Neumann KH80 DSP Monitor Measurements #3

[audafreak]

I have a similar setup a subwoofer with its own amp with adjustable bands and 3 front kh80s and 2 side speakers. I used REW to flatten the subwoofer and then the Audyssey app in our Denon to generate a room curve. Based on ASR tests I set the crossover for the kh80s at 60 Hz. Based on Toole's tests I set up the room curve with more base above 1kHz.
We mostly listen to jazz, vocals, and classical so are not too worried about very high SPL although it's certainly loud enough for the occasional Hendrix track in our 40 sq m room. I suspect you'll have very reasonable SPL in your room but not earth shattering.

I don't think it is a good idea to set the crossover for KH80 to 60Hz especially if you raise bass area.

 

[Jon AA]

Is it the Denon AVR that does the low cut filter on the KH80?

If you're using a Denon, yes, it will put a 4th order at the crossover frequency on the KH80 (after you run Audyssey). The KH80 looks to have a rolloff steeper than a 4th order at 60 Hz already, so that may leave you with a bit of a hole in the response. But it is close....

80 Hz may end up being the better choice--it certainly would be for longer listening distances as that will take a load off the little speaker and reduce distortion/compression in the bass at higher volumes.

 

[HarmonicTHD]

Definitely rather 80Hz as a starting point (or even a bit higher) , given the size of the room and the small speakers. But only measurements can tell for sure.

 

[RDoc]

Thanks Rdoc,

Encouraging feedback If you have the link to some posts about "Based on ASR tests I set the crossover for the kh80s at 60 Hz.", I'm interested.

Is it the Denon AVR that does the low cut filter on the KH80?

Best regards,

JMF

Yes, the Denon AVR does the high pass filter.

Look at the first post in this tread for the kH80 test.

As far as 60 Hz crossover, since I've got both large surrounds and a subwoofer I just wanted to get the KH80s into their linear region while still preserving some bass from the front, although, 60 Hz isn't very directional, but still. Remember that there's a 500W subwoofer below that which is pretty linear (after adjustment) from about 22 to 120 Hz and Audyssey will cross that over as well to fill in any drop. In any case, by the time all the room reflections occur, a couple of db will get (literally) lost in the noise.

People keep worrying about SPL with these speakers. If you regularly listen at levels above 90 db, I suppose that might be an issue, but, as I said, we mainly listen to jazz and classical at normal to slightly high listening levels, around 70-80+ db at our couch, occasionally a bit louder, and it's fine. I've actually tried just shutting off the subwoofer and, while the system doesn't sound anywhere near as good, it's certainly loud enough for us. Having 3 front speakers probably helps I suppose. However, we just had some friends over who are into Led Zepplin and between that, Hendrix, and Rimsky-Korsakov we did shake the room for a while, and it sounded great. That's definitely an exception for us though.

 

[RDoc]

I don't think it is a good idea to set the crossover for KH80 to 60Hz especially if you raise bass area.

The name of this group is Audio SCIENCE review.

 

[jmf11]

The KH80 are at home starting to play music with a basic behringer UM2 USB sound card. A SVS SB-1000 sub is on the way. It has 80Hz -12db slope subwoofer low pass and speakers high pass feature. This will allow first experiments of integration of the sub with the KH80.

One question to users (I didn't found the answer in the manual): what is the Neumann logo state in standby? I would expect it to be Off, but mine stay solid white and I don't know if it is that the speakers are not in standby or if it is normal.

Best regards,

JMF

 

[HQY]

The KH80 are at home starting to play music with a basic behringer UM2 USB sound card. A SVS SB-1000 sub is on the way. It has 80Hz -12db slope subwoofer low pass and speakers high pass feature. This will allow first experiments of integration of the sub with the KH80.

One question to users (I didn't found the answer in the manual): what is the Neumann logo state in standby? I would expect it to be Off, but mine stay solid white and I don't know if it is that the speakers are not in standby or if it is normal.

Best regards,

JMF

the logo should be off in the standby status. kh80 may have issue with standby or wakeup. there are discussions about the issue in the asr. i would think it's a bug.

