Those of you who believe measurements aren't the whole story, do you have a hypothesis why that is?

[Rottmannash]

Not any more, so you're safe. Doing different stuff now, though still high-speed analog.

And some of it was satellite stuff, and work on the space shuttle's radar, so destruction was not always involved.

"We're not worthy!"

 

[NTK]

For those who are interested in going deep into the subject of spatial hearing, Jens Blauert's book is open access at MIT Press.
https://direct.mit.edu/books/book/4885/Spatial-HearingThe-Psychophysics-of-Human-Sound

 

[restorer-john]

But this can be measured with the Klippel. Amir only measures one speaker, but I don't see this as the NFS being lacking.

Can you point me to some Klippel analysis of a pair of speakers reproducing an acoustic image across the spectrum? Interested.

 

[Rottmannash]

You mean you're not about fixing it until it's broken?
How did a shameless heretic like you get in here?

Shouldn't he be banned for his lack of esoteric, exotic and exorbitantly expensive gear?

 

[dshreter]

I think a useful way to think about “imaging” in speakers is in the terms convolution and deconvolution from signal processing.

Convolution is basically the transfer function applied to a signal. Instruments or a performance are the signal in music, which is convolved through the room, the microphones, and the mixing which all act as a transfer function upon the source signal that results in the recording.

Deconvolution is the application of an inverse transfer function to infer the source signal. When we listen to something we do this through inference and can conceptualize what the source signal was, and can do so for a wide variety of transfer functions. Imagine a bird chirping played back on a hifi stereo or a telephone speaker, we have no problem deconvolving what we are hearing.

Most transfer functions (convolution) are lossy to some extent. After you apply the inverse transfer function you don’t have full fidelity to the source signal. I would describe imaging as the lossiness of a system’s transfer function. If it can be fully deconvolved, retaining the spatial and other subtle details of a performance’s sound, then it has excellent imaging.

These concepts are extensively studied in signal processing. Advances in this space are part of how nearly magical technologies like MRI are possible. These are mathematically solvable problems.

 

[Blumlein 88]

I think a useful way to think about “imaging” in speakers is in the terms convolution and deconvolution from signal processing.

Convolution is basically the transfer function applied to a signal. Instruments or a performance are the signal in music, which is convolved through the room, the microphones, and the mixing which all act as a transfer function upon the source signal that results in the recording.

Deconvolution is the application of an inverse transfer function to infer the source signal. When we listen to something we do this through inference and can conceptualize what the source signal was, and can do so for a wide variety of transfer functions. Imagine a bird chirping played back on a hifi stereo or a telephone speaker, we have no problem deconvolving what we are hearing.

Most transfer functions (convolution) are lossy to some extent. After you apply the inverse transfer function you do have full fidelity to the source signal. I would describe imaging as the lossiness of a system’s transfer function. If it can be fully deconvolved, retaining the spatial and other subtle details of a performance’s sound, then it has excellent imaging.

These concepts are extensively studied in signal processing. Advances in this space are part of how nearly magical technologies like MRI are possible. These are mathematically solvable problems.

My stereo images better than MRI.

 

[dshreter]

My stereo images better than MRI.

Yeah, because sound is extremely easy to record in comparison to nuclear spin.

 

[Chrispy]

Very late to the party but read thru a coupla pages....what do you mean "believe" in measurements in the first place. I can see using them of course as a filter but why would I "believe" in them? Is it a mystical thing? If it measures well and then sounds like crap well I'm waiting for that experience still......can't imagine any useful hypothesis/theories etc but find measurements highly useful myself.

 

[Jim Matthews]

Yeah, because sound is extremely easy to record in comparison to nuclear spin.

So - the Radiologist interpreting the scan is a "Spin Doctor"?

 

[aslan7]

I think that everyone would be better off in these discussions if they admitted that the audiophile hobby is prone to extreme neurosis. People imagine--and fret about--all sorts of bizarre things about arcane "improvements" like cables, break-in, Israeli-made power supplies, isolation platforms, tubes, turntable "upgrades", etc. Remember when they used to stick rubber rings on CDs? Whatever happened to that? Then if you pay a lot of money for this stuff you are literally invested in it and, contrary to reason, profess to hear things well beyond the range of human hearing. There is also a lot of "mine is bigger than yours and cost a lot more money" attitude out there. Many people like the paraphernalia more than the music! Most hobby obsessions drift into neurosis but the audio mania ranks high. Honestly, if we all stepped back and thought about some of the crazy things we bought we would laugh at ourselves. (I once had an exotic Kieth Monks tonearm with exposed pools of mercury for contacts). But it is a lot of fun and a great from of escapism.

 

[Robin L]

I think that everyone would be better off in these discussions if they admitted that the audiophile hobby is prone to extreme neurosis.

