Audio Science: Does it explain everything about how something sounds?

[Al M.]

Later tonight we are going to the music fest and hearing the little river band over their really big stereo system, I just hope the sound guys don't turn the ultra low bass up to shake our bellies, because their music never sounded like that to start with and it might run me off for the second night due to whacky sound that does not match my preferences, fingers crossed.

I love the Little River Band. Enjoy the concert!

 

[NorthStar]

It's great to see this definition of the term "MUSICAL" = PLEASING AND ACCURATE. Along with "natural", "musical" is the highest compliment to describe a system, IMO.

I'd love to hear again Kal Rubinson's take on the term "musical" to describe an audio electronic component, or an electric/mechanical loudspeaker.
Musical as reproducing musical instruments with their pure natural tones?

Accurate? ...That's definitely another term that demands refined interpretation in the scientific audio world. ...Accurate as neutral?

In professional audio reviews there are many terms that have various meanings for various people in describing material audio components, including the loudspeakers, the wires (cables) and even the magic voodoo tweaks.

* That is a great pic Peter...from the ocean's surface.

 

[NorthStar]

I don't have current information. As of a few years ago when I took the test, it was a Proceed power amp. I don't know what was used to drive it as they were not visible.

They weren't using some Bryston amps @ one point?

 

[Orb]

I don't have current information. As of a few years ago when I took the test, it was a Proceed power amp. I don't know what was used to drive it as they were not visible.

Thanks Amir, I guess the preamp would be one of their brands, surprised they did not consider the ML gear for both components.
Cheers
Orb

 

[AJ Soundfield]

thus, to argue that warm-up effects are not audible is to argue that we can't hear these distortions (and no I am not going to prove that it's easy to hear them, but for starters, think harsh sound from transistors)

So you don't have any evidence of audibility either. Just more anecdote. Ok thanks.

On a side note, audiophile pathological SS amp measures 0.1% THD at startup, 0.001% THD (100x less) 4 hours later. Audible?

 

[AJ Soundfield]

I don't have current information. As of a few years ago when I took the test, it was a Proceed power amp. I don't know what was used to drive it as they were not visible.

Maybe ML these days? Hows it going old man?

 

[Phelonious Ponk]

Since the post was mine, I’ll respond to you, Tim.

To the first point: What qualification do you suggest I hold in order to put forth some potentially valid criticisms of a publicly listed corporation involved in research for profit? Would I need to be a peer? Hold a degree in a similar field? Be elected to a position to defend consumer rights? If it’s none of those things, then calling the above an “unfettered… attack” reeks of reactionary emotionalism and defensiveness. No one is above critique, Tim. That’s the ‘Merican way. A constitutionally supported democracy should be involved in encouraging its citizens to freely question the establishment, because if it is not, then it’s not a democracy. Asking questions about how a for-profit company conducts its research is not an “attack”, its called exercising freedom of speech.

“Uniformed?” I read the paper the company’s VP wrote on how they conduct their research (downloaded from here: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=3&ved=0CCoQFjACahUKEwj5xaOOiZrHAhUEuRQKHTWEBuk&url=http%3A%2F%2Fwww.harman.com%2FEN-US%2FOurCompany%2FInnovation%2FDocuments%2FWhite%2520Papers%2FLoudspeakersandRoomsPt2.pdf&ei=cETGVfnNC4TyUrWImsgO&usg=AFQjCNFsdYMWG0hLEgaj-6qQ7-h2zsafDw&sig2=qX_TNtPNyRaZuVerPR6WFA&bvm=bv.99804247,d.d24) and couldn’t find any statistical certainty measurements apropos any of their findings. What’s more, I’ve sat through slides and “training” from Harman employees when I worked for four years part-time in a high-end retailer during my second university stint. I’ve sold their products. I’ve read the manuals. Anything else you’d like to me have done?

Of course, I don’t claim to have anywhere near the experience of writing technical papers for a corporation that pays me to do so, and allow that there are going to be gaps in my knowledge because of that. That’s why I had some questions, Tim. Can I ask, are you okay with the research methodology?

To the second point: An insult is intended to offend someone personally. An indictment implies accusation. Neither Toole nor Olive were personally accused of anything. I was questioning the robustness of their company’s research controls, having read through the papers and not found any.