 

[Curvature]

If you're using a Denon, yes, it will put a 4th order at the crossover frequency on the KH80 (after you run Audyssey). The KH80 looks to have a rolloff steeper than a 4th order at 60 Hz already, so that may leave you with a bit of a hole in the response. But it is close....

View attachment 294965

80 Hz may end up being the better choice--it certainly would be for longer listening distances as that will take a load off the little speaker and reduce distortion/compression in the bass at higher volumes.

I usually set crossover based on distortion measurements of the mains. Ends up being around 100Hz usually to take the strain off the little woofers.

 

[RDoc]

I usually set crossover based on distortion measurements of the mains. Ends up being around 100Hz usually to take the strain off the little woofers.

According to THX 80Hz or lower can't be localized, so 100Hz would allow some sense of where the bass was coming from. I suppose with multiple subwoofers that wouldn't be much of an issue.
Where do you get distortion measurements from for speakers? Although, according to Toole, distortion has to get pretty bad before it's audible anyway.

 

[Curvature]

According to THX 80Hz or lower can't be localized, so 100Hz would allow some sense of where the bass was coming from. I suppose with multiple subwoofers that wouldn't be much of an issue.
Where do you get distortion measurements from for speakers? Although, according to Toole, distortion has to get pretty bad before it's audible anyway.

You're right, of course.

From what I remember of the original study, it was about whether or not listeners could tell where the sub was located, and it concluded 80Hz was the safest bet, although higher wasn't necessarily detrimental. There's plenty of examples where users set their crossover at 120Hz or higher and report no problems.

 

[Curvature]

Where do you get distortion measurements from for speakers?

I use anechoic data, if it's available, from ASR, EAC, Soundstage, S&R or manufacturers. If it's not, I don't worry about it.

 

[Hexspa]

You don’t necessarily need to high pass the mains. The low end acts like mini multi subs. Ed Mullen from SVS talks about it in this Sound and Vision interview, at least in the context of a 2.2 system https://www.soundandvision.com/content/ed-mullen-director-technology-and-customer-relations-svs
I’ve gotten the best results with my Yamaha HS system by not using the high pass filtering.

 

[RDoc]

You don’t necessarily need to high pass the mains. The low end acts like mini multi subs. Ed Mullen from SVS talks about it in this Sound and Vision interview, at least in the context of a 2.2 system https://www.soundandvision.com/content/ed-mullen-director-technology-and-customer-relations-svs
I’ve gotten the best results with my Yamaha HS system by not using the high pass filtering.

Hmm. I must say that to my wife (who has exceptional hearing) and I, adding a high pass filter to the kh80 mains made a huge difference in perceived sound quality. The effect wasn't in the lows though, we both remarked that it was the mids and highs that sounded much better. My rational is that by removing the low frequencies, the relatively small amplifiers in the kh80s didn't have to waste power attempting to reproduce low frequencies they couldn't do anyway. I've not measured that, so it may not be accurate, but we did both instantly notice the same effect.

 

[RDoc]

You're right, of course.

From what I remember of the original study, it was about whether or not listeners could tell where the sub was located, and it concluded 80Hz was the safest bet, although higher wasn't necessarily detrimental. There's plenty of examples where users set their crossover at 120Hz or higher and report no problems.

This discussion is moving me towards getting a second subwoofer.

 

[Curvature]

Hmm. I must say that to my wife (who has exceptional hearing) and I, adding a high pass filter to the kh80 mains made a huge difference in perceived sound quality. The effect wasn't in the lows though, we both remarked that it was the mids and highs that sounded much better. My rational is that by removing the low frequencies, the relatively small amplifiers in the kh80s didn't have to waste power attempting to reproduce low frequencies they couldn't do anyway. I've not measured that, so it may not be accurate, but we did both instantly notice the same effect.

Two of my recent posts on the topic might be useful:

• https://www.audiosciencereview.com/...atment-goldensound.45104/page-84#post-1638979
• https://www.audiosciencereview.com/...we-crave-distortion.45978/page-2#post-1639330

The KH80 is very limited in its output because of its size. That takes nothing away from how well designed it is, and its well controlled directivity >1kHz. It's simply small, which limits what it can do in the MF and lower without distorting at distances longer than a meter or so. Larger speakers also show controlled directivity down to a few hundred Hz, also important.

