Do we crave distortion?

[pseudoid]

I say why choose when you can have both. Transparent and Flavoured options via DSP, speakers and/or optional valves. Part of the nerdy appeal of hifi, no? Needn't be binary. Would hate to choose between the R1, BMW or Triumph motorcycle wise, so why so with hifi. Dayum.

Be careful talking about DSP around @Sal1950.
I think he means it about MQA.

 

[Travis]

On a related note, is there a way to change the harmonics of a musical instrument acoustically / mechanically in a controlled manner. Say, if I turn this lever this way and this much, the guitar produces 10% more second harmonics or such? I was thinking whether it would be possible to turn this into a controlled experiment but there are so many variables in recorded music I thought might be easier to test for it without any recording / reproduction system, not even sure if it makes any sense.

The different harmonics of instruments is timbre. It is what allows us to tell what type of instrument is producing the sound. An oboe from a clarinet.

A study on this is Luce, David A. (1963). "Physical Correlates of Nonpercussive Musical Instrument Tones", Ph.D. Cambridge: Massachusetts Institute of Technology.

Some instruments play only the fundamental note, with little or no harmonics, the flute is an example. Other instruments have many harmonics, and varying levels, like the tuba. Those harmonics are inherent in the design/function of the instrument and there is typically no way to add, or reduce, harmonics to a fundamental sound/note by pulling a lever, or twisting a dial. The one exception to this may be the type of strings used on those instruments, a piano, or guitar. But it isn’t much, certainly not enough to fool you into thinking a viola is a violin, cello, etc.

As mentioned previously, non-controlled harmonics can be added by a performer, depending on the instrument. The most obvious example of this is Billy Gibbons, ZZ Top, and his style of solo picking (pinched harmonic) where he uses him pick to stike the string in such a way that it produces high harmonics to the point that it creates a sound different than the fundamental, perceived as highly distorted guitar sound. Solo on Tush is full of it.

 

[audiofooled]

It makes me wonder about distortion and perceived loudness. Could it be that the effect can be quite benign and flatter some of the recordings when they are reproduced on a transparent, also full range and low distortion system?

Or, to put it differently, if such tools are valid for music production to add some of the desirable effects to otherwise fully transparent production gear, wouldn't it make sense that reproduction gear doesn't need to alter it in any audible manner? Otherwise, a favorable combination then would be a kind of hit and miss, I would say.

To me, low distortion, be it in the recording or the playback system never seems too loud, so that's one point. Even good recordings of instruments with rich harmonic content.

Other than perceived loudness, if certain amount of distortion is already in the recording, I'm not sure about the required level of the added distortion profile to become just audible to me. When I tried Klippel test, it soon became quite obvious that it's not that audible at all if I want to just enjoy the music, or even in critical listening:

 

[pseudoid]

The different harmonics of instruments is timbre...

Thank you, that is exactly what I meant.
It sounds like you were adding indscriminate 'coloration' with no regard to individual instruments' timbre.
Isn't that the point where the brain starts getting overclocked to make sense of it and gets very tired very quickly.
Unpleasant?

 

[audiofooled]

Some instruments play only the fundamental note, with little or no harmonics, the flute is an example. Other instruments have many harmonics, and varying levels, like the tuba. Those harmonics are inherent in the design/function of the instrument and there is typically no way to add, or reduce, harmonics to a fundamental sound/note by pulling a lever, or twisting a dial. The one exception to this may be the type of strings used on those instruments, a piano, or guitar. But it isn’t much, certainly not enough to fool you into thinking a viola is a violin, cello, etc.

Here's a couple of interesting short videos on harmonic series and frequency spectrum of some instruments, this being one example and there are more on the channel:

And one when the instrument is purposely "broken" and sounds like a completely different thing:

 

[Curvature]

Some sources:

• Earl L. Geddes & Lidia W. Lee, “The perception of distortion” (2003), http://www.gedlee.com/Papers/Distortion_AES_I.pdf
• Earl L. Geddes & Lidia W. Lee, “The perception of distortion” (2006), http://www.gedlee.com/Papers/AES06Gedlee_ll.pdf
• Alan Lofft, Ian Colquhoun and Tom Cumberland, “How Much Distortion Can We Hear With Music?” (2014), Colquhoun Audio Laboratories Limited LC, http://www.axiomaudio.com/distortion

Finally, conclusion is also quite underwhelming I have to admit.

