• 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!

Ideas for more meaningful speaker measurements

Holmz

Major Contributor
Joined
Oct 3, 2021
Messages
2,018
Likes
1,241
Location
Australia
Give the Highland Brewery a visit. One of 67, but I find their ales to be the best in the state.
Thanks!


All that is technical, measurable, tunable. What is not is how your brain inherent, genetic, and training, processes what you hear to decide if it is "musical" or not.
Well either god put us here ~6500 years ago, or we evolved.
If we assume the former, then genetics are not a factor.
If we assume the later, then we have some percentage of a million years, added to the millions of years from the preceding species.

I am in the later group, and even people that do not know how to read can pull out the sound in a crowded and noisy room… without trying.


And was alluded to, when you walk into a room, your brain takes a pattern from the ambient and biases what you hear with what it expects. So, a perfectly flat response in a very live or very dead room will confuse your poor brain.
I am often confused, but rarely with the sound in a room.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,522
Likes
7,041
Location
Stow, Ohio USA
Since it recently came up elsewhere, if you listen before you test and then post results after you measure, it will appear to the casual reader no different than if you measured first...

I originally penned this thread focused more purely on measurements and how their presentation affects one's perception. Notably, Stereophile reviews tend to be done with listening tests done by one reviewer and measuring done (later) by another. This seems a further step towards avoiding potential bias from measuring first.

Was wondering how intentional this might be and whether anyone noticed comparable approaches by other reviewers?

I do not seek out reviews much beyond ASR, but used to do more so in the past.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,522
Likes
7,041
Location
Stow, Ohio USA
Audioholics seem to be listening first and measuring second, for what it's worth.

Thanks!

So at least there may be an intent to avoid potential bias. Agree AH is pretty thorough about how they measure a speaker...

Less clear is their policy on the overall review. I did peruse the site. If you find something more, let me know.
 

sigbergaudio

Major Contributor
Audio Company
Forum Donor
Joined
Aug 21, 2020
Messages
2,640
Likes
5,404
Location
Norway
@Rick Sykora What are you looking to understand specifically? I guess I can ask if James (who do many of their reviews and most of the measurements as I understand it) is interested in sharing here (he has our speakers on review at the moment).
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,522
Likes
7,041
Location
Stow, Ohio USA
@Rick Sykora What are you looking to understand specifically? I guess I can ask if James (who do many of their reviews and most of the measurements as I understand it) is interested in sharing here (he has our speakers on review at the moment).
Thanks again!

Would want to know whether AH has a policy of listening AND writing that part of the review BEFORE measurements are done?
 

tuga

Major Contributor
Joined
Feb 5, 2020
Messages
3,984
Likes
4,281
Location
Oxford, England
Understood, while not perfect, my point was to ensure that more proven measures get priority, My OP for this thread though is an example of one challenge. The established FR graph is misleading but not likely we are going to get an industry committee to agree on the more realistic representation. But maybe ASR can provide a catalyst for some meaningful change.

In what way is the established FR graph misleading? (are you qualified to make such a claim)

My view is that one must learn to a) interpret the graphs and b) to correlate them as best as possible with listening.
That is our job, a difficult one at that, which requires reading and listening.

Alternatively one can only rely on the information appended to the graphs and the comments that accompany it, which may not be correct and/or sufficiently extensive.
 

tuga

Major Contributor
Joined
Feb 5, 2020
Messages
3,984
Likes
4,281
Location
Oxford, England
I'd like to hear the speakers. We have done this before on ASR, where 4 speakers of unknown brand were recorded playing snippets of sound. It was an interesting exercise. Another You Tube video shot in an audio dealership compared two high end loudspeakers of quite different proportions that was remarkable in that they sounded so much alike. In both these instances the recordings were made using affordable gear.

Listening to recordings of speakers is a worthless exercise. You either listen to them in situ or you don't. Even listening in an unfamiliar environment is an exercise of limited worth.
 

tuga

Major Contributor
Joined
Feb 5, 2020
Messages
3,984
Likes
4,281
Location
Oxford, England
...or use narrow directivity speakers, reduce all reflections below threshold and use an upmixer to taste. That's what I'm doing – or trying to do. Wide(ish) dispersion speakers are and will always be just a band-aid for the intrinsic limitations of stereo.

