Master Thread: Are measurements Everything or Nothing?

[Holmz]

….
I propose a mechanism why that may sharpen the leading wavefronts and produce music with a little more punch as a result. Add this to the rhythm chords, and you may find the music has more drive. (Or not - we could test!)
…

I doubt it… More harmonics make it sound louder at the same SPL.

And people usually like loud better than quiet.
I believe that that is the only “punch”, the rest is Kool-Aid.

 

[Mr. Widget]

Dave's output is three times higher than what your phone outputs. You must match those levels. Do the test blind. And repeat at least 10 times.

Is Dave's output higher than other high end DACs? I ask, because I have had friends swear that Chord's DACs sounded better to them... obviously one reason could be marketing, but why would Chord's marketing resonate more than DCS, Mytek, or whoever? If their standard gain structure was a bit hotter than the competition it could be an answer.

 

[Thorsten Loesch]

hmmmm... I find comparing A to B and or X already exhausting. I could not throw more into the mix unless it is about switching between a bunch of speakers and quickly assessing which are worth listening and not. And even then I prefer to switch between a known reference and C, D etc.

Not arguing that. That's why I choose to gameify this and instead make it "fun equipment tasting" in a way Audiophiles and professional musicians relate well to.

Instead of "OMG again this shell game, which bloody button" it becomes "Hmm, I wonder what this like?".

We are changing the relation of object and subject, WITHOUT violating the "blind" part.

Everyone designs tests for a specific purpose. ABX was designed for a specific purpose (ABX actually to bilk money from gullible pee-pull, just like the shell game, I'll write a bit more on that affair and why ABX [as in Audio ABX Test by ABX Company of in Troy, Michigan and David Clarke, as opposed to the more general ABX principle of forced choice testing] is gold certified cargo cult science).

I try to design my tests to give me results useful in my own endeavour and answer questions I have.

As such for example introducing ABX would run counter to my purpose, as I want amplify differences, not attenuate them into the noise floor.

It can be done easily fulfilling all the precepts of modern empirical science, to make sure we don't fool ourselves.

These days I just post my opinion with some kind of explanation which I hope isn't too technical and think I am talking in 'layman language' but always still seems to technical.

We seem mostly on the same page here.

BTW, something I wanted to mention. You may get a little kick out of this.

You may be aware of the iFi Zen Can and the various "Headphone specific" versions. The inspiration for my design was actually your Kamaeleon.

Now my way of doing it is different, I use a non-inverting Op-Amp followed by the Volume Control.

That NI Op-Amp mainly serves as input buffer (gain 1.22) and performs boost EQ active in the feedback loop.

Any cut EQ is implemented passive using an 0.82 Gain resistive divider formed with the Volume Control (2.2k & 10k Volume control).

The result is preservation of polarity and lower noise (and perhaps one stage less overall).

The actual EQ options implemented match Kamaeleon, as I felt you did an excellent job of making a simple, effective and flexible EQ system, for compensating headphone response flaws and I liked it enough to steal it. Using a combination of Boost and Cut shelf EQ's to make very wideband low Q corrections is inspired. It is not easy to do with either graphic or fully parametric EQ which needs a lot more components.

You have a great website there, excellent resource.

When one would be looking at multitones of both modes it would probably more clear as to the differences in performance between the modes where tube+ approaches potentially audible levels (music dependent).

I don't know, with complex music I find that at 85dB average playback 3% of H2 and 1% of H3 are generally inaudible and 10% of H2 or 3% of H3 while audible are not objectionable. The last serious listening based investigation into Distortion audibility was the previously mentioned one by Geddes-Lee, if you are aware of anything superseding this I'd be keen to update my understanding.

Fun exchange which is refreshing...

Indeed, we need to do this more often. If you ever come to Thailand, let me know. First shout is mine.