 

[RDoc]

Two of my recent posts on the topic might be useful:

• https://www.audiosciencereview.com/...atment-goldensound.45104/page-84#post-1638979
• https://www.audiosciencereview.com/...we-crave-distortion.45978/page-2#post-1639330

The KH80 is very limited in its output because of its size. That takes nothing away from how well designed it is, and its well controlled directivity >1kHz. It's simply small, which limits what it can do in the MF and lower without distorting at distances longer than a meter or so. Larger speakers also show controlled directivity down to a few hundred Hz, also important.

Do you have any sources to show distortion as a function of distance for small (or any size) speakers at reasonable listening levels, or even what the physical mechanism would be?

How significant controlled directivity is I'm not sure, are there any blinded tests? Dr Toole doesn't seem overly impressed by the effects of variable room reflections. Anyway, looking at directivity for the kh80, the Revel F35 and the Revel f228be, both pretty big speakers, I'm not convinced size has much to do with directivity in any meaningful way. Does 3-5 db really make any audible difference after being bounced around a real room?

Also, just looking at the curves, if they actually do mean something, it's not at all clear to me which of these would be better for reflected sound. I'd expect the more linear the lines the better, irrespective of slope but I'm unaware of any actual blinded tests.

These are normalized horizontal directivity curves by angle, the dark magenta curve is 60 degrees, the vertical scale is db relative to 0 degrees right in front of the speaker.

 

[Curvature]

Do you have any sources to show distortion as a function of distance for small (or any size) speakers at reasonable listening levels, or even what the physical mechanism would be?

SPL is a function of distance. Every doubling of distance in a small room SPL falls 3-4dB, more if heavily treated. Therefore, to get the same sensation of loudness, you have to turn up your system, which increases the distortion.

The max output for the KH 80 according to Neumann is below.

At 100Hz, for a distance of 1.4m, you hit 10% THD (the blue line; or 20dB SINAD) and get an output of just over 90dB SPL. That's psychoacoustically equivalent to 80dB SPL at 1kHz (look at the red lines in the equal loudness contours below to understand equivalencies, since we don't hear evenly per frequency). I'd say 80dB SPL is reasonably loud. It's a normal working volume in studios, and way quieter than noise you hear as a passenger while flying or the stupendous volume in movie theatres. Mind you, even at its limits the KH80 won't sound terrible. It's still a very good speaker.

Another important measurement is compression, but Amir does not include it in his suite. Compression tells you how much output is lost when the speaker plays louder than it can handle. This is per frequency, so according to the max SPL graph above, the KH 80 will do fine above 1kHz, but will struggle with the lower mids and bass, although we don't know how much. Compression basically tells if there will be a change in frequency response depending on playback level. For larger speakers this is much less of a problem. See some measurements here for a similarly good and similarly small speaker measured by Erin: https://www.erinsaudiocorner.com/loudspeakers/genelec_8331a/ It's an active speaker, like the KH80, and has limiting built in to protect it from overheating and destroying the drivers.

How significant controlled directivity is I'm not sure, are there any blinded tests? Dr Toole doesn't seem overly impressed by the effects of variable room reflections.

An big part of Toole's work was to prove that directivity was one of the most important factors in what you hear, and that directivity (specifically off-axis response) in speakers heavily influences whether or not reflections are found beneficial. I'll direct you to his book for the results and details of blind tests.

Interpreting the measured results is not easy. You're looking for relative smoothness. Use this tool to compare speakers: https://www.spinorama.org/ (KH80 vs. F35, vs. F228Be) This is a poor speaker because of the directivity error at HF: Triangle Borea BR03.

looking at directivity for the kh80, the Revel F35 and the Revel f228be, both pretty big speakers, I'm not convinced size has much to do with directivity in any meaningful way. Does 3-5 db really make any audible difference after being bounced around a real room?

