Devil's Advocate

[Anton S]

Let's have a little fun today ...

Assertion: The background noise level of a typical home listening environment is at least 40 dB SPL. True or False?

If you agree with the above premise, then this conclusion must follow …

Audio equipment noise levels below than -80 dB are inaudible in a typical home listening environment.



Assertion: In a typical home listening environment, audio equipment noise levels below -80 dB are inaudible. True or False?

If you agree with the above premise, then this conclusion must follow …

Ranking electronics with noise levels less than -80 dB based on their S/N ratio is moot.



Assertion: Harmonic distortion levels produced by an audio system’s transducers are orders of magnitude greater than that of its electronics. True or False?

If you agree with the above premise, then this conclusion must follow …

An audio system’s transducers are by far the weakest links in any audio equipment chain with respect to harmonic distortion.



Assertion: An audio system’s transducers are by far the weakest links in any audio equipment chain with respect to harmonic distortion. True or False?

If you agree with the above premise, then this conclusion must follow …

Harmonic distortion differences among various electronic components are negligible by comparison with those among transducers.



Assertion: The generally accepted goal of recorded music listening is musical enjoyment. True or False?

If you agree with the above premise, then this conclusion must follow …

Analytical listening, which distracts from one’s musical enjoyment, conflicts with the generally accepted goal of music listening.



Assertion: Not all recorded music productions are created equal. True or False?

If you agree with the above premise, then this conclusion must follow …

Refusal to adjust an audio playback system’s transfer function, as needed, to obtain a pleasing and convincing recreation of a recorded performance is self-defeating with respect to the generally accepted goal of recorded music listening.

 

[voodooless]

Except that you can hear signals below a noise floor…

 

[Timcognito]

Don't frequencies of the noise play a role?

 

[Geert]

Ranking electronics with noise levels less than -80 dB based on their S/N ratio is moot.

Not if the ranking represents a dimension of engineering excellence.

 

[NTK]

Ranking electronics with noise levels less than -80 dB based on their S/N ratio is moot.

It is sort of like whether you should be totally happy with food that just met the minimum standard on the amounts of natural or unavoidable defects in foods for human use that present no health hazard.

Food Defect Levels Handbook

Levels of natural or unavoidable defects in foods that present no health hazards for humans.

www.fda.gov

 

[GaryY]

Even if I fully agree, then I can say we can't hear warm sound of tube/ A class or AB class because it's not audible and there is no need to look for warm sound.

 

[Prana Ferox]

Most of this is fair but you do have to remember that noise / distortion in an upstream component is amplified by the gain of downstream components just like the audio signal. So a very low noise component can be made quite audible downstream. Phono playback being the most obvious example of this, or high sensitivity tweeter hiss.

 

[DLS79]

Let's have a little fun today ...

Assertion: The background noise level of a typical home listening environment is at least 40 dB SPL. True or False?

If you agree with the above premise, then this conclusion must follow …

Audio equipment noise levels below than -80 dB are inaudible in a typical home listening environment.

Closed back headphones, noise canceling headphones and IEMS are a thing.

Assertion: The generally accepted goal of recorded music listening is musical enjoyment. True or False?

If you agree with the above premise, then this conclusion must follow …

Analytical listening, which distracts from one’s musical enjoyment, conflicts with the generally accepted goal of music listening.

So do people who make a living making music or other forms of audio recordings no matter because their job requires Analytical listening?

 

[DVDdoug]

I agree with most of that. But noise does get "complicated".

-80dB is very quiet. If you open a file in Audacity, listen to it, and then lower the volume in 10 or 20dB steps you probably won't hear anything but the time you get to -80, except you might hear some amplifier hiss.

I don't even bother looking at distortion specs. I've never heard distortion (from a digital source) unless something was broken or unless an amplifier was over-driven into clipping, etc.

I don't really worry about frequency response from modern electronics either.

Tubes are another story. It's not easy or cheap to make a good tube amp, especially a power amp.

Analytical listening, which distracts from one’s musical enjoyment, conflicts with the generally accepted goal of music listening.

That is a psychology question and it depends on the listener. I don't intentionally listen for defects, except for the rare occasions when I'm cleaning-up digitized vinyl.