"The results of this study indicated that within the constraints of this simple study that the perception of linear distortion is dependent on the level and the delay time of the linear distortion and the playback level."

This is important because of a later statement:

As has been discussed here.

Later, Geddes wrote a book called Premium Home Theatre. He summarizes his thoughts on the topic by saying:

However, this is a stub of a claim and it is not discussed anywhere else. Geddes wrote anecdotally in his posts on DIY Audio and has repeated many times in presentations that he finds nonlinear distortion to be basically an uninteresting topic because its effect is so minimal compared to linear distortions, something mentioned in the papers and his book.

It's unfortunate because he's viewed as an authority and, despite that, allows loose, hard-to-interpret comments like the above.

The different harmonics of instruments is timbre.

Again, timbre perception is a mix of the spectral content, whether harmonic or enharmonic, the internal phase relationships, and envelope. It is not just harmonics.

Whenever we say "timbre", the reference is correctly tied to the human sensation, not something inherent to the instrument. Originally of course it was used to discuss instrument sounds, but the ambiguous philosophical discussions that followed about the difference between what's making the sound and how we perceive the sound have since been disambiguated, or at least they have been in hard psychoacoustics... Just like pitch is subjective while frequency is not, and the two do not have clean correlations, the same thing applies to physical location vs. human ability to localize; they are related, but fundamentally different. It's why you have "soundstage" instead of sounds anchored to two speakers. Stereo hijacks a part of our auditory processing and pushes a localized sound, which we are so used to accurately assigning to some object, to a phantom location. I think something similar is can be said about instruments in general, that they work because they hijack certain evolutionary features: they are more similar than they seem, despite the variety in appearance. I think that comment can be extended to speakers and their particulars as well. There's less difference at work than it seems.

 

[12Many]

I find this thread fascinating.
One question I had is: What is distortion? Is it harmonics, or noise? What about a note that varies form the original? A C becomes a C minor. Broadly it would be a change from the original, a distorted version. What are the types of distortion?

 

[Curvature]

I find this thread fascinating.
One question I had is: What is distortion? Is it harmonics, or noise? What about a note that varies form the original? A C becomes a C minor. Broadly it would be a change from the original, a distorted version. What are the types of distortion?

• Linear distortion: Changes in frequency response (where magnitude is not the same per frequency) and group delay (where the delay is not the same per frequency). Linear distortions are reversible through filters. Very audible.
• Nonlinear distortion: Added spectral contents to the signal (extras tones). Nonlinear distortions are hard to predict and cancel. Not very audible.

Edit: All sorts of modulation distortion like IMD or Doppler are nonlinear, before someone else brings them up as "other types".

 

[pseudoid]

One question I had is: What is distortion?

Simplistically speaking, it is the signal content on the input in comparison to that same signal (and its content) on the output.
Viewing @amirm measurements, they are graphically shown and these Input-to-Output differences are clearly described.
In both real-time and in the frequency-domain (and sometimed, when necessary, in phase-relation/difference) graphs/plots.

 

[Curvature]

What about a note that varies form the original? A C becomes a C minor. Broadly it would be a change from the original, a distorted version.

It's an interesting question.

If you play any instruments you'll probably be familiar with instances where poor techniques, like pressing too softly or inaccurately on a stringed instrument, causes the intended pitch to sound, for lack of a better term, distorted. That's probably the best example of a distortion. Moving from note to note causes big changes in, well, everything, every physical quality, so as long as the sound is within some intended operating range, it's fine. Take pitch: instruments drift while you play them, and some people find this very objectionable, or at least they have a lower threshold for detecting shifts. But even the notion of a scale and different tunings have shifted over time, and between cultures notes that tend to fall in between pitches established by convention are considered "wrong" ("distorted"), even though acoustically there's no reason why a microtonal change could not be considered just another note.

You can also use very interesting techniques that cause unusual (for the instrument) timbre, e.g., over- and underblowing for wind instruments.

Edit: Typoes.

 

[pseudoid]

One question I had is: What is distortion? Is it harmonics, or noise?