To the last point: Like I said, I read the papers, found no information as to how they control for confounding variables or conflict of interest, and responded to a post from a person who we know has a vested interest in defending the research due to their company being in a position to gain financially from its dealer status with the corporation in question. Yes, Amir is one of the founders of the forum, and has declared his dealer status. That still doesn’t make him, Toole or Olive above critique.

And you’re right, I had many questions about how the research was designed and executed, and I still do. But since my line of questioning has only seemed to create defensiveness, I’ll abandon it because I just don’t think it’s worth entering into personally-charged slanging matches over first-world problems.

I will say this though. I believe in the scientific method in as much as conflicts of interest can be eliminated and statistically certainty can be substantiated. And I also believe in the heuristic approach to building a narrative of enquiry where that individual’s experience is owned as their own but shared within a sympathetic community. This hobby is filled with both approaches and I’m glad it is. But I’m glad scientists didn’t dismiss outright the highly unusual and scientifically unsupported anecdotal claims of a woman in 1989 whose experience of her dog taking an specific interest in a skin legion lead to the correct diagnosis of a malignant melanoma. Had they done so, we would not have uncovered a line of potentially life-saving medical research* in which dogs are now being trained to use their oncological senses in the detection of volatile organic compounds related to tumours. Given the standard scientific Prostrate-Specific Antigen test already has a high false positive rate (less than 1 in 4 men who test abnormally have prostrate cancer) leading to unnecessary invasive surgery, further study based on canine prostrate detection in urine seems worth pursuing, as crazy as it may seem to those who hold rigorously to the established norms of PSA.

*The initial study in 2004 had power in excess of 95% to show a statistically significant result (P < 0.05) for a mean success rate of at least 55%, irrespective of the method of analysis used. They assessed power by 1000 stochastic simulations of the experiment with each dog having an expected success rate of between 45% and 60% (mean 55%). The results were analysed by t test and bootstrap techniques, to ensure that the power was adequate under both forms of analysis.

P.S. For anyone interested, I just started a new thread in the Jazz Forum on favourite John Zorn releases.

You'll forgive me if I don't wade through all of that. I refer you to pg. 90, Post # 892. It doesn't appear that you have a lot of questions about how the research was designed and executed, it seems that you made a lot of assumptions about how the research was designed and executed, assumptions so dismissive that painted the research as so amateurish that yes, it was an insult to Toole and Olive.

Tim

 

[Al M.]

So you don't have any evidence of audibility either. Just more anecdote. Ok thanks.

On a side note, audiophile pathological SS amp measures 0.1% THD at startup, 0.001% THD (100x less) 4 hours later. Audible?

Yes, that should easily be audible.

From:
http://www.lindos.co.uk/test_and_measurement/SOURCE=Articles/SOURCE=Articles|VIEW=full|id=5

3 Crossover Distortion – how THD fails as a measure of audibility

In the early days of audio, one form of non-linearity dominated in all systems. Valves (tubes), tape recordings and transformers all tended to produce less output at high levels of input, in a symetrical manner (affecting positive and negative sides of the signal equally). This ‘soft limiting’ squashes the peaks of the sine-wave, producing a form of distortion known as ‘odd-order’ which contains only odd harmonics of the input (3rd, 5th etc). Processes that cause asymetric distortion generate only even harmonics (2nd, 4th etc) and are rarer. Odd order distortion products are sometimes considered more objectionable than even order, since they are not musically related in the way that even order products are (by octave intervals).

With the introduction of transistor amplifier though, a new form of distortion arose, known as ‘crossover distortion’ and caused by a kink in the transfer characteristic as the sine-wave crosses zero. This is a form of ‘high order’ distortion, and produces odd harmonics (3rd, 5th, 7th, 9th etc) which extend right up the frequency range, with little reduction in amplitude, and because the ear analyses sounds in terms of frequency components, and is most sensitive to frequencies in the 2kHz-8kHz region, it turns out that we are particularly sensitive to even small amounts of crossover distortion, when compared to the ‘low-order’ distortion of valves and tape, which generate mostly 3rd harmonic.

While early experiments had determined that 0.1% THD (-40dB) was the very minimum that could be heard, it soon became apparent that this was not the case for transistor power amplifiers, where 0.01% or less could be heard as harshness on the sound. Nevertheless, THD measurements continued to be quoted, and audio measurement itself got a bad reputation among the ‘hi-fi’ fraternity, who turned to listening tests as the only way to assess audio equipment.