People have no idea how easy it is to self-hypnotize.

 

[Kal Rubinson]

(I once had an exotic Kieth Monks tonearm with exposed pools of mercury for contacts).

I had one of those and, despite the necessary maintenance, I loved it. In fact, I still have a few parts left.

 

[JJB70]

I think a lot of it originated with turntables. With turntables there is a degree of playing with gear to set them up and optimize performance and it is not just delusional tweakery. When digital came along some manufacturers and magazines etc needed something to step into that and so invented all sorts of weird ideas trying to pretend that digital was like analogue.

 

[audio2design]

@Duke , nice to see you here instead of that other "forum" ....

- Harman's subjectively-preferred steady-state in-room response curves, whether blind-generated by trained listeners or untrained listeners or both or based on extensive loudspeaker preference data, are all non-flat (and non-constant-slope) and counter-intuitive.

- I will tackle this one first because there is an ENORMOUS, i mean HUGE, FUNDAMENTAL, wickedly big flaw in Dr. Toole's and Harmon's preference curve, and testing methodology. I mean giant, brutally large, drive a bus through it flaw. Sorry Floyd.

----> The Harmon Preference Curves ARE NOT tests of speaker response preference. <----


.... let the barbs be thrown.


The huge, ginormous, fundamental flaw of the Harmon preference curves is that they use music as the test signal.


When you use music as a test signal, you are not exclusively testing the speaker that you are listening to. You are also testing the recording engineer, the mixing engineer, the system(s) that the recording and mixing engineers used when developing the music you are listening to, and even the volume they were doing their work at. <---- Let that sink in!

- Most mixing and recording engineers have worse than average hearing both because of exposure and also because most of them are old farts.
- Most mixing and recording is done near field (even if later checked on multiple systems)
- Most mixing and recording is done at a different volume than listening

The "preference" curves for speakers are highly highly influenced by the people, equipment, environment and volume level used in the development of the so called test signal, i.e. music. When you realize that, then the preference curves start to make total sense.

A 5-inch woofer and a 15-inch woofer may have identical frequency response curves but will sound very different even at the same SPL. So we cannot equalize a minimonitor to sound like a big JBL.

- I am pretty confident that if we restricted both drivers to a narrow frequency band that both could cover, a volume level that kept both to lower distortion levels, the frequency band was restricted enough to keep dispersion similar, and you did not listen too near field, then you would not be able to identify a difference that was indicative of size of the driver.

Reports of subwoofers subjectively improving the mids and highs (whether or not the mains are high-pass-filtered), and add-on supertweeters subjectively improving the bass, have no correlation with measurements.

- Unfortunately, the people who often make these claims often don't have a high correlation with reality attachment either, but let's not go too far down that path.

- Assuming tolerable subwoofer integration and the fact that multiple subs (and mains) usually results in a more even bass response with node elimination the net result is both a tendency to use higher listening levels and a lack of masking functions from excessive room modes. Both will result in improved mid and high perception and yes this does correlate with measurements. Flatter room responses which on average does correlate with subwoofer usage, and certainly your subs, will always result in improved perception of mids and highs.

- Supertweeter usage is somewhat rare and any consistency of implementation is totally out the window since people pretty much just slap them in without having any idea what they are doing. Real ultrasonic frequencies (which is unlikely what is coming out of most supertweeters) can create a masking function for audible higher frequencies, but more likely is that the supertweeter is evening out a poor high frequency room response which like room modes at low frequencies can eliminate a masking function hurting overall sound impression. Then again, in this case I am willing to consider placebo.

It is not generally apparent from measurements which horns have "horn coloration" and which ones do not.

- I would counter that it is not generally apparent what horn coloration is or even that it is real and any different from any other coloration that could be assigned to any speaker w.r.t. resonances, etc.

If we have two identical sounds, except that one lasts a little bit longer than the other, the longer one will be perceived as being louder, despite the measured SPL being identical.

- Nothing at all unusual about this. This is true for any human sensory function including light and touch. Duration always plays a part in perceived intensity, but then again, when have intensity and energy ever been the same?

Measurements fail to predict imaging, much less spaciousness. Research by Wolfgang Klippel (cited by Floyd Toole) indicates that “the feeling of space” makes a 50% contribution to “naturalness” and a 70%(!) contribution to “pleasantness”. The virtual uselessness of measurements in this area is arguably a significant shortcoming if Klippel's findings are in the ballpark.

- See my quote above w.r.t. using music as a test signal, and for imaging, that is really really a tough one as what contributes to imaging in real music is all over the map with a preponderance for level based cues, far less frequently timing cues, some artificial (usually) frequency response cues, etc. Spaciousness .... even more of a mixed bag.