Both and more.
Would your question become self explanatory if you think of the situation that 'we' are NOT trying to create anything more than the original version of the media source?
Think of these measurements as 'reproduction' centric. it may help.

 

[fatoldgit]

This has always been my theory with regard to long time vinyl users (of which I was one but not now) preferring vinyl playback to digital. Because they have become accustomed to the inherent distortion in the vinyl playback system so a clean digital sound grates with them.

I remember reading a couple of properly conducted tests in the early 2000's where some University Music professors used their students as test subjects.

The students exclusively had listened to MP3's during their formative years and when asked to rate which sounded better, MP3 or CD versions of the same tracks, played back on a transparent audio system, they all preferred the MP3 tracks. Cause thats what they had been conditioned to (MP3 distortion).

I cant stand autotuned vocals but kids today dont mind cause thats what they have become accustomed to.

So yes, we can like distortion.

Question to the OP: what are the ages of the 10 test subjects?

Peter

 

[Curvature]

Both and more.
Would your question become self explanatory if you think of the situation that 'we' are NOT trying to create anything more than the original version of the media source?
Think of these measurements as 'reproduction' centric. it may help.

Good way to put it.

And then noise is noise. In electronics, a semi-random distribution of energy. Nothing like noise in speakers, unless it includes severe mechanical issues called "rub & buzz". If noise is high enough, it'll come through the speaker. Like hiss in actives. If amplifying a soundsystem to max without any signal playing, you'll hear the noise rising.

 

[Curvature]

This has always been my theory with regard to long time vinyl users (of which I was one but not now) preferring vinyl playback to digital. Because they have become accustomed to the inherent distortion in the vinyl playback system so a clean digital sound grates with them.

I remember reading a couple of properly conducted tests in the early 2000's where some University Music professors used their students as test subjects.

The students exclusively had listened to MP3's during their formative years and when asked to rate which sounded better, MP3 or CD versions of the same tracks, playback back on a transparent playback system, they all preferred the MP3. Cause thats what they had been conditioned to.

I cant stand autotuned vocals but kids today dont mind cause thats what they have become accustomed to.

So yes, we can like distortion.

Peter

Whenever comments like this come up they make me want to dig in my pile of papers for evidence that says: No!

Vinyl versions of songs have completely different mastering stages. Vinyl in general has FR limitations compared to digital and that comes through in comparisons, as well as mono'ed bass and some level compression. If you like vinyl, which has very, very high distortion from the playing head on (like, -25dB or >5%!), the distortion is still a small factor next to frequency response. To be clear, the distortion is audible, but does not have a high weight subjectively.

MP3 distortions, where they exist, are gross. It is highly unlikely that they are preferred, since that contradicts what we know from controlled studies done elsewhere.

 

[pseudoid]

A C becomes a C minor. Broadly it would be a change from the original, a distorted version. What are the types of distortion?

I was actually attempting to deflect because I would not know how to (begin to) answer this part of the question, if it is viable.

 

[Travis]

I say why choose when you can have both. Transparent and "flavoured" options via DSP, speakers and/or optional valves. Part of the nerdy appeal of hifi, surely? Needn't be binary. Would hate to choose between the R1, BMW or Triumph motorcycle wise, so why so with hifi. Got old school Tannoys sitting on top of two BK XXLS400FFS subs, with Neumann KH310s on top. Can easily switch between speakers and DSP presets for each via MathAudio.

Well there’s the rub, and it’s a big one isn’t it. The effects of this rub are on here on a daily basis. Just check out the “Tube Sound” thread.

Just limit it to amps, forget Preamps, Receivers, DACs, turntables, cartridges, speaker colorations. Here, in general, amps should measure with the least amount of distortion/SINAD as possible. Because? The science says that’s “true high-fidelity.” (I’m of course generalizing to keep this short(er). People who prefer tube sound or it’s sub parts SET, P/P, A, AB, tape sound, analogue sound, D amp sound (or lack thereof) vinyl sound, and a host of other things are continually reminded that, (pick a device), is more outdated than a Triumph, and ultimately inferior resulting in something less than high-fidelity (ASSUMING its even audible and not in your head). It’s typically inferior, and outmoded technology because of either distortion (tube amps), noise/noise floor, lower dynamic range, etc. Then . . .