Thor

 

[birdog1960]

Not trying to be contrarian but it will seem that way. I think many of the initial premises are straw men. I suppose someone said each of the things initially presented (and a few, many more - "my ears"). "you all never listen" and " cult" do come to mind tho. I don't do measurements. Don't have the equipment and no plan to try. I go with my ears and have sent more than a few unsatisfying components back. That doesn't mean I don't appreciate being aware of measurements. Terrible measurements steer me away. Great ones ( especially at a reasonable cost) evoke interest. It's not a dualist proposition. just my 2 cents. And an idea for value conscious members: how bout a thread on best used equipment (with great measurements) available at various price points?

 

[krabapple]

Totally agree. But today's objective results don't necessarily capture everything relevant.

Your circular argument defines 'relevant' as ' what I like'.

Yet 'what I like' is influenced by things that aren't audible.

So it's a silly argument when it comes to audio quality.

The rest of your silly argument is the usual unqualified 'we aren't measuring everything important'. A 'god in the gaps' argument.

(Did some other forum send its troll infantry to ASR this weekend?)

 

[krabapple]

Right - I'm a recording engineer. We do matched levels all the time (with SPL meters, btw.

Not good enough, btw.

It appears you don't even know what you don't know. But you're sure you're right.

 

[Thorsten Loesch]

Is not noise and distortion cumulative?

Not quite. Noise and harmonics add up according "SquareRoot of SumSquare".

So say having 2uV noise and 2uV noise in series will not produce 4uV noise, but only 2.828uV.

And if we have (say) 90dB SINAD from a DAC/Amp and 60dB SINAD from a transducer the system SINAD will be effectively that of the Transducer (60dB) and contribution of the DAC/Amp will be in the rounding error territory.

I've heard this description many times. Is "dull" the same for everyone? I doubt it. Is "boring" the same for everyone? I doubt it. On the flip side, is "exciting" the same for everyone? I doubt it.

It was for the individuals involved in my case..

In my small circle of friends, there is often sharp disagreement about these qualities. Is it not safe to assume that there is even more disagreement about these qualities among the general listening population? For example, might not I (for one) disagree with your opinion on this matter?

It is safe enough to assume that two people in a room will be unable to agree about anything.

That is why it is best to perform tests in such a way that all these psychological factors are removed from the equation.

I don't listen to that.

But many people do.

Are you insinuating that large numbers of people listening to garbage is a reason to deny me what I prefer?

No, I am suggesting that we need to put our debates into a context. For the music many people listen to, anything we are now debating over in terms of quality is probably better than required.

It is us few who actually need better than that and it's a shame that instead of working together toward better understanding, we draw up frontlines and snipe at each or dropping artillery on each other in a way that is reminiscent of what is going down in Bakhmut right now.

I apologize, but I don't quite understand your point here.
Let's say, for the sake of argument, that a live performance is the absolute ideal. We will ignore highly processed studio recordings for now.
Do you agree that there's nothing that can equal being right there in front of the performers in their performance venue?

I agree. A GOOD recording engineer can however make a recording that, replayed on an adequate system communicates much of the experience to the listener, if recording and system are good enough even to "suspension of disbelief".

In other words, "high quality audio" (including the recording part) should get close to being "equal being right there in front of the performers in their performance venue?"

But which is the better or more perfect recording of that event? Is it one that has less distortion or one that has more distortion?

Ok, we now need to define "distortion".

Just using microphones to convert sound to electrical signals is not a distortionless process. Microphones distort the 3D perspective of acoustic even in stereo recordings, microphones add noise, harmonic distortion and whatever else they add, even if it is not currently a quality we quantify using objective measurements.

Among Recording Engineers Microphones a very hotly debated subject.

If we characterize the original event as being the baseline, then which condition more closely approaches it?

I would suggest that it is that condition which allows the listener to suspect her/his/it's disbelief and connect emotionally to the music as if they were in the performance venue. And that is not guaranteed by minimising noise and distortion.

Listen for example to Miles Davis "Kind of blue" or John Coltrane "Blue Train". There is noise, KOB has severe mike or tape overload on Mile's horn in some sections.

Nevertheless the performances are incredibly able to influence people emotionally, which is why we still listen to 'trane and Miles in 2022.

Having a more accurate recording of an event

What is "more accurate"? How do define accuracy?

