The difference between good and great

[CtheArgie]

If may add some points and I hope you can bare with me. Most of you are using classical music in music halls as the "reference" for the sound you wish to get at home.

It seems to me that this "reference" only applies for unamplified music but most music nowadays IS amplified. Jazz concerts are amplified, even if some of the instruments are not, they are amplified for you to listen to them. Diana Krall (to pick one name) will never sound the way she sings without microphones because we likely never heard her without them. Even her piano is amplified. I heard Joh Pizzarelli many times and he is always amplified. The best way to hear him "as if live" is in his Thursday's social media concerts from home. He signs to one microphone and his guitar is unplugged. Actually, those concerts are remarkably realistic.

There is a Youtube series of how major rock songs are produced, and you can see (and listen) that most of them never intended to be as if live. The recording engineer and mastering is all "artificial". When they release the song to us, they have an expectation of how they want us to hear it. They use their equipment, and if they are any good, they will do the mastering using good equipment. But sometimes, the songs are released to be heard on headphones. Again, they use the "medium" as if it was another instrument.

So, for the majority of music, you want your equipment to be as faithful as it is in the mastering suite with he sounds they approved for us. Problems arise when you have bad engineers or if it is done at home with cheap stuff. I don't know how Billie Eilish and her brother decided on the master of their record, but I think they did everything at home (there is also a video of it).

So, realistic sound for non classic music is a different beast to the concert hall. But in this case, depending on where you sit you will get a different perspective. You can never replicate what you hear there. My wife sang in a Master Chorale at Disney Hall a piece with small choirs all over the hall, not only on stage. The hall was an instrument! How would you ever reproduce this?

 

[Pio2001]

Have any of you ever had experiences where the, technically speaking, better loudspeaker left you less satisfied in a direct comparison, and that beyond suspicion, there are no hard facts as to why this is? Or have the listening experiences always followed suit.

Hi,
I've been recently comparing two speakers (JBL 305P and Neumann KH-120) in very interesting setups : with their anechoic frequency response matched as far as I could (for this purpose, the JBL were equalized as well as possible to mimic the neutrality of the Neumann), and also with their in-room frequency response equalized (à la Dirac, with the same target, with a very high level of accuracy).

Among the four setup, both in-room eq were disappointing, although surprisingly close to the ear.
What's interesting is that the JBL 305P with custom equalization to make them flat had a better sound, to my ears, in the mid-range, than the Neumann.
This was confirmed with in-room MMM measurements showing a smoother frequency response from the equalized JBL than from the naturally flat Neumann. I think that it would have been quite difficult to predict this from both spinorama: the directivity index of the Neumann doesn't look worse than the JBL's.

Another unexpected phenomenon occured : the JBL produce much more treble in-room than the Neumann (about 2 dB more above 2 kHz). This is in contradiction with their directivity plots. It is a side effect from the room equalization below 1000 Hz, that shifted the respective reference level of both speakers, taking the sound power at 1500 Hz as the reference (I sorted this out afterwards, it was not on purpose). And the sound power DI of the JBL and Neumann is 2 dB apart at 1500 Hz.

But the phenomenon that is closest to what you are asking is the crystalline treble of the JBL compared to the dull treble of the Neumann.
Nothing in the measurement made so far explain this.

There is however a possible explanation : the gated measurement I made of the JBL are inconsistent with the in-room MMM ones above 6 kHz. Maybe there is something to explore here.
I have yet to try the following : measure the JBL from the listening position with very small MMM averaging, for example 10x10 cm, or 20x20 cm. I think it may reveal the differences that I'm hearing in treble, that would be completely averaged out with a 1x1 meter MMM measurement.

 

[Inner Space]

Have any of you ever had experiences where the, technically speaking, better loudspeaker left you less satisfied in a direct comparison, and that beyond suspicion, there are no hard facts as to why this is? Or have the listening experiences always followed suit.

Yes. Background - there was a question on another thread whether measurements could predict great imaging abilities. The consensus was no, not really, but a test was mentioned, that I have used before - play identical mono pink noise through both channels, and listen for a narrow vertical image between the speakers ... if it's truly vertical, doesn't stretch or bloat sideways, or lean left or right, then the system is capable of great imaging.

So ... in one of my rooms I have a system designed to be ultra-accurate, ultra-low-distortion, and ultra-precise. (Benchmarks into Yamaha NS-5000s.) It passed the imaging test fantastically well. The pink noise image was blade-thin - dimensionless, really. Best I ever heard.