Two camps (if they agree it’s audible)

1. Yes there are people prefer tube sound, for example, because they like distortion, this rolled off, or what the measurements can explain is the difference; or

2. Your preference is misplaced, you are listening to low-fi, and it’s probably because you were raised on ear buds/poor MP3 recordings, assuming you can even hear what you perceive to be a preference.

The reviews that Amir does on amps, with ultra expensive and hyper accurate instruments, and engineering science, have established that the key parameters are THD, SINAD, as well as others, but the lower SINAD the further to the left they go on that scale.

People preferring “Distortion” (assuming it is the same or similar as added in by these plugins) is fundamentally contrary to what is considered to be the “best” in the reviews isn’t it? (Refer to Measurements, Everything or Nothing).

If the preference of distortion by the 10 people that had the blind test by the OP is fundamentally contrary to what hi-fi should be, then you will the test questioned, it wasn’t double blind, they picked up on something, the distortion is completely different, or something else.

If the preference of those 10 isn’t contrary to what the preferred measurement thresholds are, then you will see response akin to those in camp 1 above.

 

[Curvature]

The science says that’s “true high-fidelity.”

1. Yes there are people prefer tube sound, for example, because they like distortion, this rolled off, or what the measurements can explain is the difference; or

2. Your preference is misplaced, you are listening to low-fi, and it’s probably because you were raised on ear buds/poor MP3 recordings, assuming you can even hear what you perceive to be a preference.

The science is asking: what is it that you are actually hearing? Are you interested in really learning how you hear what you hear? Are you interested in learning how devices work? And where they work differently, what distinguishes them, and furthermore, is that difference important or not?

If you can't be bothered to figure things out, you'll end doing arbitrary things for arbitrary reasons. Most of which is honestly ok, but where someone starts to care about what happens and why, and ask where they can find the most value, and how to understand the choices out there, then it would be wrong to mislead them.

 

[pseudoid]

People preferring “Distortion” (assuming it is the same or similar as added in by these plugins) is fundamentally contrary to what is considered to be the “best” in the reviews isn’t it? (Refer to Measurements, Everything or Nothing).

Untrained humans are like bad behaving pets inside a house.
Most of the Bose 901 buyers were like that... but I don't mean they pee'd on the floor.
That aside; I would think that the ONLY humans who can unequivocally answer that question (re: good distortion) are those in the production-side of the music. Anything else becomes a subjective criteria of added personal "preference".

 

[Travis]

Harmonics of a single cleanly played note on an electric bass guitar, electrical signal into an ADC,

What note or open string was that? EDIT, already answered thank you

 

[Travis]

The science is asking: what is it that you are actually hearing? Are you interested in really learning how you hear what you hear? Are you interested in learning how devices work? And where they work differently, what distinguishes them, and furthermore, is that difference important or not

I certainly agree with all of that. But also here, the way the science is applied is evaluation related to measurements. At least in the discussion of reviews, that is a significant part of the comments. Many, me included, do dig into the scientific areas mentioned, what are the differences, and are they significant. (There are many things they don’t know exactly why we can hear the things we do, like how we can tell that a perfect middle C came from a piano, or clarinet, or cello).

Generally speaking, with amplification, the accepted gold standard, since the 40s, has been harmonic distortion, standardized as THD specifications and further refined as THD + N. There are other things involved as well, but that’s the main one and the one that was chosen (expressed as SINAD) to be used in comparing all devices with that measurement from least to most and color coded. Objectively speaking, the lower the distortion the better (sounding) the amp. They correlate, better SINAD, better reproduction. The output signal varies by such a small degree from the input signal (distortion) that’s it is well below audible.

The interesting part about the OP, to me, was it raises some very interesting questions. The most fundamental question: Is this correlation correct? For a significant majority, 90+percent correlation? To get to that point other things have to be answered.

Is the distortion that the OP adds even analogous to the “distortion” being measured in the reviews (SINAD, THD, N). If it’s not, then it’s pretty much “salt and pepper” as someone mentioned earlier.

If it is analogous, and the tests are valid, and a significant number of people prefer distortion, what does that mean for distortion in amps? Does it mean for the people who prefer the distortion that tube amplification is the “best sounding” amp along with many other possible ramifications.