I for example propose that recording using a modified Decca Tree with an ORTF Array in the front position and two large format omni's in the rear positions with time-alignment in a DAW and then cloning the Omi Track and using a mix of frequency response shaping to shift from intensity stereo at high frequencies (that's what the ORTF array does) to time delay stereo at low frequencies (that's the job of the Omni's) gives a more accurate representation of the actual ensemble, as well as given a more accurate tonality of the ensemble than alternative minimalist stereo recording techniques.

I am even going as far as using cancellation of the enseble in the tracks (imperfect but surprisingly good) to create a "ambience tracks" where the ensemble is significantly reduced and by combining these ambiance tracks with response shaping and delay (not reverb) with the ensemble can create a better representation of the acoustic space of the venue in the recording than alternative minimalist stereo recording techniques.

I actually got my "Tonmeister" in 1988 with an early, all analogue version of this microphone setup, in what was then east berlin.

And I am very sorry for talking about microphones, positioning etc and not all noise or distortion. These are generally for professional recording gear in the 1980's assumed to be below audibility, if the equipment is used in a competent way. We (Tonmeister) rarely debate these, microphones and speakers are the focus.

means that there is the possibility of modifying it by adding distortion, IF the listener wishes. The same is true of the whole reproduction chain; a more accurate (or "cleaner") reproduction chain can always be modified by the addition of various distortions so as to please various people.

I think making pre-processed but not finally mixed mastered source tracks available with a simple to control interface to shift from the final mix more towards the source, by instrument where this applies would be more useful. Adding or subtracting HD is not really useful, unless the level is extremely high (> 3% H2 & > 1% H3 @ 85dB average SPL).

I think it's incorrect to state that large segments of the population "like" distortion.

I agree. I would go even further in saying that it's disingenuous, cynical and self serving.

A recording that is of comparatively poorer quality and a reproduction chain that is of comparatively poorer quality cannot be modified to be more accurate (or "cleaner"). That's been lost. For the recording, it's been lost in production, and for the playback chain in your home, it's been lost in reproduction.

Completely agreed.

Let's not take pure water and add a smidgen of dirt to it on the off chance that someone might not notice.

No, that's not quite it. Let's not over purify water beyond the point of necessity and then pretend this water is better for the person drinking than that which only been purified to the point of necessity.

Thor

 

[birdog1960]

Your circular argument defines 'relevant' as ' what I like'.

Yet 'what I like' is influenced by things that aren't audible.

So it's a silly argument when it comes to audio quality.

The rest of your silly argument is the usual unqualified 'we aren't measuring everything important'. A 'god in the gaps' argument.

(Did some other forum send its troll infantry to ASR this weekend?)

What I like matters to me and I think many others. Sure there's other variables, the biggest for me being value. I like my subaru wagon. But a BMW x5 or a volvo has better specs (and nicer interiors). I'm still interested in comparing the measurements for what's important to me. imo, can't beat a WRX for value but that's me.

 

[Thorsten Loesch]

About 20 years ago, I was asking myself similar questions you now have with Thorsten Loesch aka Kuei Yang Wang (none of these are his real names). I do not think there are simple answers and as opposite of TL, I have stopped discussions about such things in a public forum. It goes to nowhere. Yes, I have AB test records, with individuals or small groups up to three people. And I will not go into any deeper discussion anymore. It is just wasting time. Just to mention, living under conditions of a DDR/CZ engineers gave us something, even if we are happy that those times have ended.

Pawel (I think in Czechia it would use a w not a v?) , yes, I have at times used nome du plume/guerre instead of my real name, but Thorsten Loesch is real. Check the videos linked in my profile.

Yes, there are no simple answers. There are only simple comforting lies and complex and inconvenient truth's.

I think too few people these question their beliefs and what they are told, something that (should) come as absolutely essential and naturally to anyone growing up and living under "Real Existierendem Sozialismus" (der -ismuss wo jeder mitmuss or die Stasi hilft Dir bei mitmuessen).