***

About Lindos, from whose website the above is taken:

Lindos Electronics has made audio test sets for more than 30 years. A small British firm, we're passionate about high quality audio. We're trusted by big players like Sony and the BBC to test the performance of their audio equipment, and now we're passing on our audio expertise to the recording industry with the creation of the Minisonic Mic Kit.

Source:
http://www.lindos.co.uk/home

 

[PeterA]

Yes, that should easily be audible.

From:
http://www.lindos.co.uk/test_and_measurement/SOURCE=Articles/SOURCE=Articles|VIEW=full|id=5

3 Crossover Distortion – how THD fails as a measure of audibility

In the early days of audio, one form of non-linearity dominated in all systems. Valves (tubes), tape recordings and transformers all tended to produce less output at high levels of input, in a symetrical manner (affecting positive and negative sides of the signal equally). This ‘soft limiting’ squashes the peaks of the sine-wave, producing a form of distortion known as ‘odd-order’ which contains only odd harmonics of the input (3rd, 5th etc). Processes that cause asymetric distortion generate only even harmonics (2nd, 4th etc) and are rarer. Odd order distortion products are sometimes considered more objectionable than even order, since they are not musically related in the way that even order products are (by octave intervals).

With the introduction of transistor amplifier though, a new form of distortion arose, known as ‘crossover distortion’ and caused by a kink in the transfer characteristic as the sine-wave crosses zero. This is a form of ‘high order’ distortion, and produces odd harmonics (3rd, 5th, 7th, 9th etc) which extend right up the frequency range, with little reduction in amplitude, and because the ear analyses sounds in terms of frequency components, and is most sensitive to frequencies in the 2kHz-8kHz region, it turns out that we are particularly sensitive to even small amounts of crossover distortion, when compared to the ‘low-order’ distortion of valves and tape, which generate mostly 3rd harmonic.

While early experiments had determined that 0.1% THD (-40dB) was the very minimum that could be heard, it soon became apparent that this was not the case for transistor power amplifiers, where 0.01% or less could be heard as harshness on the sound. Nevertheless, THD measurements continued to be quoted, and audio measurement itself got a bad reputation among the ‘hi-fi’ fraternity, who turned to listening tests as the only way to assess audio equipment.

***

About Lindos, from whose website the above is taken:

Lindos Electronics has made audio test sets for more than 30 years. A small British firm, we're passionate about high quality audio. We're trusted by big players like Sony and the BBC to test the performance of their audio equipment, and now we're passing on our audio expertise to the recording industry with the creation of the Minisonic Mic Kit.

Source:
http://www.lindos.co.uk/home

First a great post from Ack about amp warm up, now this one from Al. This thread is getting back on topic. Thanks guys.

 

[NorthStar]

So we heard the little river band at the music festival in our town. I had low expectations and they were somewhat fulfilled. Live concerts are about being surrounded by people talking and folks working the sound board that are deaf. We had too much bass, then next in loudness was vocals, then an extreme third were instruments. So, a poorly balanced "stereo" system which led to a huge loss of musical information but we enjoyed the atmosphere at the show non the less, as we always do. Diana commented how muffled everything sounded, I went scouting around and found a spot where we got a little bit better balance, but as I say, low expectations through years of amplified events has pre-biased me to low expectations. The light show was good, the fog, etc. and the songs brought back some memories and some of the folks were interesting to look at as well I much prefer the orchestra and even local municipal bands, that are unamplified and glorious to listen to. The little River Band, if I want to hear their music, sounds much better on their recordings than at the live gigs. But you all know that all ready.

Cool of you for sharing...and you said it; it's the "live atmosphere" with some of the other folks 'interesting' to look @ as well. ...And the fond memories of our youth.

 

[amirm]

First a great post from Ack about amp warm up, now this one from Al. This thread is getting back on topic. Thanks guys.

In what way do you think Al's answer contributes to your cause?

 

[BlueFox]

For the last year or so I have been turning the amps off when I go to bed, and then on a few hours before music time. This thread caused me to revisit the result of leaving them on overnight. Yesterday, Friday, I turned them on about 2PM, and they have been on since then. While I have amped up my electric bill with them being on, idle, and the AC constantly running because of that, this is a rigorous, scientific experiment, so I will absorb the cost.

Anyway, at 11PM Saturday night, they sound awesome. Using my calibrated and tuned ears, the background noise is non-existent, or at least lower, the music is clearer and more defined, and appears a tad (technical term) louder.