 

[olieb]

Audio equipment noise levels below than -80 dB are inaudible in a typical home listening environment.

-80dB is certainly not the threshold when noise becomes inaudible in all home listening situations.
If I listen with an average SPL of 90dBspl (not even that loud for a fortissimo passage) this will often mean that 0dB corresponds to 105-110dBspl, in many classical recordings even more. So noise level from equipment will be about 30dBspl (or higher).

While the noise level in a quiet room is about 40dBspl that is not what you hear. A-rated (for the sensitivity of hearing) it will be closer to 30dB(A) as room noise typically is low frequency dominated.
30dB white noise from a DAC or amp can be heard on top of that (hiss) in low level passages (pianissimo) or pauses (with high volume as above).
(Recordings from the loudness wars will behave differently of course.)

And if you use digital volume control (with some headroom) even -90dB or -100dB SNR might still result in audible hiss though this will be quite faint in most cases.
Below that you will start to be on the safe side unless you screw up your gain structure too much. (And don't forget: noise from dac, preamp and power amp will add up.)
But then there are headphones where noise becomes more apparent again.
On the other hand, often enough the recording room has the highest noise level in the chain.

 

[MaxwellsEq]

Noise is interesting and complicated. Room background noise, analogue electronic noise and digital noise are all quite different things. Not all noises mask other noise (and in physical and some electronic noise, we can recover information below the noise floor! Noise cancelling have different frequency behaviours again changing the masking implication. So your -80dB model is not necessarily correct.

 

[Blumlein 88]

Try this for just a goof. Made it for another thread. A snippet of music with some quiet parts. Another song is mixed in 56 db below the average level of the music you'll hear. As the loud music has an average level of something like -18 db, the average level of the quiet part is around -75 db. Let us know if you hear it. Try speakers and then headphones.

Mixed signals.zip

Shared with Dropbox

www.dropbox.com

 

[restorer-john]

This is one of those threads destined to go nowhere fast.

 

[Anton S]

My sincere thanks to everyone who participated. Observing the distribution of presumptions and internalizations to what I tossed out as chum in the water, as well as the widely diverse responses, was genuinely fascinating. Thank you one and all for playing this little game.

 

[GXAlan]

Audio equipment noise levels below than -80 dB are inaudible in a typical home listening environment.

The best way to think about noise in your home is the total residual voltage of your electronics and what that translates into your speakers…

You are correct in that active speakers from companies like Meyer Sound, Genelec, and Neumann will have more noise at idle than a premium “audiophile” setup, but at typical listening levels, you are far from having audible noise.

Ranking electronics with noise levels less than -80 dB based on their S/N ratio is moot.

Again, we should be ranking amplifier noise by residual voltage and doing the math to determine listening distance and speaker efficiency to determine what matters or not.

An audio system’s transducers are by far the weakest links in any audio equipment chain with respect to harmonic distortion.

Agreed, except certain tubes intended to run into overdrive at even lower volumes.

Harmonic distortion differences among various electronic components are negligible by comparison with those among transducers.

This is still cumulative and there are differences in frequency dependent distortion.

Analytical listening, which distracts from one’s musical enjoyment, conflicts with the generally accepted goal of music listening.

No, that’s just your goal. Analytical listening can be just as satisfying as relaxed listening.

Analytical listening in a different form:

Refusal to adjust an audio playback system’s transfer function, as needed, to obtain a pleasing and convincing recreation of a recorded performance is self-defeating with respect to the generally accepted goal of recorded music listening.

Premise is accurate, but imagine a poor recording of a family member’s recital performance. I have an old vinyl LP of Abraham Lincoln’s speech as featured at Disneyland/Disney World. As it was produced in the 1960s, it is a technically flawed recording and yet it is enjoyable as a convincing recreation of the experience at the World’s Fair.

 

[Vacceo]

At least 40 db for starters? I just measured my room and it hardly reaches 30db peaks, most of the time is around 23db...

 

[tmtomh]

So to sum up:

Assertion 1 is often true but not always.
Even when Assertion 1 is true, the stated conclusion is not, actually, inevitable or always true.

Assertion 2 is almost always true.
The stated conclusion does not automatically follow from the assertion.