[real existing socialism, a slogan in the 80's to disguise the systemic failings of the eastern block system, the rest is a funny ditty in german, it means that socialism is an ism where everyone must join in, or the secret police will help you develop the correct attitude]

Thor

 

[Thorsten Loesch]

Whoever he is, I don't find his diatribes against ABX (which he later modifies to 'ABX as usually practiced' ) novel or at all convincing, nor his anecdotes from 'tests' conducted way back when.

But it's Christmas and life is short.

Yeah, I suggest you look into the history of ABX in Audio and specifically of the ABX Company of Troy Michigan and it's principals, who defined the Audio ABX test as in current practice, the various criticisms levelled against the method, the various experiments (including Oohashi et al Hypersonic Effect) that illustrated the inability of the ABX test to reveal audible differences, the audibility of which was confirmed by EEG.

Then, combined with the Richard Feynman piece on "Cargo Cult Science" you may understand why I call ABX "Cargo Cult Science", "Confidence trick" and "shell game".

Thor

 

[krabapple]

Everyone designs tests for a specific purpose. ABX was designed for a specific purpose (ABX actually to bilk money from gullible pee-pull, just like the shell game, I'll write a bit more on that affair and why ABX [as in Audio ABX Test by ABX Company of in Troy, Michigan and David Clarke, as opposed to the more general ABX principle of forced choice testing] is gold certified cargo cult science).

I know who David Clark (no e) is and there was never anything secret about him designing and selling the ABX Comparator device; the name and location of his company is right there in the affiliation info in this JAES article he published about it in 1982. And I have, for decades now, followed the critiques of double blind methods used in audio comparison, including ABX in particular.

You seem like a bad actor. And citing Oohashi, who was an outright crank, makes me chuckle.

 

[Thorsten Loesch]

Musicians have rhythm and pace, Equipment does not.

A certain british turntable (actually several) has, by design. It's motor is so weak, the increased stylus drag will slow down the record and speed back up when modulation is less.

A test LP like the HiFi News Test LP that has 315 Hz tracks at multiple levels combined with a frequency counter can be used to to evaluate this.

Past that, adjusting the pitch of music changes the listeners reaction. Pitching flat (down) from a=440 Hz creates a sense of relaxation while pitching sharp from a=440 Hz creates more excitement. Many DJ's in the vinyl age (me included) deliberately used running the pitch slightly sharp even if not needed for beat mixing and later in the evening starting to pitch a little flat to soothe the drunken animals wanting to make a ruckus.

Thor

 

[krabapple]

What I like matters to me and I think many others. Sure there's other variables, the biggest for me being value. I like my subaru wagon. But a BMW x5 or a volvo has better specs (and nicer interiors). I'm still interested in comparing the measurements for what's important to me. imo, can't beat a WRX for value but that's me.

'Likes' and 'dislikes' that are easily swayed by non audio confounders are quite bad bases for comparing objective performance.

This is why audio gear is measured, and listening comparisons are done blind, level matched, etc.... if you want reliable data on performance.

As for car analogies to audio gear : those have always been just plain dumb.

 

[Thorsten Loesch]

The premise "anything that matters can be measured" is myopic.

Correct. And the corollary "because it is measured it matters" is equally myopic.

It is the sign of cargo cult science.

Real science is self-correcting and this process of self-correction forms part of the continous advancement and progression.

Cargo Cult science resists correction and is static. It does not advance, it does not progress and as Feynman put's it "The planes don't land!".

Thor

 

[Thorsten Loesch]

I doubt it… More harmonics make it sound louder at the same SPL.

And people usually like loud better than quiet.
I believe that that is the only “punch”, the rest is Kool-Aid.

This is correct, it used for example in the Aphex Exciter (something I add into any Club sound system I ever set up). But the distortion levels for this are quite high, several percent of H2 or over a percent of H3. Outside transducers it is extremely rare to encounter such levels in equipment that functions correctly and is operated within it's capabilities.

The difference between (in my case) 0.001% THD&N and 0.015% THD&N at what would 120dB SPL (with the headphone Amir used) can be safely considered inaudible, based on every thing I know.