- Are measurements useless? ... No, at least in room measurements of direct/reflected can provide information on what you may be able to achieve for both. Every room is different so looking at the speaker on its own is never going to be enough, and most people don't know how their speaker will behave, but they can make some guesses and they can adapt their room to the speaker.

 

[Robin L]

@Duke , nice to see you here instead of that other "forum" ....



- I will tackle this one first because there is an ENORMOUS, i mean HUGE, FUNDAMENTAL, wickedly big flaw in Dr. Toole's and Harmon's preference curve, and testing methodology. I mean giant, brutally large, drive a bus through it flaw. Sorry Floyd.

----> The Harmon Preference Curves ARE NOT tests of speaker response preference. <----


.... let the barbs be thrown.


The huge, ginormous, fundamental flaw of the Harmon preference curves is that they use music as the test signal.


When you use music as a test signal, you are not exclusively testing the speaker that you are listening to. You are also testing the recording engineer, the mixing engineer, the system(s) that the recording and mixing engineers used when developing the music you are listening to, and even the volume they were doing their work at. <---- Let that sink in!

- Most mixing and recording engineers have worse than average hearing both because of exposure and also because most of them are old farts.
- Most mixing and recording is done near field (even if later checked on multiple systems)
- Most mixing and recording is done at a different volume than listening

The "preference" curves for speakers are highly highly influenced by the people, equipment, environment and volume level used in the development of the so called test signal, i.e. music. When you realize that, then the preference curves start to make total sense.




- I am pretty confident that if we restricted both drivers to a narrow frequency band that both could cover, a volume level that kept both to lower distortion levels, the frequency band was restricted enough to keep dispersion similar, and you did not listen too near field, then you would not be able to identify a difference that was indicative of size of the driver.




- Unfortunately, the people who often make these claims often don't have a high correlation with reality attachment either, but let's not go too far down that path.

- Assuming tolerable subwoofer integration and the fact that multiple subs (and mains) usually results in a more even bass response with node elimination the net result is both a tendency to use higher listening levels and a lack of masking functions from excessive room modes. Both will result in improved mid and high perception and yes this does correlate with measurements. Flatter room responses which on average does correlate with subwoofer usage, and certainly your subs, will always result in improved perception of mids and highs.

- Supertweeter usage is somewhat rare and any consistency of implementation is totally out the window since people pretty much just slap them in without having any idea what they are doing. Real ultrasonic frequencies (which is unlikely what is coming out of most supertweeters) can create a masking function for audible higher frequencies, but more likely is that the supertweeter is evening out a poor high frequency room response which like room modes at low frequencies can eliminate a masking function hurting overall sound impression. Then again, in this case I am willing to consider placebo.



- I would counter that it is not generally apparent what horn coloration is or even that it is real and any different from any other coloration that could be assigned to any speaker w.r.t. resonances, etc.



- Nothing at all unusual about this. This is true for any human sensory function including light and touch. Duration always plays a part in perceived intensity, but then again, when have intensity and energy ever been the same?




- See my quote above w.r.t. using music as a test signal, and for imaging, that is really really a tough one as what contributes to imaging in real music is all over the map with a preponderance for level based cues, far less frequently timing cues, some artificial (usually) frequency response cues, etc. Spaciousness .... even more of a mixed bag.

- Are measurements useless? ... No, at least in room measurements of direct/reflected can provide information on what you may be able to achieve for both. Every room is different so looking at the speaker on its own is never going to be enough, and most people don't know how their speaker will behave, but they can make some guesses and they can adapt their room to the speaker.

That's "Harman", not "Harmon".

 

[aslan7]

I had one of those and, despite the necessary maintenance, I loved it. In fact, I still have a few parts left.

Yes it was good on the few occasions it worked. I was very glad to trade it back in for an SME. Of course I am a big digital fan but in keeping with my theories can’t bear to part with my old Thorens/SME combo.

 

[aslan7]

People have no idea how easy it is to self-hypnotize.

I am obviously pro science but when I look at the top Prima Luna amp and preamp can’t help thinking it’s a lot cooler looking than my VTV Eval 1. I wouldn’t mind having my original Ampzilla back either.

 

[audio2design]

That's "Harman", not "Harmon".

That's pedantic, not pedantic.

 

[Jim Matthews]

That's pedantic, not pedantic.

At least it was brief.

 

[Andrew s]

Measurements can't be the whole story. It is the psychoacoustic theory that they relate to that is "the story". While some areas are well supported by good data others are not as @audio2design pointed to above.

Even when well justified it can be misunderstood. Take the threshold of hearing is it drawn where 100% can't hear it 50/50 or something else?

Maybe rather than only measuring kit we should be strengthening our psychoacoustic theories?

Regards Andrew