Assertion 3 is true for any competently designed and well-performing electronics, but not if we're talking about tubes used with low-distortion speakers and/or headphones.
The stated conclusion is true when the assertion is true, because the conclusion is more or less just a restatement of the assertion.

Assertion 4 is the conclusion from Assertion 3.
The stated conclusion is true whenever the assertion is true, because once again the conclusion is essentially just a restatement of the asserton.

Assertion 5 is true.
The stated conclusion contains its own unsupported and invalid assertion and does not follow from or have anything particularly to do with the assertion.

Assertion 6 is true, though so vaguely defined as to be meaningless because it conflates differences in quality (better/worse) with differences in kind (same/different).
The stated conclusion is not valid, because of the fatally - and in my view intentionally - fuzzy use of "equal."


So Assertion 2 is based on a incomplete understanding of noise, and Assertions 3 through 6 rely in one way or another on the rhetorical fallacy of begging the question. (And Assertions 3 and 4 appear to implicitly rely on ignoring the additive quality of distortion as the signal passes through all the components of an audio system.)

Only part I'm still not clear on is when the "having a little fun" part of the thread is supposed to happen.

 

[SKBubba]

To me, just some old geezer who isn't an EE, it seems logical that there are lots of environmental sources/causes of noise and distortion. Some you can correct, some you can't.

And there's a multiplier effect in the playback chain. If the source has x% of distortion, the amp has y% of distortion, and speakers have z% of distortion, the end result gets exponentially worse along the way. (Or something. I'm not good at math.)

So, it seems desirable to seek the lowest possible/achievable noise+distortion in every component in the chain to minimize the overall undesired effect on the final output, or "sound". Tempered of course by budget, desired features, etc.

I am happy to be corrected and educated on any of my misconceptions.

 

[Justdafactsmaam]

Let's have a little fun today ...

Assertion: The background noise level of a typical home listening environment is at least 40 dB SPL. True or False?

If you agree with the above premise, then this conclusion must follow …

Audio equipment noise levels below than -80 dB are inaudible in a typical home listening environment.



Assertion: In a typical home listening environment, audio equipment noise levels below -80 dB are inaudible. True or False?

If you agree with the above premise, then this conclusion must follow …

Ranking electronics with noise levels less than -80 dB based on their S/N ratio is moot.



Assertion: Harmonic distortion levels produced by an audio system’s transducers are orders of magnitude greater than that of its electronics. True or False?

If you agree with the above premise, then this conclusion must follow …

An audio system’s transducers are by far the weakest links in any audio equipment chain with respect to harmonic distortion.



Assertion: An audio system’s transducers are by far the weakest links in any audio equipment chain with respect to harmonic distortion. True or False?

If you agree with the above premise, then this conclusion must follow …

Harmonic distortion differences among various electronic components are negligible by comparison with those among transducers.



Assertion: The generally accepted goal of recorded music listening is musical enjoyment. True or False?

If you agree with the above premise, then this conclusion must follow …

Analytical listening, which distracts from one’s musical enjoyment, conflicts with the generally accepted goal of music listening.



Assertion: Not all recorded music productions are created equal. True or False?

If you agree with the above premise, then this conclusion must follow …

Refusal to adjust an audio playback system’s transfer function, as needed, to obtain a pleasing and convincing recreation of a recorded performance is self-defeating with respect to the generally accepted goal of recorded music listening.

I would not set the bar at typical room noise.

 

[MaxwellsEq]

And there's a multiplier effect in the playback chain. If the source has x% of distortion, the amp has y% of distortion, and speakers have z% of distortion, the end result gets exponentially worse along the way. (Or something. I'm not good at math.)

The OP is only talking about noise. Distortion and noise sound very different (we are sensitive to noise and not really to distortion) and are mostly created by different mechanisms.

The end-to-end multiplication effect you mention is not absolutely correct. Throughout the chain, gain can indeed increase noise and/or distortion, but we also attenuate along the chain, otherwise everything would be always at maximum volume and way too loud. Attenuation done properly reduces noise and distortion. The art of the system is to have exactly the correct amount of gain at the right places with the correct amount of attenuation at the right places to minimise noise and maximise signal.