I stand to be corrected of course, if there is new research that requires to correct my position and update what I know.

Thor

 

[Thorsten Loesch]

No measurements of any subject in any field is complete. Yet we are able to examine what is wrong with you when you get sick and fly to Mars. For a DAC we measure with ease what impacts audibility to levels way better than we can hear. No one has demonstrated they can hear what we can't measure.

Of course if you don't know electronics or audio science, you can be sold any story that favors a manufacturer. We don't have that problem here.

Amir, let me challenge you here. You say:

"For a DAC we measure with ease what impacts audibility to levels way better than we can hear."

Please provide reliable scientific evidence for:

1) "what impacts audibility" in terms of measurements
2) "than we can hear" that is what levels of what impacts audibility according to 1 are the audibility limits.

Then we can put your extensive measurements into context to gain useful information.

There is that old saying that Information is Data placed in a relevant context.

Your measurements provide a huge amount of Data I peruse frequently, placed in my context.

But to the general public there is no context provided and thus they may mistake data for information.

I know you make no such claims, but I think without a huge disclaimer to this effect you permit the average Joe Bloggs to make such mistakes and even encourage them.

And IF, as your statement implies, you have all the evidence and backing to provide the context to your data, why not do so?

Thor

 

[birdog1960]

'Likes' and 'dislikes' that are easily swayed by non audio confounders are quite bad bases for comparing objective performance.

This is why audio gear is measured, and listening comparisons are done blind, level matched, etc.... if you want reliable data on performance.

As for car analogies to audio gear : those have always been just plain dumb.

yes, because cars are way more complicated. But we compare car audio systems. narrow horizons are dumb. I'm looking at it as a consumer (and I'll bet many manufacturers are too). You seem to be looking at it as a pure science, which most would agree, it isn't, in terms of listening enjoyment..If it was, there would be one speaker design, one amp design, one digital or analog source (even with cost as no object) etc that was considered superior in a given room.

 

[Thorsten Loesch]

No one has shown what is audible that we don't measure. Give an example and proof and it would be trivial to create a measurement for it.

Amir, let me pose a counter challenge.

No one has shown, that except at extremely gross levels anything you measure is audible and an indicator of sound quality. Give an example of what is actually audible.

Thor

 

[Thorsten Loesch]

Fletcher-Munson not only could measure frequency response in 1933, but measured ear's frequency response including threshold of hearing: https://en.wikipedia.org/wiki/Equal-loudness_contour

"The Fletcher–Munson curves are one of many sets of equal-loudness contours for the human ear, determined experimentally by Harvey Fletcher and Wilden A. Munson, and reported in a 1933 paper entitled "Loudness, its definition, measurement and calculation" in the Journal of the Acoustical Society of America.[2] "

You see the danger in talking about specific science but not having done the research to know it? In this forum, you are talking to people who have studied the science and engineering. In contrast, you are repeating marketing lines from manufactures that benefits them but not consumers. You need to get past that to know that we are more than qualified to measure what makes a high performance DAC, and what doesn't.

And Fletcher-Munsons measurements were pretty inaccurate too (Blue vs Red). Just saying. But good to see that SOME things in Audio work according to scientific principles and are self correcting.

Thor

 

[DonR]

A certain british turntable (actually several) has, by design. It's motor is so weak, the increased stylus drag will slow down the record and speed back up when modulation is less.

A test LP like the HiFi News Test LP that has 315 Hz tracks at multiple levels combined with a frequency counter can be used to to evaluate this.

Past that, adjusting the pitch of music changes the listeners reaction. Pitching flat (down) from a=440 Hz creates a sense of relaxation while pitching sharp from a=440 Hz creates more excitement. Many DJ's in the vinyl age (me included) deliberately used running the pitch slightly sharp even if not needed for beat mixing and later in the evening starting to pitch a little flat to soothe the drunken animals wanting to make a ruckus.

Thor

Mechanical systems (outside of speakers) are interesting only in a historical sense to me. DJing is content creation, not hi-fi reproduction.