such as the one used in 8340 has slightly better dispersion at higher frequencies than a 1 inch dome.

One has to listen to both constructions to hear how the sound is, what is prefered, and also knowing that a 2 way loudspeaker always, no matter how good it is and how optimized it is, is a compromise in the chosen crossover frequency, dispersion, IM distortion and bass ability with high spl.

 

[Tangband]

View attachment 249109

This makes perfect sense with a metal dometweeter with breakup at maybe 28 KHz . 48 kHz sampling frequency attenuates everything higher than 24 kHz , so the tweeter resonance will never be excited.

 

[jhaider]

Hats off to Neumann. Such superb performance (now validated twice, first by Sound und Recording and now by @amirm) makes one legitimately wonder if there is room for improvement on this model in its size/footprint class. Sticking within the assumed design brief, I can only think of one: greater bass output. As it stands, from ~300Hz down is actually matched in headroom by JBL’s teeny tiny (and cheaper I think) 705p. That JBL 5” woofer is ridiculously potent. But they also need to leave enough on the table to make a KH 150 II in a dozen years or so.

705P

New Neumann KH 150

Also I guess the addition of “prosumer” features such as Wi-Fi and AirPlay 2 would be nice for us, though out of scope for their target market. Hopefully Neumann's improved the standby circuit compared to the super flaky one on KH 80, so that’s not a thing any more.

Aside from those, it’s hard to see any “improvements” that aren’t really “design changes,” such as a wider-dispersion waveguide and maybe tweeter changes to accommodate the lower crossover point required for wider coverage with no dispersion disruption. But that's again a different speaker.

But we are hearing the sum of on-axis and reflections from off-axis.
That's why Estimated In-Room Response comes in instead of on-axis only.

Are we? Go back to @Sean Olive's curve drawing study, which found trained listeners’ impressions of a loudspeaker’s spectral balance tracked the on axis/LW.

IMO PIR is a wildly overrated metric, and speakers with horiztonal dispersion disruptions that voice the listening axis response to improve PIR are…not my personal preference.

No. KH150 has unbalanced directivity. DI steps up to low treble due to dimensions and shape of the wave guide. It should be listened close to on-axis. KH150 is a "monitor" so all that could be irrelevant for many others (but not for me).

Are you saying you think it’s sonically worse to go from wide to narrow at lower frequencies than to have a dispersion disruption at the mid tweeter cross?

Personally I don’t recall a case when I’ve found the latter acceptable, but the former seems pretty benign in use.

 

[D!sco]

Pictures on page 4.

I was curious, so I looked up the price of all the chips. This barely scratches the surface of the time and research investment, but the parts themselves are rather affordable.
Texas Instruments Amplifier chip, $16.08
Analog Devices DSP chip, $15.33
Cirrus Logic Codec chip, $12.57
STM ARM logic chip, $18.03
Heatsinks, caps/coils/resistors, ~$50-100
Some drivers, $50-200

Interesting that the logic chip used is ARM-based. I wonder if it could run CamillaDSP...

 

[Tangband]

I was curious, so I looked up the price of all the chips. This barely scratches the surface of the time and research investment, but the parts themselves are rather affordable.
Texas Instruments Amplifier chip, $16.08
Analog Devices DSP chip, $15.33
Cirrus Logic Codec chip, $12.57
STM ARM logic chip, $18.03
Heatsinks, caps/coils/resistors, ~$50-100
Some drivers, $50-200

Interesting that the logic chip used is ARM-based. I wonder if it could run CamillaDSP...

To be clear - isnt those the chips thats inside the cheaper kh80 , and not the kh150 ?

 

[tcli]

Absolutely stunning performance. I can't stop thinking about how much time they spent working on both DSP and materials to get such a flat anechoic On-Axis... And I mean: repeatable, dead flat FR, even when tiers-measured.

Flat on-axis is very easy to get with a DSP. Smooth directivity transition from low to high frequencies is largely more difficult.