But ... so many vocal tracks are booth-recorded in mono, and then pan-potted into the exact center of the image, which means putting the exact same signal on both channels, just like the pink noise in the test. The result is that the vocalist seems to have no dimension at all. My brain interprets it as the vocalist standing sideways on. It's very disconcerting.

 

[TimVG]

I'd still be willing to bet this has something to do with directivity. It's hard to compare measurements from different sources somethimes, and the G2 does not appear to have particularly wider directivity than the neumann, but it does seem to show more constant directivity behavior after 2kHz , which I find I tend to prefer in terms of tonality. Lots of waveguided/horn speakers will instead have a frequency response that becomes increasingly tilted off-axis.

This was also my guess. I was (and am still) surprised though that even away from boundaries (let's say around 4m) across the horizontal plane, that directivity still affected imaging in such a way with a single loudspeaker.

Just to be clear, people should not interpret this as invalidating the known science. Yes, the KH80 measures flatter, the G2 is no slouch though. Strictly tonally speaking, both were very acceptable to me and sounded quite similar.

 

[Fonzombie]

I've been pondering over this question for a while now

Being quite familiar (for a person without a background in audio from a technical pov) with the known research, obviously we want to tick all the boxes. Neutral direct sound, excellent directivity properties, lack of resonances, lack of distortion, etc.. The whole package.

What makes good and what makes great I wonder? "It depends" is very likely the answer, as with most things.
Without quoting specific posts. I've read (sometimes heated) discussions over budget loudspeakers with good measurements sounding less good than more expensive models with similar numbers, to which distortion is often brought up. On the other hand, we've also had (sometimes heated) arguments about that aspect when a loudspeakers with non-exemplary distortion numbers is perceived as audibly good, or great even. While sighted bias is often a factor, this is not always the case.

Another recent example, to which we're still awaiting Amir's review, is the JBL 4349. Harman's spinorama was posted in the thread, and while good, on paper it's not spectacular. Yet it would seem odd to me, that they would release a loudspeaker in this price range that didn't go through the blind test protocol. We weren't dazzled with the numbers however.

The Revel Salon2, while producing excellent measurements, technically speaking has been surpassed by several models from Revel themselves, as well as other manufacturers. I've yet to hear to read about anyone not favoring the Salon2 compared to those other Revel models though, apart Harman marketing.

Have any of you ever had experiences where the, technically speaking, better loudspeaker left you less satisfied in a direct comparison, and that beyond suspicion, there are no hard facts as to why this is? Or have the listening experiences always followed suit.

Where exactly are we with regards to the research? What have those +100 measurements and reviews told us so far?

Talk about down the rabbit hole lol. I have the JBL 4349s - initially I hated them, they were far too bright. Tried them with a Primaluna evo 400, MC462 and a Marantz 2275... and they sounded like drek on each one. But I disliked the JBL 4367 too, which I believe measured really well? I do however LOVE the 4429 and the 4365. I was right about to sell the 4349s when I tried one more amp on them and that made all the difference. They image beautifully and scale up very well... on this amp, the 4th amp, and to my ears. Thank god cus i really didn't want to ship them. The amp that paired well was the Line Magnetic-845 Premium. I can sit and listen for hours now - before it was immediate listener fatigue in 15 minutes or less. I still think the 4429s are the better speaker though - hands down. To my ears.

 

[Tsuchi]

I'll just say that there are so many variables involved when comparing the sound of speakers/headphones without measurements. We have to be conscious of the tricks our minds can play on us and things we also don't account for because they aren't directly noticeable.

But you also can't determine which is ultimately better by measurements alone. There will always be a place for listening to speakers/headphones in person. That process is important to discovering what to you is bad, ok, good, and great.

 

[sejarzo]

... When they release the song to us, they have an expectation of how they want us to hear it. They use their equipment, and if they are any good, they will do the mastering using good equipment. But sometimes, the songs are released to be heard on headphones.

I worried about this starting back in around 1993, when I read an interview of Ian Anderson around the time that The Best Of Jethro Tull - The Anniversary Collection was released. He remastered all the material so it sounded good to him on his wife's CD Walkman with the stock headphones.

 

[Hon]

Vivaldi's fun listening. Bach is deeper than the deepest ocean.

But what about The Beatles?

 

[Robin L]

But what about The Beatles?