“Habits die hard. The introduction of stereo in the 1950s gave us a left-to-right soundstage, but close-microphone methods, multitracking, and pan-potting did nothing for a sense of envelopment—of being there. The classical music repertoire generally set a higher standard, having the advantage of the reflectivity of a large performance space, but a pair of loudspeakers deployed at ±30° or less is not an optimum arrangement for generating strong perceptions of envelopment (as will be explained later, this needs additional sounds arriving from further to the sides). Perhaps that is why audiophiles have for decades experimented with loudspeakers having wider dispersion (to excite more listening room reflections), with electronic add-ons and more loudspeakers (to generate delayed sounds arriving from the sides and rear). In the control room, audio engineers now employ spatial simulation algorithms to break the monotony of a frontal soundstage by adding some low-correlation spatial sounds, even including binaural crosstalk-cancellation. All have been intended to contribute more of “something that was missing” from the stereo reproduction experience that, even at its best, serves only a single listener.
The real solution is more channels, giving the capability of delivering anything from a single point image through to spatial envelopment and immersion, as appropriate.

Excerpt From
Sound Reproduction
Floyd E. Toole

Toole writes "more channels" but he means "more speakers".
The music industry is not giving us more channels.
 

Holmz

Major Contributor
Joined
Oct 3, 2021
Messages
2,018
Likes
1,241
Location
Australia
Toole writes "more channels" but he means "more speakers".
The music industry is not giving us more channels.

I thought he advocated mono speaker listening?

Your post reminds me of the Lena Lovich song, “Lucky Number.”

 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,522
Likes
7,041
Location
Stow, Ohio USA
In what way is the established FR graph misleading? (are you qualified to make such a claim)

My view is that one must learn to a) interpret the graphs and b) to correlate them as best as possible with listening.
That is our job, a difficult one at that, which requires reading and listening.

Alternatively one can only rely on the information appended to the graphs and the comments that accompany it, which may not be correct and/or sufficiently extensive.

This OP is my opinion. Agree with yours, but just suggesting that there is room for improvement on the scaling.

I can justify the standard scaling somewhat as it includes the bass rolloff. My claim about the scaling is not disputable. You will experience it if you try for yourself on most any speaker (including the Directiva ones, btw). It is not about the measurement system capability, the variability is simply more apparent due to the different vertical spacing. As mentioned, in earlier times, amplifier manufacturers tried the same approach to make the response appear flatter too.

If the manufacturers do not provide the scaling with the shorter vertical range, you are not going to be able to readily see the difference.
 
Last edited:

tuga

Major Contributor
Joined
Feb 5, 2020
Messages
3,984
Likes
4,281
Location
Oxford, England
This OP is my opinion. Agree with yours, but just suggesting that here is room for improvement on the scaling.

I can justify the standard scaling somewhat as it includes the bass rolloff. My claim about the scaling is not disputable. You will experience it if you try for yourself on most any speaker (including the Directiva ones, btw). It is not about the measurement system capability, the variability is simply more apparent due to the different vertical spacing. As mentioned, in earlier times, amplifier manufacturers tried the same approach to make the response appear flatter too.

If the manufacturers do not provide the scaling with the shorter vertical range, you are not going to be able to readily see the difference.

I think that the 50dB-wide vertical scale is informative enough, it is more or less standard too.
Too much smoothing is common though, but that is not a problem with Amir's measurements.

I would like to see an In-Room response measurement to compare with the PIR but te speaker would have to be positioned as if part of a symmetrically-placed stereo pair.
I like the set published here: https://www.fidelity-online.de/avantgarde-acoustic-uno-xd-messungen/
It would be nice to see measurements of the amplification inside active speakers.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,522
Likes
7,041
Location
Stow, Ohio USA
I think that the 50dB-wide vertical scale is informative enough, it is more or less standard too.
Too much smoothing is common though, but that is not a problem with Amir's measurements.

I would like to see an In-Room response measurement to compare with the PIR but te speaker would have to be positioned as if part of a symmetrically-placed stereo pair.
I like the set published here: https://www.fidelity-online.de/avantgarde-acoustic-uno-xd-messungen/
It would be nice to see measurements of the amplification inside active speakers.

While admittedly my OP started with a small incremental improvement (and one that is easy for a wide audience to comprehend). If manufacturers either do not measure or do not share detailed measurements today, does not seem likely they are about to sign on to supply more measurements (unless it clearly benefits them). Today, most major manufacturers do not even provide the 50 dB scaled frequency response. From my perspective, it a is major compromise and we still have yet to see it widely published.

While frequency response linearity is not everything, it has been well-studied and accepted. We all know the speaker is (by far) the weakest link in the audio chain with respect to frequency response. In the last 3 years of Amir's Klippel measurements, I saw more measurement progress than I have of 50 years of manufacturer measurements. So while progress in one respect, it represents only a small fraction of all the different speakers marketed today.

