Overnight Sensation - Klippel Analysis @ Erin's Audio Corner

[djg]

[Paul Carmody]

Here's what I can conclude from this:

A) Klippel's Near Field Scanner has incredible resolution. Forget about the old 1/3 octave smoothed, 1/6 octave smoothed or even 1/24th octave smooth, which is what I personally consider to a be minimum when designing crossover. That thing has a resolution of about 2Hz. So you see the warts, in all it's filthiness, under a dermatoscope. Everything looks totally, unless, you too, have been used to seeing things under a Klippel NFS (Have you seen seen a wart under a dermatoscope- yech!- if you haven't; ask for a gander the next time you see your physician/dermatologist)

B) Let us suppose that Erin's measurements are legit. And Paul's measurements use a 5ms gate. Which means a resolution of 200Hz. So between the octave of 1Khz and 2Khz, which is a span of 1000Hz, that's 5 data points. So let's take Erin's measurements and apply a 1/5 octave smoothing. Would it look the same as Paul's? I personally think it would. First off the NFS has a scale of 50dB, Paul's has a scale of 60dB, so apart from the frequency resolution interval of ~2Hz that shows every lump or bump as a jagged little hill, the height is also stretched vertically. And the tell tale sign is the that dip at 7Khz and peak at 4Khz is present on both measurements. The bass boost centre around 80-90Hz is there. The uncertainty region in a near field + fair field merge has ALWAYS between 500Hz to 2KHz. And here that 500-2KHz is on average at a higher level than the 2-8Khz region. In both measurements.

So I think roughly pans out, and it's all about measurements resolution superiority of the NFS vs near field + far field merging/blending.

C) The room determines how this speaker sounds. I've moved enough times in my life to know this. So despite the anechoic quality measurements of the NFS, extrapolated to 2m, how does it really sound in the room? I bet a small speaker like the OS is usually placed close to front wall. The real in-room response is not something that the NFS can do accurately, although it does try "Predicted in-room response".

Last summer I took Jeff Bagby's Revolution mini into my kid's 3x3L bedroom. I thought maybe he would like a pair.
It sounded so bloated in the upper bass, and too boomy in the mid-bass. "Ugh! What a waste of a good speaker" I thought.
Think I'll keep it for myself in my listening room.

Very good perspectives here. I think what we're running up against with the Klippel measurements--and I notice this with a lot of them they test at ASR--is that while the resolution is quite good, there's stuff it can't take into account. In other words: the room where the speaker finally lives. As all of us know, the room plays a huge part in how the speaker sounds to the listener. And while some may make the argument that all listening rooms should be properly treated with soundproofing or room correction software, well that's not really very realistic, is it?

I reminds me of marketing class back in college. They talked about how the shaving companies studied the way men shaved, and they noticed that most men were holding the razor blade wrong. So they had 2 options: either train the man how to properly hold the razor, or change the shape of the handle so men will end up holding it correctly, without even realizing it. Obviously the companies went for option 2. (Telling customers "you're doing it wrong" is usually not a good way to win their favor )

Even before Klippel, I noticed a very similar phenomenon at the DIY meetups. I'd notice that in those large hotel meeting rooms, the speaker would often take on a different characteristic than it would in a living room. Small, sealed speakers would have the bass range filled in and sound massive. Whereas larger speakers with vented bass tended to sound bloated and the bass could be overwhelming.

I think the hope of the Klippel spinorama has been "FINALLY we can measure what makes some speakers sound great. We can finally quantify all of it." But I don't think it's achieving that goal. Don't get me wrong, I like seeing the squiggles they measure over at ASR of various speakers. it's interesting. But sadly, it's not he last word. So we're left--again--to trust the listener's impressions, which I'll be the first to admit are not that reliable. So maybe it's 2 steps forward, 1 step back?

[djg]

Are we all holding our speakers wrong?

[Billet]

Recently, I was playing with crossover settings by ear on an older speaker project. I came across a configuration that sounds really nice. When I put these changes into my modelling software, it shows a double hump response that looks a lot like the Overnight Sensations response graph in post #1.

I like the sound and have no plans to "fix" it. I am afraid if I try to make a flat response out of it, it will lose the magic and sound just like my other speakers. The elusive "sounds good" is difficult to achieve and even more difficult to quantify.

[fatmarley]

Some people forget that music replay is an elusion and certain distortions, not only sound better, but can actually sound more like real, live music.

