-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Neumann KH 80 DSP Monitor Review
- Thread starter amirm
- Start date
- Status
I've read it, to me they measure pretty much the same or with minimal differences FR wise (we miss all the horizontal and vertical dispersion curves). The point I'm trying to make is: is it possible that a speaker that measures better than another sounds worse under the same identical conditions (same room, same position, etc.)? If it is possible, and I don't think so, why are we wasting time to measure it?
Yep, especially taking into account the tested SPL level and reference angle. Neumann is about as close to flat as we'll get, though the JBL has a little more bass and may compress less.
And at the end of the day, that's what matters about amir's efforts here. It's the wonderful thing about measurements. We can decide for ourselves what we'd rather take the risk on =]
I've certainly had situations where I preferred the sound of a worse measuring speaker to a better measuring one, but it's also the exception to the rule.
Funnily enough, In my case, that was the JBL L100 Classic, which has some notable deviations from flat compared to many/most some of the speakers I've measured, yet it's one of my favorite sounding speakers so far, like top 3. I have some theories on why I might've liked it so, but no firm answers =]
Maybe JBL just has some secret sauce we're missing
Well, it certainly can be possible. The generalized preference rating model is 86 percent accurate and is designed around living room listening. That still leaves 14 percent up to other factors, especially in different listening setups.
I personally like a slight boost around 2k for stereo listening to account for interaural crosstalk and make dialogue clearer. In my setup, everything seems to sound better with a bit of EQ that way (except for the L100 classics, which have that built in). Yeah, it's not ideal because it introduced a peak off axis, but I always listen "seriously" from the same position, so I don't mind.
Also, sighted impressions are still valuable for individual choices. If one speaker slightly beats another in a blind test, but looks like a regurgitated baked potato, you can bet I'll pick the prettier speaker. Measurements and blind tests help inform purchasing decisions and who I want to support as a manufacturer, but at the end of the day I do have to look at the speakers too - and if that colors my impressions, so be it.
Last edited:
Measuring speakers spinorama is definitely not a waste of time - it is the only scientifically acknowledged way to separate good speakers from bad.
On the other hand, IMHO sticking a SINAD equivalent score to a speaker based on it's spinorama measurement may lead to a wrong conclusion that speaker with not so much better score will sound better in a room. Certainly, speaker with a lot worse score will sound worse in a room, but based on what @mitchco posted above my poinion is that we should have no better than traffic lights resolution (like those 3 colours of SINAD charts) for evaluating speakers based on spinorama measurements.
Yep, especially taking into account the tested SPL level and reference angle. Neumann is about as close to flat as we'll get, though the JBL has a little more bass and may compress less.
And at the end of the day, that's what matters about amir's efforts here. It's the wonderful thing about measurements. We can decide for ourselves what we'd rather take the risk on =]
I've certainly had situations where I preferred the sound of a worse measuring speaker to a better measuring one, but it's also the exception to the rule.
Funnily enough, In my case, that was the JBL L100 Classic, which has some notable deviations from flat compared to many/most some of the speakers I've measured, yet it's one of my favorite sounding speakers so far, like top 3. I have some theories on why I might've liked it so, but no firm answers =]
Maybe JBL just has some secret sauce we're missing
Well, it certainly can be possible. The generalized preference rating model is 86 percent accurate, and is designed around living room listening. That still leaves 14 percent up to other factors.
I personally like a slight boost around 2k for stereo listening to account for interaural crosstalk and make dialogue clearer. In my setup, everything seems to sound better with a bit of EQ that way (except for the L100 classics, which have that built in). Yeah, it's not ideal because it introduced a peak off axis, but I always listen "seriously" from the same position, so I don't mind.
It still needs to be established if following with JBL and Neumann what Mitch did in his experiment would yield an audible difference between these 2 speakers when listening to them in a room.
Agreed 100%. Also I would expect more bass from the JBL since they have a bigger woofer and are much bigger than the Neumann (they need a subwoofer in any case)
Follow me: taken to the extreme consequences if we accept that under the same conditions a worse measuring speaker sounds better than a better measuring speaker (and I do not accept that) the logical conclusion is that we don't need measurements or that we are unable to derive any pratical utility from them.
Follow me and read Mitch post once again: we may easilly establish that a speaker that has slightly "worse" spinorama than the other one would sound equally good in a room once EQ-ed to the same target curve. We agreed that Neumann spinorama looks better than the JBL one, but if we put them in a room and EQ-ed them both to the same target we probably wouldn't be able to distinguish them.
