Announcement: ASR Will Be Measuring Speakers!

[dukanvadet]

I do hope that all y'all are prepared for the quantitative values that conventionally loaded loudspeakers (BR, TL, acoustic suspension, infinite baffle, etc.) using conventional dynamic drivers (cone, dome, planar) are gonna return for harmonic distortion and electrical efficiency (as opposed to sensitivity -- although most modern loudspeakers are pretty insensitive as well)!

You might even discover why some of -- ahem -- us either came back to, or never abandoned horns.

In all seriousness, it will be really nice to see some quantitative assessment of the kinds of loudspeakers that don't make it to JA's test bench @ Stereophile.

Two metrics that are not very important.

 

[Juhazi]

Two metrics that are not very important.

Well, I have read many complaints and questions, why the highly acclaimed KEF LS50 doesn't sound good in a large wooden villa or loft. Stress test with high spl can reveal eg. a struggling midwoofer or too low and soft xo for the tweeter. Canucks and some Krauts do measure distortion at high spl.

LS50 90dB/2m (10.6V)

R3 90dB/2m (7.87V)

Source of measurements https://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=16&Itemid=140

 

[Mashcky]

I’m thrilled for the future of ASR. This audio hobby was insanity for me before reading Dr. Toole’s book and joining this forum. Gone are the days of struggling to find a competent DAC. Some day it will be the same for speakers! As many have said, speaker testing is an imperative for the consumer and the industry.

I have decided to become a supporting member in a small way commensurate with being a graduate student, and plan on continuing support in the future however I can.

@amirm I hope this project is deeply fulfilling – if not financially then personally. I’m looking forward to it!

 

[maxxevv]

Not sure if I missed it in the past 12 pages of this thread, perhaps we should have a list of speakers with a broad appeal and generally "affordable" range for starters ?

I know @amirm has started testing on the ubiquitous JBL 305p's. Popular powered/ passive monitors in the sub-US$500/- range would be of great interest to a lot of people I think ?

 

[bobbooo]

One slight problem is that i think that would give an advantage to speakers that can be used fullrange. A smaller speaker that are meant or likely to be crossed over to subwoofer may get a bad score that would turn out really good once you use it with a sub?

This is a very important point. Very few speakers, even full-range floor-standers, provide full extension down to 20 Hz. A separate subwoofer would be needed in most situations to produce true full-range extension, and so it would be these 2.1 systems that would likely score the highest.

I think it would then make sense to measure each speaker not only in isolation, but also in conjunction with a good subwoofer that has already been measured to perform well, to judge the speakers' full sonic potential in a realistic set-up. @amirm would this be doable with the Klippel system?

Alternatively to measuring speakers and subs together, maybe you could have two separate scores for speakers - one that encompasses the full 20Hz to 20kHz, and another that has a greater lower-limit (the standard sub crossover frequency of 80Hz might be a good choice), so their performance in both 2.0 and 2.1 systems can be accurately judged.

Also, in what frequency range will subwoofer performance be judged when they are measured in isolation? Here I'm thinking about the upper limit i.e. will a sub that extends not just lower, but higher, produce a better score? If so, just like penalizing speakers for lack of low-end extension when they were designed to be and perform best coupled with a sub, I think it would be equally illogical to penalize a sub for lack of upper-bass extension.

 

[mhardy6647]

Two metrics that are not very important.

Actually, I don't disagree -- except in the case of doubling in the bass, which can still be a problem.
I care about both sensitivity and efficiency, but I am a tad bit out of the mainstream when it comes to amplification.

 

[MZKM]

Bass region yes, a speaker can be designed to use wall support for instance. Also I was more thinking of wide vs narrow dispersion even if the dispersion is smooth.

That likely should't be factored into sonic quality, as that is more there for those in narrow rooms or wide rooms.

 

[dukanvadet]

Well, I have read many complaints and questions, why the highly acclaimed KEF LS50 doesn't sound good in a large wooden villa or loft. Stress test with high spl will reveal eg. too low and soft xo for the tweeter. Canucks and some Krauts do measure distortion at high spl..

I dont think anything in this post proves that those subjective complaints are because of high THD.
If I where to guess i would say that since the ls50 uses a coaxial in a two way so there will be higher doppler distortion when the cone starts moving. Does the same problem appear with the 3-way with the same driver?

Edit: i saw one of the graphs where for the 3-way so i assume it has a steeper XO. My guess is still a plausible explanation if the 3-way works.