[PWR RYD]

I built a pair of Paul's Overnight Sensations probably 12 or 13 years ago before I started to design my own. They honestly do and sound EXACTLY as Paul describes on his website. On a work bench or low shelf, nearfield, at 80 dB or less, they sound great. Very fun to listen to design. If you move them 2 - 3 meters away from you, out on a stand, and expect 80 dB at your seated position.... they just can't do that. They never were supposed to do that.

[davidroberts]

Bottom line is measurements are an evaluation tool. They have their place but at the end of the day the most important tool is the ears of the listener/builder/buyer. If they enjoy them that is all that matters.
I have several designs from builders like Paul, Curt C and John Marsh in my house at the moment (not to mention one of Pete's Anarchy's I have yet to finish) and my daily driver is still an old set of Cinderella's.
I have heard more than one terrible speaker that someone else thought was the best thing since music was created. My response has always been "Great job. I am happy you like them".

[ErinH]

Very good perspectives here. I think what we're running up against with the Klippel measurements--and I notice this with a lot of them they test at ASR--is that while the resolution is quite good, there's stuff it can't take into account. In other words: the room where the speaker finally lives. As all of us know, the room plays a huge part in how the speaker sounds to the listener. And while some may make the argument that all listening rooms should be properly treated with soundproofing or room correction software, well that's not really very realistic, is it?

I reminds me of marketing class back in college. They talked about how the shaving companies studied the way men shaved, and they noticed that most men were holding the razor blade wrong. So they had 2 options: either train the man how to properly hold the razor, or change the shape of the handle so men will end up holding it correctly, without even realizing it. Obviously the companies went for option 2. (Telling customers "you're doing it wrong" is usually not a good way to win their favor )

Even before Klippel, I noticed a very similar phenomenon at the DIY meetups. I'd notice that in those large hotel meeting rooms, the speaker would often take on a different characteristic than it would in a living room. Small, sealed speakers would have the bass range filled in and sound massive. Whereas larger speakers with vented bass tended to sound bloated and the bass could be overwhelming.

I think the hope of the Klippel spinorama has been "FINALLY we can measure what makes some speakers sound great. We can finally quantify all of it." But I don't think it's achieving that goal. Don't get me wrong, I like seeing the squiggles they measure over at ASR of various speakers. it's interesting. But sadly, it's not he last word. So we're left--again--to trust the listener's impressions, which I'll be the first to admit are not that reliable. So maybe it's 2 steps forward, 1 step back?

I encourage you to look at the Estimated In-Room Response. In every instance where I have overlaid this with the actual in-room response (via moving mic method), it lines up with about a dB or two (above the Scrhoeder frequency). This is true in two entirely different rooms: my living room (big, open floor plan) and my home theater (upstairs bonus room).

As for soundstage: radiation pattern (horizontal spread and DI, namely) + frequency response matching between stereo sides (EQ'ing the left & right speakers to match in-room).

What you see below matches what I heard from these speakers in my room.

The real problem is that so many DIY designs are based on insufficient methods. It's not necessarily even that resolution; it's the improper merging of NF/FF, simulation + FF or just simply not having a proper FF measurement to begin with. It's more common that people would like it to be. But hopefully when I post data and people say "well, *my* data doesn't look like that" they'll take the extra steps to understand why rather than brush it off. Unless they design speakers by ear. Which makes measurements entirely moot for them. ;)




- Erin

[djg]

Douchebag got by the censor software, I'll remember that.

[ErinH]

Gotta be honest. I'm surprised to see measurements being dismissed here so readily. I figured - if any place - this would be the place where this kind of stuff is found interesting, helpful and beneficial. My feelings aren't hurt. I mean, it's just data. I'm just surprised to see the attitude toward the results. *shrug*

[fpitas]

Your measurements look pretty different than Paul's. Until we understand that, I'm not sure what to think.

[dld]

I agree Erin. Better to be humble and try to learn. Maybe a lot of people like what the like. They're a lot of very popular commercial speakers that measure poorly. Maybe way fewer golden ears than we think.

[djg]

When entertainment is beamed directly into our brains, speakers will seem so quaint. I just hope there's adblocking available.

[ErinH]

Wasn't that the premise of Jim Carrey's The Riddler in Batman Forever?

[djg]

Well, we can trust Bill Gates and Mark Zuckerberg, right?

[fpitas]

In a way, you can. Depend on them to cash in.

[tktran]

now:
“pfft! Meta whatever”

2042: my kids will ask

”in the olden days, did you have friends on the forums?”