But that certainly doesn't mean that spinorama mesaurements are useless as we need it to separate bad speakers from good ones.
Krunok, you are not really following me. Don't introduce EQ now which is a further step. Bear with me, it is only logic:
I believe that we need measurements and that better measuremens always translate in better sound if you compare two speakers under the same conditions (and without EQ in the equation for now). But if this statement is untrue, so we can have a worse measuring speaker sounding better than a better measuring speaker or said in another way if better measurements don't always translate in better sound (when comparing two speakers under the same conditions) the utility of measurements is zero, nil. Its a demostration ad absurdum.
If you have two speakers and one measures like shit but the other measures perfect you are going to buy the one that measures perfect because you expect that it translates to a comparatively better sound. If you break this equivalence, better measures = better sound, measurements don't help you anymore in making a decision, I think it's pretty clear (unless you are an engineer).
EQ is an afterthought: I buy a worse measuring speaker because I think that I can correct it and make it a better sounding speaker compared to another that is better measuring but less able to be corrected.
I believe that we need measurements and that better measuremens always translate in better sound if you compare two speakers under the same conditions (and without EQ in the equation for now). But if this statement is untrue, so we can have a worse measuring speaker sounding better than a better measuring speaker or said in another way if better measurements don't always translate in better sound (when comparing two speakers under the same conditions) the utility of measurements is zero, nil. Its a demostration ad absurdum.
If you have two speakers and one measures like shit but the other measures perfect you are going to buy the one that measures perfect because you expect that it translates to a comparatively better sound. If you break this equivalence, better measures = better sound, measurements don't help you anymore in making a decision, I think it's pretty clear (unless you are an engineer
).EQ is an afterthought: I buy a worse measuring speaker because I think that I can correct it and make it a better sounding speaker compared to another that is better measuring but less able to be corrected.
Last edited:
DDF
Addicted to Fun and Learning
- Joined
- Dec 31, 2018
- Messages
- 617
- Likes
- 1,360
The possibilities that come to mind are the differences in source (they used many CDs) but most likely its due to differences in averaging constants used. These have a significant impact on the outcome. Perhaps you're using longer term averages, which would more suppress the higher frequencies. If you want to really compare, you'd need to look at the standards and mimic the method.
I do not think that conclusion, even if meant as an exaggeration, is logical at all.
All we know is that Amir preferred the seemingly worse measuring speaker among two good speakers in precisely one uncontrolled test. The spinorama, for all its utility and high predictive power, is still an imperfect predictor (86 percent), not a measurement of every possible parameter(when two speakers measure similarly do other factors emerge in preference?), and does not account for sighted listening impressions (which are a natural part of how we listen to speakers).
Put another way, there is a lot we'd have to do for listening tests to always match up with measurements.
1) Create the perfect model. Currently, our best predictive model for all loudspeakers is "only" 86 percent accurate. Doing so may requiring factoring in more details about speakers like compression, distortion, accounting for radiation patterns, etc
2) Do more rigorous listening tests: For best results get at least approximately 300 listeners as in the Olive paper.
3) Figure out how to factor in sighted impressions
4) Create a way to neutralize the circle of confusion which will always cause some speakers to sound better on some tracks.
I do not think the goal was ever to perfectly predict which speaker will win every listening test every time. Your premise that better measurements always = better speakers is negated by simply considering directivity: There is no evidence that wide directivity is inherently better than narrow directivity or vice versa. Even if it seems most people prefer wide, it's been shown to be a matter of preference, intended use, and matching with recording. So someone might prefer a worse measuring speaker with wide directivity if they prefer wide directivity.
So you are saying that better measurements don't always translate in better speakers. Ok, let's for a moment pretend that I agree with you: if that is true why you need measurements? Is there a secret sauce that we can't measure? Is directivity the secret sauce?
Krunok, you are not really following me. Don't introduce EQ now which is a further step. Bear with me, it is only logic:
I believe that we need measurements and that better measuremens always translate in better sound if you compare two speakers under the same conditions (and without EQ in the equation for now). But if this statement is untrue, so we can have a worse measuring speaker sounding better than a better measuring speaker or said in another way if better measurements don't always translate in better sound (when comparing two speakers under the same conditions) the utility of measurements is zero, nil. Its a demostration ad absurdum.