Yes, warmup does exist.

 

[Billy Shears]

For the last year or so I have been turning the amps off when I go to bed, and then on a few hours before music time. This thread caused me to revisit the result of leaving them on overnight. Yesterday, Friday, I turned them on about 2PM, and they have been on since then. While I have amped up my electric bill with them being on, idle, and the AC constantly running because of that, this is a rigorous, scientific experiment, so I will absorb the cost.

Anyway, at 11PM Saturday night, they sound awesome. Using my calibrated and tuned ears, the background noise is non-existent, or at least lower, the music is clearer and more defined, and appears a tad (technical term) louder.

Yes, warmup does exist.

I Have the same with my Roksan Integrated. Hifi news Tested the Caspian M2 and it produced measuarbly less distortion after 30min Warm up.

 

[853guy]

That is not their "research." It is a Powerpoint presentation meant to educate people so it has the high level message. The research are the ASA and AES Journal papers.

Dealer training of that sort doesn't teach you anything. Now if you went to Harman recently and sat through their 2-day presentation, then you would have had an informed opinion because you would actually then sit through the blind listening tests. But even that would not get you educated enough to know this very complex field. It took me years to read hundreds of papers, spend days with the researchers to learn this topic. It is not remotely something you pick up as a part-time person in a Hifi shop.

All of this analysis has been done and extensively so in every *paper*. Here is one of countless examples where the statistical analysis goes on for pages: "Differences in Performance and Preference of Trained versus Untrained Listeners In Loudspeaker Tests: A Case Study"

"In both tests, there was a highly significant
difference in preference between the different
loudspeakers; F (3, 180) = 175.5, p < .0001 for the 4-
way test, and F (2, 224) = 91.441, p < .0001 for the
3-way test. A Scheffe post hoc test performed at a
significance level of .05 showed significant
difference in the means between all pairs of
loudspeakers in both tests.
Other main effects that were statistically
significant in both tests were listening group; F (10,
60) = 8.844, p < .0001 for the 4-way test and F
(13,112) = 5.459, p < .0001 in the 3-way test.
In the 3-way test, program was statistically
significant, F (3,336) = 6.302, p < .0004. Since the
effect size was very small it is not discussed any
further. The other main effects are discussed in more
detail in the following sections. "

Again the analysis goes on and on. Please, please assume some competence on the part of our industry research communities to demand such rigor and for presenters to include them in all of their work.

Thanks, Amir, yes my bad indeed. I read only what I could find, which was essentially similar to what I'd already heard from Harman reps and nothing I found contained any statistical analysis. Nothing on experiment design nor experimenter/subject COI either. Are these papers available to the public or for AES/ASA members only?

 

[853guy]

You'll forgive me if I don't wade through all of that. I refer you to pg. 90, Post # 892. It doesn't appear that you have a lot of questions about how the research was designed and executed, it seems that you made a lot of assumptions about how the research was designed and executed, assumptions so dismissive that painted the research as so amateurish that yes, it was an insult to Toole and Olive.

Tim

I can admit when I'm wrong, Tim, no problem. I did make an assumption, you're right. And what's more, it seems that there's a lot of statistical analysis accompanying Harman's research. As to Toole or Olive being insulted - impossible for me to know, Tim. That's your assumption.

 

[Orb]

I copied this exerpt, if it does not give pause to those who understand electronics, I don't know what will.

"If we now consider speech or music subject to crossover distortion in a power amplifier, another very important consideration arises. Typical speech and music waveforms do not cross zero very often. There may be violins or cymbals contributing significant high frequencies to the waveform, but most of the time these will be riding on top of bass notes such that zero-crossings are much less frequent than they would be for a pure tone. Where a relatively pure tone does arise from a violin note, it will not have many audible harmonics, whatever the distortion mechanism, because the 3rd harmonic of a 3kHz tone is at 9kHz, and the 5th is at 15kHz which will not be heard by most listeners. Most people are surprised to find that they can tell absolutely no difference between a 6kHz square wave and a 6kHz sine wave, but of course the 3rd harmonic is at 18kHz, and even listeners who can hear to 18kHz will usually have greatly reduced sensitivity at this frequency."

the first bold is obvious, the second bold also as we do not send square waves to speakers in music listening. A lot of this stuff is marketing stuff carefully crafted and can fly right over folks heads giving subjectivists the "science" they need which is just really confusing market speak.