 

[Juhazi]

Interesting that in coffee’s instance its basically the inverse of Harman’s listening preferences. The purest form of the coffee (neutral) loses to sweetness. Sound and taste aren’t a perfect corollary however.
Another interesting thought is that if you preferred dark coffee sighted but picked milky coffee blind, would you then deny your sighted preference and drink it milky? Why shouldn’t sight factor in if you’re drinking it sighted.

Here is an interesting graph of preliminary studies for the Harman curve. Untrained listeners prefer loud bass! "Training" will change the preference for smoother response, but isn't this just manipulation? A bit more in Sean Olives blog

 

[amirm]

Well, if it’s made available, you could see what the correlation would be to the subjective ratings in Harman’s listening tests and see if it does better than the 86% that they achieve.

But I don't have access to preference rating of some 100 speakers Harman used. They have only published data on a few speakers that way.

 

[Nightlord]

Here is an interesting graph of preliminary studies for the Harman curve. Untrained listeners prefer loud bass! "Training" will change the preference for smoother response, but isn't this just manipulation? A bit more in Sean Olives blog

View attachment 44940

Can't help wondering why the common point was picked like that. If you matched them as closely as possible over the entire range, it would seem that trained listeners preferred more midrange... Lies, damned lies and statistics....

 

[Juhazi]

RTA curves of a single speaker were used at Olive's study, if I remember right. In the graph they were matched at 500Hz. Anyway, tonal difference between curves is huge, as anyone who has played with eq and measurements knows.

If untrained listeners' curve is wrong, isn't it as crazy as saying that voters of Putin and Trump were wrong?

 

[MZKM]

But I don't have access to preference rating of some 100 speakers Harman used. They have only published data on a few speakers that way.

Maybe kindly sent such a script to Sean and ask him to kindly run it ; it would also benefit his research in getting a higher correlation score.

Wishes aside, my critique still stands; that to improve upon an 86% correlation, some weighting of performance needs to be done based on frequency, and that weighted THD should be included (albeit with a low impact on score). But yeah, 86% ain't bad, so if that's what you want to use for now, it still will be pretty meaningful.

I don't think you need to have different scoring for towers and bookshelf speakers as others are suggesting, as objectively the tower version of bookshelf (all else equal) will perform better. Because it wouldn't stop after that, what about having another scoring list for bookshelves that weigh vertical performance more heavily for those looking for near-field usage.

 

[MZKM]

RTA curves of a single speaker were used at Olive's study, if I remember right. In the graph they were matched at 500Hz. Anyway, tonal difference between curves is huge, as anyone who has played with eq and measurements knows.

If untrained listeners' curve is wrong, isn't it like saying that voters of Putin and Trump were wrong?

I see your point, but those spending hundreds/thousands/tens of thousands on speaker gear (and high end headphones) are a different market segment than the high schoolers with 18" subs in their car that is so loud it rattles the windows.

 

[murraycamp]

They have only published data on a few speakers that way.

Enough to at least confirm consistency with proposed methodology?

 

[amirm]

Enough to at least confirm consistency with proposed methodology?

Unfortunately not. You need a lot of sample data to curve fit to half a dozen variables. I am going to keep thinking about this though.

 

[GD Fan]

This is very intriguing news - sufficiently so that I finally joined up as a new member to contribute to the project. Hopefully many others will follow suit! Best of luck.

 

[williamwally]

This is very intriguing news - sufficiently so that I finally joined up as a new member to contribute to the project. Hopefully many others will follow suit! Best of luck.

Same here!

 

[Bruce Morgen]

JBL Paragon.

I have to laugh at seeing that thing. I mentioned to Manny, the white-aired audio buyer at Arrow Electronics where I worked in the showroom circa 1970, that I'd love to hear one of those monsters. He replied in his thick NYC accent (Manny pronounced "Garrard" as "gay-rard") that the one time he'd heard one it sounded simply awful -- edgy and rapidly fatiguing -- but that the woodwork was gorgeous.

 

[rgpit]

I have to laugh at seeing that thing. I mentioned to Manny, the white-aired audio buyer at Arrow Electronics where I worked in the showroom circa 1970, that I'd love to hear one of those monsters. He replied in his thick NYC accent (Manny pronounced "Garrard" as "gay-rard") that the one time he'd heard one it sounded simply awful -- edgy and rapidly fatiguing -- but that the woodwork was gorgeous.

I agree with Manny's opinion on the edgy fatiguing sound of the JBL Paragon. They were a beautiful piece of audio furniture. Coincidentally I retired from Arrow Electronics a few years ago. I was a field sales engineer in their industrial electronics division.