If you have two speakers and one measures like shit but the other measures perfect you are going to buy the one that measures perfect because you expect that it translates to a comparatively better sound. If you break this equivalence, better measures = better sound, measurements don't help you anymore in making a decision, I think it's pretty clear (unless you are an engineer
M8, well I am an EE and when I was a student I had 2 semestar course of measurement in electronics, 1 semestar course of logic and additional of math logic, so please don't lecture me.
Now, speaking of your statement that "better measuremens always translate in better sound if you compare two speakers under the same conditions (and without EQ in the equation for now)" - this is something that definitely is not true.
Better SINAD with DAC or amp doesn't mean at all that it will sound better than DAC or amp in a properly conducted blind listening test. If both DAC/amps have reasonably good SINAD they would both sound transparent and couldn't be distinguished one from another.
The same is true with speakers: speaker with "better" spinorama might not all sound "better" "under the same condition". Now, speaking of that "same" condition, you have only 2 choices: either put them in an anechoic room and listen to them there or put them in a room and use EQ to EQ-ed them both to the same target with the same tolerance. I am not at all interested if one speaker would sound "better" when listened as a single mono speaker in an anechoic room if they would both sound the same when EQ-ed and listened in a real room.
This should be pretty clear even if you are not an engineer.
ok so in the end you are both saying that better measurements don't translate to better sound: "speaker with "better" spinorama might not all sound "better" "under the same condition"
So I don't need a spinorama and I don't need to compare two different speaker spinoramas. The Neumann have a (much) better spinorama than the JBL but maybe the JBL sound better who knows, they might or might not. OK
So I don't need a spinorama and I don't need to compare two different speaker spinoramas. The Neumann have a (much) better spinorama than the JBL but maybe the JBL sound better who knows, they might or might not. OK
We cannot hear all the differences we can measure and thus very often we cannot differentiate between 2 products that measure differently. Both need to measure "good" of course, as those which measure "bad" sound bad.
IMO there is a chance that speaker with better spinorama would sound better in anechoic environment, when listened to mono signal coming from a single speaker, assuming you are a trained listener, etc.
However, there is a good chances that when you put 2 good measuring speakers like JBL and Neumann in your room and properly EQ-ed them you will not be able to hear the difference. As Mitch demonstrated..
That certainly makes spinorama measuremens very usefull as with it you can separate good from bad measuring speaker, as no EQ can help bad measuring ones.
OP
- Thread Starter
- #258
Your logic is good.
The practical issue at hand is that you have no reliable listening test results to use in judging a speaker's sound. So unless you can get the samples and compare them yourself in controlled environment, then you can't make a purchase decision. The measurements are a close substitute when performed in the manner we are doing. Ultimately, nothing is a guarantee. A cheap compass may not get you out of a jungle but is sure as heck better than randomly guessing.
that's my last as I have to go: you are saying that if you take two different speakers, one measuring better than the other, and you correct both to the same target curve at the end they sound the same. I bet so, I wouldn't expect nothing different but it is obvious don't you think?
I'm saying better measurements don't always translate into preferred speakers for one person with every song. I'm saying that there's still room for preference among great speakers. I'm saying we don't know the worth of every measurement, and we don't have every measurement available in readily accessible form. And I'm saying that the predictive power of the spinorama is meant to cover a wide breadth of people and will never give you perfect results with a single listener.
I, once again, bring up the example of the Revel Salon2 vs JBL M2. Both speakers measure well, but the M2 definitely looked better measurements-wise. The Salon2 won in a blind test by a clear margin. But even then, if I recall correctly, some people had a strong preference for the M2, which was accentuated on some specific tracks.
The way I see it, we need measurements because buying speakers is a gamble. For the most part, measurements can tell me more about a speaker than any written review can (coming from someone who writes reviews for a living), and probably better than a sighted test at a dealer can. If I'm considering two speakers, I'll get the one with better measurements because there's a better chance that I'll like it. I might be wrong, but I probably won't be.
Obviously better measurements mean better sound in the grand scheme, it's just not as absolute as you're making it to be. The JBL's might sound better to you, an individual. But if you shipped everyone on ASR a 305P and a KH80 and crossed them both perfectly with a sub and used room EQ below schroeder, and then put everyone on house arrest and forced them to perform a blind test test with 100 songs, I'd be willing to bet Harman's entire research budget that the Neumann's would be preferred by some margin.
Last edited:
- Status