Other stuff I have not read as carefully and just see it as salesmanship mostly. There is of course truth that a THD+noise measurement is not the same as a spectrum analysis measurement, which is the best way to see or understand what is happening, and to say the IMD is not so much as it is cracked up to be is being way to dismissive IMO.

Anyway, an interesting link to a marketing paper.

Ah man,
if only you managed to let THX and Benchmark know before the development of the Achromatic amplifier that was implemented in the Benchmark AHB2
You would had been their hero for stopping them wasting time and money on developing an efficient design totally removing crossover distortion
Ok it does more than this but one benefit of the feed-forward rather than negative feedback (found in nearly all other designs).
Funny enough the AHB2 is one of the best measuring products out there and is the best in some ways.
And before any dismissive comments, please note this was developed by THX; not what one would call your normal audio manufacturer.
Cheers
Orb

 

[AJ Soundfield]

Yes, that should easily be audible.

– how THD fails as a measure of audibility

I love this place.
Thanks Al.

cheers,

AJ

 

[AJ Soundfield]

First a great post from Ack about amp warm up, now this one from Al. This thread is getting back on topic. Thanks guys.

Peter, if you and friends have not a care in the world about science when subjectively enjoying music on your stereos, why the dire need for what you hear as you watch/know/read manual/etc, etc. your system warming up, to have a scientific objective explanation?
Clearly not a subjectivist, so what do you consider yourself to be?

..."how our ears tell us things that audio science can not yet explain."

Another paradox, given your acceptance of (incorrect) science explaining what you "hear" after hours/days/weeks.

cheers,

AJ

 

[AJ Soundfield]

I Have the same with my Roksan Integrated. Hifi news Tested the Caspian M2 and it produced measuarbly less distortion after 30min Warm up.

Did they test whether it was audible? Or is that irrelevant here to self identified subjectivists?
Keep in mind, I'm not saying an audiophile amp can't be so pathological as to meet this condition, it's just that I need a bit more evidence than "I said so". Thread title related you know.

cheers,

AJ

 

[amirm]

Yes, that should easily be audible.

From:
http://www.lindos.co.uk/test_and_measurement/SOURCE=Articles/SOURCE=Articles|VIEW=full|id=5

3 Crossover Distortion – how THD fails as a measure of audibility

Well, THD fails in almost every manner to predict audibility of distortions. And the way we know that is not by reading a claim of 0.1% audibility in that web site but listening tests. Here is one from the presentation/paper published in AES, MEASUREMENT OF HARMONIC DISTORTION AUDIBILITY USING A SIMPLIFIED PSYCHOACOUSTIC MODEL - UPDATED:

Focus on the graph on the left. The horizontal axis is amount of THD distortion. The Vertical is listener detection of the same. One means it was not detected. Five means it was obnoxious. the circles are results of listening tests at those THD distortion levels.

If THD predicted the audibility of distortion, then all the circles would all line up on the blue line. As is clearly seen, it does not do that. Studies show correlation of about 30%. Not good. The reason it is not good is because THD sums all the distortion products and gives you one number. But audibility of those distortion products is not the same due a psychoacoustics principal called "masking." Low order distortions are less audible than high order. In that regard, I call THD psychoacoustically blind measurement. Even the Wiki gets that right: https://en.wikipedia.org/wiki/Total_harmonic_distortion

"For many purposes different types of harmonics are not equivalent. For instance, crossover distortion at a given THD is much more audible than clipping distortion at the same THD, since the harmonics produced are at higher frequencies, which are not as easily masked by the fundamental.[23] A single THD number is inadequate to specify audibility, and must be interpreted with care. "

So the question of whether 0.1% THD is audible is wrong as is the answer. Except in gross amounts where we know it is audible, it is not a metric to be used.

The graph on the right is the proposal from the authors to use psychoacoustics to add weighting to the distortion products according to how we hear. When we do that, the correlation improves drastically. The listening test result circles now track the line much more closely than THD although still not perfectly. There have been a number of such proposals but unfortunately the audio industry continues to use the stale THD because it is easy to measure and it is what people know.

Hopefully members see how there is proper audio science that has evaluated such things (see more here: http://www.gedlee.com/distortion_perception.htm). And that the knowledge there, is the proper one, based on listening tests and how we hear. In that regard, we better embrace it if what matters is what we hear .