Announcement: ASR Will Be Measuring Speakers!

[FrantzM]

To me this is a turning point in Audio reviewing and testing. For more than 50 years ( Really !: 1970 ~ 2020) we have witnessed the rise and acceptation of subjectivism and relativism (whatever you wallet can bear) in Audio reviewing. There has been a conscious effort to push back against any semblance of objectivity in Audiophile circles, to the extent that "Hi-Fi" became a derogatory expression. This has led to confusion. All efforts have been put forth to lead the public to believe there cannot be measurable when it comes to audio. Subtly but efficiently price became the metric. This has led to countless >$200,000 speakers, >$100,000 DACs. amps, and to me the epitome of stupidity >$50,000 speaker cables>$10,000 USB cables and $14,000 power cords.

NO MORE!!

ASR has become the standard bearer for DACs. Speaker Testing was the necessary next step in the evolution (and return to basics) of Audio reviewing. I, for one, am thankful.
Congratulations Amir! Years from now, the Audio reproduction field will remember this day.

 

[Indien29]

I have a feeling that you will find some truly awful speakers that no amount of DSP and room treatment could improve significantly, are should be avoided to avoid the old problem of ending up trying to "polish a turd." As with DACs and Amps, I assume that there are idiots out there who have no business designing and selling speakers, and some of those will be "outed".

During the spinorama measurement analys, it would be easy to see the homonegeity of the in axis / off axis measurement who are the most important think, linearity of the global curve can be in this case, easy to be linearised with DSP / biquad of FIR filter correction.

With a important difference from the "in axis" and "off axis", no chance to be lucky to make a good corection, the speaker will be not a good speaker, this is the most important think that the Klippel / Amir work will show and what, from my side, I'm espected to visualise, thank's to Amir !!!

 

[briskly]

This should not need to be said, but CEA-2034 data should be present. The same goes for electrical impedance as this can be measured.
Representing the SPL map on the ring around the vertical and horizontal axes will make it easier to compare to existing measurements. I think this will be more familiar for DIY'ers. The NFS system offers an expedient method to extract more spatial resolution than typical 10-15 degree rotations, which make it more suited to revealing Airy/Fraunhofer patterns drivers trend to as ka>1, as well as other interesting behaviors of waveguides and cabinet diffraction. I would not suggest axial normalization here.

When I think of nonlinear testing, what I think of is characterizing the weak points of the system, which will not be easily compared between varied driver configurations. You can certainly try the standard harmonic sweeps, and even a possible test of non-coherence, whatever good that will do. Note that harmonic distortions will collect on the central axis of each radiator. How will you assure that the mic+pre linearity is sufficient to well characterize the DUT?
I would suggest ignoring joint time-frequency representations (STFT, wavelet, Wigner-Ville, etc.) as they don't seem to produce data better shown only in the frequency domain. I'm sure that someone will be interested in seeing the acoustic phase and the impulse response, but I do not believe these necessary either.

 

[Shadrach]

That's an impressive commitment amirm.
In order to evaluate speakers you will need some. I can't see many people sending 100kg packets, or entrusting their sometimes thousands of pounds worth of audio investment to a delivery system. Where are you going to get enough speakers to test to make the project useful?
Do you plan to contact manufacturers to see if they will supply samples?

 

[tkr]

How does this system rank in terms of accuracy?

For instance:
• Stereophile uses quasi-anechoic which is only valid above ~350Hz, and is plotted using connect the dots (meaning actual measurement points across the band, not measuring all frequencies).
• SoundStage/NRC are anechoic and are actual sweep measurements, but only valid above ~100Hz due to small chamber size.

Are these measurements equivalent to an anechoic chamber and is fully equivalent across the audible band?

I own speakers from Nubert and when they moved into an new building they have been very happy about their anechoic chamber. To keep the size and cost of it down they build it as a half-room (not sure if that is even an english word, but I have found no better translation for "Halbraum") doing ground plane measurements, which gives them results down to 70 Hz and can be expanded down to about 30 Hz with some tricks and tweaking applied. There are good pictures from the build phase showing the size of the damping elements needed to achive These results (https://www.nubert-forum.de/nuforum/viewtopic.php?f=1&t=42573&hilit=schalltoter).

To me as a hobbyist it seems that there has been quite a lot of Innovation going on regarding measuring and designing speakers in recent years. Before that it seemed rather static in comparison. If more companies design good speakers as they can measure them more easily with e.g. Klippel, that is great news.

Unfortunately it will be rather hard for speakers that are only sold in specific regions to make it over the pond to be measured in comparison. My floorstanders exceed the size limit for normal DHL parcels (> 120 cm in one dimension) in their packaging and another parcel service would charge its selling price for shipping both ways.

 

[Juhazi]

I want to thank and congratulate Amir too! I've been sceptical about this project, but we will see where this leads to. The basics of loudspeaker design and acoustics is well known, and an educated guesser can deduce a lot by just looking at a construction( eg. baffle size, driver diameter, xo points). Small details are important too, there are many ways to spoil a a basically healthy concept. Manufacturer's advertizing folklore is another issue.