So while am open to discussing what Amir (and other reviewers) might do better too, seems to me that unless motivate more manufacturers get on board, will not likely get better data in the next 50 years. Am just hoping I can hear well to enjoy this hobby for at least the next decade! :)

Hopefully, I have articulated my intent better than in my OP. I have made a small edit to the OP to clarify.
 

sigbergaudio

Major Contributor
Audio Company
Forum Donor
Joined
Aug 21, 2020
Messages
2,640
Likes
5,404
Location
Norway
We also have (sort of)the opposite problem, that with too much and detailed measurements, they are misinterpreted or its significance overrated. not necessarily in speakers, but certainly in the coveted SINAD of DACs and amps. By reading on ASR it's easy to think that if an amp or DAC doesn't have at least 100dB SINAD, you might as well throw it away, and it will be unlistenable.

On a Norwegian forum we recently did an interesting ABX-test. A short track was sent through an DA/AD loop through a Yamaha LS-9 mixing console. So Laptop-> AES (Digital) -> Mixer -> XLR Analog out -> XLR Analog in -> AES (digital) -> Laptop.

This was done in a loop 10 (!) times. The result? No one was able to successfully ABX the difference between the tracks.

The test was then repeated and the track was looped through the DA/AD loop in the mixer 100 (!!) times. It was now possible to do a successful ABX test, but you still had to concentrate to tell the tracks apart.

Obviously the SINAD after those 100 rounds was not great. And those worried about introducing DSP in their chain for instance to enable EQ probably won't need to worry. :)

1649657787107.png
 

sigbergaudio

Major Contributor
Audio Company
Forum Donor
Joined
Aug 21, 2020
Messages
2,640
Likes
5,404
Location
Norway
Thanks again!

Would want to know whether AH has a policy of listening AND writing that part of the review BEFORE measurements are done?

Got an answer from James that he agreed to being shared:
In my view, I am going to have preconceived notions about a loudspeaker no matter what, so I don’t concern myself with the order in which it is heard. If I look at a design, I will form some inescapable assumptions about it even if I don’t have hard measurements of its performance. If I see its measured performance, that is going to just alter those preconceptions in another way. There is no way around those expectation biases. So I just try to listen with as open mind as possible, regardless of whatever I know or think I know about the system’s performance.

He also mentioned that he and Gene has an interesting youtube video (originally a livestream) that specifically talks about these issues:
 

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,923
Likes
7,616
Location
Canada
I admit I don't really understand the fixation on listening before measuring. Like, I guess vice versa creates some bias? But so do 100 other things. You're still doing sighted tests, rendering your listening test questionable AT BEST...

If a reviewer tells me they listen before measuring I just think "So what? You eliminated 1 bias, what about the other 99?". it's just totally irrelevant.
 

gnarly

Addicted to Fun and Learning
Joined
Jun 15, 2021
Messages
992
Likes
1,390
Hi Rick, a few measurements i would like to see are:

And this one for certain:
The new AES75 standard for max SPL. https://www.aes.org/publications/standards/search.cfm?docID=116
Looks like Meyer's M-Noise effort has become a potential standard. https://m-noise.org/

I continually wonder how much linear SPL , including dynamic headroom, is available. Standard compression tests seem to be geared more to max average SPL, and fall short of measuring linear peak capability.
From my experiences with home and pro gear, I think max linear SPL, both average and peak, is a big component of SQ.


Then these for digging deeper:

Some form of automated multi-tone test, to get a better sense of how audible/important IMD is.
Joseph Crowe for example, has an interesting simplistic method. https://josephcrowe.com/blogs/news/distortion-and-sound-quality-what-is-reasonable

I've begun playing with multi-tone, and wavelet bursts, trying to dig deeper into why boxes with the same tuning & same polars, can still sound different. (MEH's using the same horn but different driver complements, tuned for the same on & off axis response.)

Also been trying measurements like found in ARTA's Speech Intelligibility section for the same just stated purpose.....MTF, STI, etc
Measurement junkie i am ...:)
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
Even if one were to generate a swept sine measurement at 0 dBFS (my own set max reference level) and thereafter play back music at maximum volume... the actual transient peaks acoustically measured esp. in the bass will exceed the level as seen in the former. I may have forgotten the reason for this -- disregarding intersample peaks for now -- hmmn... why is this the case? Is it that some crest factor also needs to be accounted for?

1649695752532.png

Volume Leveling to reduce volume of streamed audio content disabled in JRiver (normal EQ still on)
*Purple traces are individual left and right response. Red trace is both L+R swept sine at 0 dBFS

LZpeak value (likely mostly from bass) is way higher than what I would've expected.
 
Last edited:
Top Bottom