The lower the frequency, the less change in dB SPL is required before a difference is registered, again according to the equal loudness contours above. By the time the KH80s get to 400Hz, they are more or less omnidirectional, while the F35 maintains control to 200Hz, and that's even more the case for the F228Be. This also has to do with how speakers interact with rooms, which is a complex topic in itself. Take this recent trainwreck/excellent thread here: https://www.audiosciencereview.com/...peakers-and-room-treatment-goldensound.45104/

Physical size also affects how the speakers radiate and couple with your room. The KH80 has a precise, small image because of the closely spaced drivers and good directivity. The Revel speakers will sound more diffuse, although the tone will be similar.

Here are two speakers widely considered excellent: Dutch & Dutch 8c, Genelec 8361A. They are much bigger than the KH80. Due to driver placement and other elements of design, they have excellent directivity control. And then compare the KH80 to the larger Neumann KH420 for example. Above 1kHz it'll be similar in tone and presentation, but for lower frequencies it'll be a different experience. Now take a speaker with a 28" front width and a horn: Danley SH-60.

 

[Hexspa]

Hmm. I must say that to my wife (who has exceptional hearing) and I, adding a high pass filter to the kh80 mains made a huge difference in perceived sound quality. The effect wasn't in the lows though, we both remarked that it was the mids and highs that sounded much better. My rational is that by removing the low frequencies, the relatively small amplifiers in the kh80s didn't have to waste power attempting to reproduce low frequencies they couldn't do anyway. I've not measured that, so it may not be accurate, but we did both instantly notice the same effect.

It goes both ways: adding bass also changes the sound and many people perceive an improvement in the midrange when adding a subwoofer. There are reasons for this such as the lower harmonic frequencies giving a stronger tonality to the higher ones. Regardless, as long as we're talking subjective preference (you say better) it's tomato/toe-mah-toe.

Removing bass frequencies reduces masking of the higher ones so you will notice an improvement in clarity. Erin's Audio Corner showed that the difference in harmonic distortion from high passing a speaker is not that great. Of course you remove the LF bandwidth that is higher-distortion but there are also marginal gains in the higher frequencies as well. Another member replied to you and said you also change the directivity somewhat, or at least remove a lot of the LF energy that's radiating around your room. Depending on the acoustic response of your room, this will reduce the level of ringing in the bass so you'll perceive an improvement in the time domain. Of course, you're not really solving any problems here but simply avoiding them.

I'll mention that a member here mentioned in the Avantone Mix Cube review thread that part of the improvement in "transient response" I was perceiving could be due to the high-passed nature of the speaker. This was something I'd not considered and I've been using DSP filtering on my Yamaha HS50 and subjectively perceive them as a bit sharper. Now, this is sort of a complex anecdote but, suffice to say, removing bass will certainly make the higher frequencies seem "clearer" or "cleaner" despite the fact they may have not changed much at all.

 

[Curvature]

Erin's Audio Corner showed that the difference in harmonic distortion from high passing a speaker is not that great.

A speaker of what size, though? The result is definitely dependent on the main speaker's inherent output capability.

at least remove a lot of the LF energy that's radiating around your room. Depending on the acoustic response of your room, this will reduce the level of ringing in the bass so you'll perceive an improvement in the time domain.

This is roundly incorrect. Search for information about multisub setups and why they work.

You may find this comment I made earlier useful: https://www.audiosciencereview.com/...n-bass-range-better.45230/page-2#post-1615755

 

[Hexspa]

A speaker of what size, though? The result is definitely dependent on the main speaker's inherent output capability.

This is roundly incorrect. Search for information about multisub setups and why they work.

You may find this comment I made earlier useful: https://www.audiosciencereview.com/...n-bass-range-better.45230/page-2#post-1615755

I think it was his review of the KEF R3 Meta.

About my comment being "roundly" incorrect, let's clarify a few things before jumping to conclusions, shall we? I wasn't talking about a subwoofer - my comment was simply on a single loudspeaker in a small room <10,000ft3. The fact is that rooms ring and high passing the bass out of a speaker will prevent the room from ringing via the speaker as a stimulus; at least below where you high passed and especially for high-Q modes. Secondly, you seem to be assuming I don't know about "multisub" and its benefits. Know that I am aware of how placing LF transducers at opposite polar points of a room's modal pressure zones can minimize ringing and smooth the SPL response.