About 1mm thin. Fun, but they were a Boy Band, after all.

 

[Lamprologus]

Talk about down the rabbit hole lol. I have the JBL 4349s - initially I hated them, they were far too bright. Tried them with a Primaluna evo 400, MC462 and a Marantz 2275... and they sounded like drek on each one. But I disliked the JBL 4367 too, which I believe measured really well? I do however LOVE the 4429 and the 4365. I was right about to sell the 4349s when I tried one more amp on them and that made all the difference. They image beautifully and scale up very well... on this amp, the 4th amp, and to my ears. Thank god cus i really didn't want to ship them. The amp that paired well was the Line Magnetic-845 Premium. I can sit and listen for hours now - before it was immediate listener fatigue in 15 minutes or less. I still think the 4429s are the better speaker though - hands down. To my ears.

Do you find that you can play loud without loosing control?

I’m looking for something that can tame the brightness of these speakers but I’ve not considered low power tube amps as I like to play loud occasionally.

 

[111db]

It seems to me that this "reference" only applies for unamplified music but most music nowadays IS amplified. Jazz concerts are amplified, even if some of the instruments are not, they are amplified for you to listen to them. Diana Krall (to pick one name) will never sound the way she sings without microphones because we likely never heard her without them.

I remember well the live performance market feeding frenzy in the '70s to obtain Shure SM58 mics "just like on Midnight Special". Shure couldn't keep up with the demand. We are so accustomed to the plumped-up sound of a human voice as captured by a cardioid mic at close proximity that we consider it to sound unnaturally thin without that coloration.

 

[111db]

I have just reviewed this entire thread with great interest. It is for these kinds of discussions that I am here! As a mechanical engineer and DIY speaker guy, I am particularly interested in understanding the relationships between measured parameters and perceived performance, with the goal of gaining greater control over the design objectives of my projects. I hope to find more discussions of this kind as I "mine" the forum, but in the meantime, here are a few of my reactions to this topic and previous comments.

I have long felt that one of the differentiating factors between speakers was large-signal capability - perhaps some combination of compression and distortion. But as more reviews become available that include these measurements, the differences between speakers are smaller than I expected (except for very small speakers). Peter Aczel mentioned the concept of "wave launch" on a number of occasions, in the context of what makes a big speaker sound big, but I don't know of any way to quantify this or of any research that backs it up.

I think upper-bass and low-mid directivity has a lot to do with the apparent 'bigness' of a speaker too. I remember the 8C sounding bigger than they are, perhaps because of the cardioid constant directivity.

I think @Kvalsvoll was the one who talked about this once, about how low end directivity also affects the transient response and that you can't completely match tonality for both continuous signals and transient signals for speakers of different sizes.

And if course, more controlled directivity on the lower mids and bass means less adverse room interaction.

"Wave launch" is a welcome and picturesque term for something I have been thinking about for some time, namely, what is it about a really big woofer that---all else equal---adds dynamic vitality to the sound? We understand well that for midranges and tweeters a large diaphragm causes increased beaming at higher frequencies, and we try to avoid that. But for woofers the same principle applies, and beaming from a big driver (or horn mouth) can be a good thing, concentrating midbass energy toward the listening position in a taut, powerful wavefront. I also suspect that boundary turbulence and/or diffraction occurs at the edges of direct radiator diaphragms just as at the sharp edges of bass ports, and the ratio of edge circumference to radiating area is proportional to the amplitude of that "noise" in relation to the signal.

I also need good imaging, which probably implies good manufacturing tolerances as much as anything else.

My most recent system tweaks have resulted in a heightened understanding and expectation of how beguiling a spatial projection from two speakers can and should be. As a result, I have been able to listen to many classical recordings with a freshly critical ear to the effectiveness of the recording setup in capturing a convincing and pleasing sense of space and localization/separation. I think that a significantly overlooked speaker performance parameter is the sample-to-sample consistency of frequency- and time-domain performance, an outcome of tight manufacturing controls and/or meticulous matching of individual component pairs. "It takes two" to directionally image, and the fewer the disagreements that occur between speakers (and reflections), the more focused the imaging can be. DSP active multiamping can help in this regard by eliminating crossover components that would otherwise require close matching.

Finally, I do wonder whether there are physical performance parameters---or proxies thereof---that have significant influence on the listening experience, but that are not currently identified or measured so as to be potentially brought under control by a designer. Maybe a potential new topic? Or an old one that I haven't yet discovered?