What I woul like to see most, is measurements of speakers that deviate from the mainstream - horns, dipoles, omnis etc. Many acoustic tricks they do don't show up in "normal" frontal on-off-axis or room response measurements. These are often advertized with glorious and extravagant claims of superiority, which should be investigated independently and with precision and reliability which this Klippel scanner now provides.

Seems like an interesting year 2020!

 

[Nightlord]

This is wonderful news indeed!

In my crazier dreams of making my own speakers, I've been pondering building something similar to this, but in the slower manual sense. Great to see that equipment like this exist and that one now made it into the hands of a fellow 'lunatic'.

 

[Soniclife]

I do plan on performing listening tests however.

That will be interesting, and important to see how preference aligns with the predicted preference. Are there people you could recruit to a listening panel if you find a few interesting speakers worth comparing?

 

[Soniclife]

Amir has finally lost his mind!

I'm preferring the word eccentric. Reminds me of the Victorian era when men went to their sheds to do science, and many great things came from it.

 

[bunkbail]

Wow. You're insane, Mr Amir. Didn't expect you'd follow through about this whole speaker testing this soon. Congratulations on getting the Klippel machine!

Any idea which speaker will be the one to be featured on the 1st review? This is so exciting!

 

[PaulD]

This is a wonderful initiative Amirm - CONGRATULATIONS!!! Your own personal speaker testing Darlek!

As for summarising and presenting the data, I think it will need to be multidimensional, something like a linearity figure across octave (or 1/3rd octave) bands, and the linearity includes frequency magnitude linearity as well as distortion (THD, IMD etc, as you find appropriate).

There is no doubt that speakers (and probably acoustics) are the weak link of the reproduction chain, so this initiative could be very important. I am lucky to have a suite of professional studios I operate that have controlled acoustics and professional monitoring (such as the M2s) and I find that I am very sensitive to speaker distortions. I also find that most people have never heard very low distortion speakers, so they have no idea how they sound. I often demonstrate our mastering room to people just so they have that experience. Your efforts here could literally revolutionise this for the public, allowing them to reliably find the highest linearity speakers - I certainly hope so!

 

[urfaust]

Amazing news. Good luck for all the preparation and setup work.

I can't wait to see some of the nearfields i know. On a sidenote, Is there a way the speaker's distorsion could be correlated in those measurements?

 

[tomtoo]

....

Any idea which speaker will be the one to be featured on the 1st review? This is so exciting!

Wait, let me look into my magic glass bowl. Ah fuc*ing distortions over the ocean today. Let's finetune the Picture.

But wait i see a ordinary Box, and sry a moment pls, it's black. Maybe Batman's speaker? But small??! Ahh i see a Horn? Nah it's just a waveguide. No compression driver. What could it be? ; )

 

[bunkbail]

Wait, let me look into my magic glass bowl. Ah fuc*ing distortions over the ocean today. Let's finetune the Picture.

But wait i see a ordinary Box, and sry a moment pls, it's black. Maybe Batman's speaker? But small??! Ahh i see a Horn? Nah it's just a waveguide. No compression driver. What could it be? ; )

S400?

 

[anmpr1]

Good luck. I mean it. If you haven't already (and I'm sure you have) I'd review the work done by Richard Heyser, Don Keele and Floyd Toole, etc. These men have set a standard.

Measuring speakers is a big challenge. Especially large floor standing speakers, plus unusual speakers with differing 'wave launches'--speakers that use room reflections as part of their design, dipoles, subwoofers.

I am really looking forward to your work in this area. Do you contact manufacturers for test samples? Don't hurt your back lifting.

 

[arboleda]

This is awesome. Up’d my Patreon level to help a bit. Can’t wait. I imagine Amir will iterate and learn, and this will be a very valuable source of information.

 

[RickSanchez]

This is a fantastic initiative! Thanks Amir!

 

[snapsc]

Amir,

Congratulations and Good Luck....you have undertaken a daunting task...but one that is really needed; shining a light on how speakers really measure so that later there may be some correlation with how they sound.

One thing that always seems to be missing is a way to simulate different room sizes so that people could get a better idea of how a speaker might measure and de facto how it might sound in their own room.

Does the system have any capability to run such simulations as I'm pretty sure that lugging it from room to room is out of the cards.

Thanks again for undertaking this project!

 

[bobbooo]

I do plan on performing listening tests however.

It would be good to see how subjective impressions match up with objective measurements, but unless the listening tests are blind, they won't have much scientific merit, due to expectation bias etc. You could eliminate one source of expectation bias at least by doing your listening tests before seeing any measurements of the particular speakers you're testing, so your subjective impressions aren't influenced by objective measurements (the same goes for your DAC/amp etc listening tests).

Thanks so much for expanding your tests to speakers, it's a monumental project that will be of huge benefit to us consumers.

 

[pozz]

It would be good to see how subjective impressions match up with objective measurements, but unless the listening tests are blind, they won't have much scientific merit, due to expectation bias etc. You could eliminate one source of expectation bias at least by doing your listening tests before seeing any measurements of the particular speakers you're testing, so your subjective impressions aren't influenced by objective measurements (the same goes for your DAC/amp etc listening tests).

Thanks so much for expanding your tests to speakers, it's a monumental project that will be of huge benefit to us consumers.

Just FYI, the types of speaker measurements Amir is proposing to make have a very good correlation with listening tests.