[djg]

Yes, we had to type, with our fingers.

[tktran]

Gotta be honest. I'm surprised to see measurements being dismissed here so readily. I figured - if any place - this would be the place where this kind of stuff is found interesting, helpful and beneficial. My feelings aren't hurt. I mean, it's just data. I'm just surprised to see the attitude toward the results. *shrug*

Erin,

you know how I feel about the NFS (it’s amazing- I almost bought one 3 years ago)
https://www.audiosciencereview.com/f...04/post-191634

And you may know that I’ve been doing designs using full spherical measurements for awhile now.
But anyway, I digress.

I don’t think your measurements are being dismissed. It’s a form of cognitive dissonance. I mean people are, well, human!

We have our weaknesses. We’re not infallible. We are shocked, get a bit defensive, or be in denial, particularly about our babies (creations). And change is hard
Every designer should have been doing it, at least since 2015, when the the ANSI/CTA 2034A was published.

Paul,
guess what? We don’t need a NFS!
One can get pretty close if one is willing to learn new techniques and methods, and makes diligent attempts to put these to practice (ie. manual turntable, 10 ft+ away from closest boundary)

Diplomacy is an art, and some are at different stages of growth.
For instance, despite having limited English, a fellow decided to use translators to post his measurement/modification(!) of a speaker as well as make interpretations of correspondence between the buyer and designer said speaker (!!) , and simultaneously make accusations of another review/reviewer (!!!)

I think it was terribly foolish, and I feel for you, bro. Being dragged into that…

Paul is far more diplomatic that that fellow. My guess is that he’s saying;
well shucks that sure don’t look good.
But it’s still sounds pretty good.

I’m just not sure he’s yet ready to turn the full 360 degrees with VituixCAD2 or SoundEasy; the only two measurement systems that display CTA2034A measurements during crossover design.

I’m gonna go out on a limb, here Paul, and say, “your designs are great for the DIY community, but they could be even better, with better measurements”

You may have your new Pit Vipers.


But you you took a 360 degree view, you may have something even better than my Peregrine Falcons:

[djg]

Just think how many Overnight Sensations would sell if they were any good.

[davidroberts]

Gotta be honest. I'm surprised to see measurements being dismissed here so readily. I figured - if any place - this would be the place where this kind of stuff is found interesting, helpful and beneficial. My feelings aren't hurt. I mean, it's just data. I'm just surprised to see the attitude toward the results. *shrug*

I would not say the measurements are being dismissed. I have followed your site for some time now. In fact I completely agree with you on a lot of your reviews. I too think Andy's GB25 is one of the most amazing drivers ever (I have followed your car audio builds for a long time).
But sound is an individual preference. While a mic and measurements may be constant there are no two people that will perceive sound the same way. It may be what they are conditioned to, their Audible range of hearing or the shape of their heads or ears. It is all a perception based on senses and how they are processed.

Your measurements are a vital tool that folks can use to identify what they may like it dislike in a speaker. But at the end of it all it is still how an individual processes that information that makes the difference.

Some folks like mustard but despise ketchup. It's all a matter of personal choice.

[gworrel]

Gotta be honest. I'm surprised to see measurements being dismissed here so readily. I figured - if any place - this would be the place where this kind of stuff is found interesting, helpful and beneficial. My feelings aren't hurt. I mean, it's just data. I'm just surprised to see the attitude toward the results. *shrug*

I also don't think the measurements are being dismissed. It is more a matter of attempting to reconcile many favorable reports and reviews with what are relatively poor looking measurements.

At one time it seems to me that +/- 3dB was considered pretty decent for a loudspeaker. The Klippel measurements put this one very close to that range. While I have no doubt the response flaws revealed in the measurements are audible, it seems doubtful that the speaker is objectionable considering its size and cost. A 100Hz bump in the bass is a tried and true way to fake a decent bass response. It's not ideal, but sometimes you have to work with what you got.

I highly value the measurements done by you and ASR. I would love to pick up a pair of speakers that measure especially well to compare to my own home brew designs. I think your focus on speakers is especially valuable since that is where the real improvements can be had in home audio. ASR's measurements of electronics often become a matter of looking for better measurements for the sake of better measurements even when the differences are not audible.

[djg]

I always thought external ears are very influential to perceived sound.

[ErinH]

Are you stalking my social media selfies again?

[Wolf]

Good 'ol Quark!