 

[pseudoid]

Thank you Amirm for both the review of a diamond-in-the-rough find and for your 'musica nuda" youtube link.
I've never had a hankering for 'active' speakers; but this KH-150 is a legitimate game-changer for me.
The clincher was your first graph: That plot alone tells me that the need for an additional EQ-stage is starting to become moot point for monitors..
Error Correction: The only down side I see is the lack of balanced (XLR) inputs for such a great active speaker described as a "studio" monitor.

 

[Sancus]

IMO PIR is a wildly overrated metric, and speakers with horiztonal dispersion disruptions that voice the listening axis response to improve PIR are…not my personal preference.

We Don't Hear The PIR is something I would bold under every review lol. Everybody wants to simplify how a speaker sounds to one line but that's not how we hear things.

If it was that simple, then the pref score would be perfectly accurate, and PIR would be 100% of it, instead of 38%.

 

[kimmosto]

Well, I don't think so.

Ok, maybe I thought that woofer's LP is steeper at 3-5 k. Anyway, this directivity index is not two different slopes or totally separate lines. Just tiny hump at XO and small dip above if/because phase match is not possible to on-axis anymore. Max 1.5 dB difference to smooth average cannot cause any harm alone because it can be compensated with tiny S-curve in ON, LW, ER, PIR and SP.

 

[Matias]

The only down side I see is the lack of balanced (XLR) inputs for such a great active speaker described as a "studio" monitor.

But there is XLR analog input, this is a given in the most basic monitors let alone the best ones such as these.

 

[kimmosto]

Are you saying you think it’s sonically worse to go from wide to narrow at lower frequencies than to have a dispersion disruption at the mid tweeter cross?

No. Smooth transition with shallow slope in DI from LF to HF. No steps, not zero (or negative) or too high slope. If slope of DI is very low (flat DI), LF could/should be lifted in on-axis to get about the same balance with conventional box speakers. For example half space designs (in-wall etc) should have few dBs extra bass and low mid. This is just statistics related to traditions and gear used by music production especially 3-4 decades ago. And if listening distance can vary from zero (head between the speakers) to several meters without need to rotate speakers, design should be close to zero diffraction and off-axis reponses to front sector should turn down above ca. 2 kHz. Like tweeter without wave guide in a box with very long radius rounding.

 

[PeteL]

But there is XLR analog input, this is a given in the most basic monitors let alone the best ones such as these.

I don't see it from the picture.

 

[RobL]

I don't see it from the

Here you go!

 

[ROOSKIE]

We Don't Hear The PIR is something I would bold under every review lol. Everybody wants to simplify how a speaker sounds to one line but that's not how we hear things.

If it was that simple, then the pref score would be perfectly accurate, and PIR would be 100% of it, instead of 38%.

I know, this is so clear from Toole's book and as you state that fact that we need the Whole SPIN and a host of other measurements to create an accurate picture.
Funniest part is that unlike the great actual measurements posted here, it is not even a measurement, it is as stated a prediction & 38% of a metric that Harman themselves has abandoned according to previous threads about it.

I believe in Toole's book he mentions viewing the PIR and actual in room measurements in 1/2 octave or even 1/1 octave resolution. Just to get the basic idea of a downward trend and the general amount of energy in the room.

 

[Arc Acoustics]

Max 1.5 dB difference to smooth average cannot cause any harm alone because it can be compensated with tiny S-curve in ON, LW, ER, PIR and SP.

I don't know about the hearing threshold of DI, I do not have that abundant experience, but it is rather cheerful.

 

[DJBonoBobo]

Here you go!

Yes, analog input is XLR. The RCA socket is an additional digital SPDIF input.

 

[Mario Sanchez]

I opened my KH80s... weird. Here's those pictures.

View attachment 248964
View attachment 248965
View attachment 248966
View attachment 248967
View attachment 248968
View attachment 248969
View attachment 248970
View attachment 248971

This is someone else's picture:
View attachment 248972

View attachment 248973

From someone else.
View attachment 248974

Well, these are all from me (it was a KH80 A G CCC unit) , it's a shame that the board wasn't as clean as the 120 was, and looks a bit cheapo. Still a far cry from badly engineered though, everything was either braced or bolted down where they can't be (like the PCB part).

 

[DMill]

Tis the season.