Question about audible differences: does a lower noise floor improve the sound?

[kemmler3D]

1. Is that possible that the absence of noise makes the equipment and/or speakers perform better because the noise is not interfering with the actual music signal?
2. If so, are there any real world tests that explore this? I would imagine such a test would have to be one in a anechoic chamber or an otherwise sound isolated space using a very sensitive microphone. But even if you were to capture that sound, you still have to compare it somehow and make sense of it.

No, inaudible noise (or anything else inaudible) can't have an audible effect if it's just linearly combined with the signal.

IMD seems like an appealing concept here, until you remember that IMD is usually -40dB below the signal even with a speaker cone, and we started out with the noise being inaudible, so we're 40dB below inaudible for that.

Noise won't somehow cause an exotic type of phase distortion, that's just wild nonsense.

And, as far as I know, inaudible noise doesn't cause subconscious effects on hearing that you then become secondarily aware of as bad sound quality. This is coming really close to arguing that placebo effect-type changes to the sound actually have a concrete cause... if you imagine they do. (think about that one.)

Could the speaker theoretically perform better if there was no noise at all in the signal?

The speaker doesn't know if it's playing noise or signal. Inaudible noise doesn't make itself audible through secondary effects, it's just there mixed in with the signal. Or, you can think of it like this, to the speaker, everything is signal.

 

[JSmith]

the absence of noise

Yeah that's why we test for noise levels... you know, the N in THD+N.


JSmith

 

[MattHooper]

I just watched Amir's video called "GR Research B24 AC Cable Review: Does it Make an Audible Difference?" and something occurred to me.

I get that running the digital file comparison test is proof that there is no audible difference between two cables, because the difference between the two files is just noise below the audible threshold. But audiophiles claim that there is an audible difference that can be explained by having a lower noise floor directly translating into better instrument separation and bigger soundstage. I have two questions:

1. Is that possible that the absence of noise makes the equipment and/or speakers perform better because the noise is not interfering with the actual music signal?
2. If so, are there any real world tests that explore this? I would imagine such a test would have to be one in a anechoic chamber or an otherwise sound isolated space using a very sensitive microphone. But even if you were to capture that sound, you still have to compare it somehow and make sense of it.

Well, unless we’re talking about something like vinyl playback in which background noise can in fact, be louder or quieter…

I have experienced the impression of a lower noise floor - the de jour phrase is
“ blacker background” - between different versions of tube amplifiers. But I do not attribute this to differences in “ noise floor” which I’m sure where well below my hearing threshold to begin with. It was either my imagination, or some sort of measurably audible difference, perhaps some slight frequency response deviation.

I’ve also experienced an impression of a noise floor or a “ blacker background” by absorbing more room reflections in my room.

With more room reflections, it tends to add a sort of “ room hash” over the sound that should’ve whitens and lightens everything.
Judiciously dining down the reflectivity seems to reduce this and the sonic impression is that of the Sonic images sitting against a
“ darker background.” At least that is what I perceive.

 

[pma]

The requested level of S/N strongly depends on types of music listened. Classical with deep ppp will ask for higher S/N than constant level DR4 pop. Ambient room noise level is also important. As always, there is no simple answer to a non-trivial question.

 

[Mnyb]

It is, but it also indicates AI is not capable of swimming in that sea of BS. So what use is it? Inaudible is inaudible, but sub-audible isn't. Yeah, right. I guess AI would be highly effective as an automated BS sales software. Lord help us as we already have too much of that.

I think it's perfectly capable of making the sea of BS and it's used for these purposes already , "content" creation and troll farms will be the arena for AI currently . Until it improves , it will be a dilemma as all trustworthy contents will be diluted to insignificance by then soo no learning material anymore

 

[antcollinet]

What I am saying is, might the equipment care if the noise is simply not there? Could the speaker theoretically perform better if there was no noise at all in the signal?

No - as I say it has no idea if it is reproducing noise or if the noise is signal. It will simply do what it is supposed to do.

 

[Blumlein 88]

I think it's perfectly capable of making the sea of BS and it's used for these purposes already , "content" creation and troll farms will be the arena for AI currently . Until it improves , it will be a dilemma as all trustworthy contents will be diluted to insignificance by then soo no learning material anymore

I'm seeing that already. My phone likes to push you to googles AI search. It seems pretty handy, but it always has wrong information woven in somewhere if it involves anything much at all. As this proliferates it will be much harder to find info, to learn something about a subject you don't already know. Now if they solve the problem like they were promising us they pretty much had in versions soon to be available as of a year ago, then no other type search could compete. Instead Brandolini's law will need revision. Brandolini's AI law. An AI can spew BS at a rate 1 million times faster than it can be corrected.

 

[JSmith]

An AI can spew BS at a rate 1 million times faster than it can be corrected.

Absolutely... in fact may get to the point of model collapse;

AI trained on AI churns out gibberish garbage

Eventually, it collapses—'poisoned with its own projection of reality.’

www.popsci.com

JSmith

 

[antcollinet]

Haha! That’s hilarious and it’s why I take everything ChatGpt says with a grain of salt.

Replace that grain with a couple of sacks. AI simply regurgitates the misinformation it is trained on. And it will do that with utterly convincing confidence.


Plus you asked a bad question, you asked:

Why do audiophiles claim that removing noise from an analog signal improves the imaging and soundstage?

It told you why (it 'believes') audiophiles claim this. It didn't say audiophiles are correct in those claims, or that those reasons are valid.

 

[Blumlein 88]

I used the Perplexity AI search and asked the question differently. It lists its sources. Reddit threads were a source, Audiogon forums, ASR forums and DIY audio along with a few others. It was a mismash of truth and audiophile lore. My question didn't mention audiophiles. It seems to me all these AI searches have deteriorated in the past 6 months. I haven't used the for pay versions of AI for more limited tasks like writing programming or such.

 

[boxerfan88]

audiophiles claim that there is an audible difference that can be explained by having a lower noise floor directly translating into better instrument separation and bigger soundstage.

Let’s add the room into the equation.

Our listening room, at the quietest possible, probably has a noise floor of 35dBA, or at best 30dBA.

Majority don’t play music at full volume, my guess is 75-85 dBSPL average.

All those noises 100dB down that some people think they can hear into the noise floor… soundstage… instrument separation … i really wonder if it is true or not… hmm…

(In a quiet room… most of us cannot even hear our heart beat…)

 

[Galliardist]

I remember "blacker backgrounds" and such from the late 1980s. They only happened with vinyl systems, some of which clearly measured having audible noise in a suitably quiet room (less than 45dB down). I was suspicious for a short time that reviewers were more comfortable with barely audible noise, than they were with the genuine silence of digital recordings. It didn't take long for the idea to become more random, though.

 

[Blumlein 88]

Let’s add the room into the equation.

Our listening room, at the quietest possible, probably has a noise floor of 35dBA, or at best 30dBA.

Majority don’t play music at full volume, my guess is 75-85 dBSPL average.

All those noises 100dB down that some people think they can hear into the noise floor… soundstage… instrument separation … i really wonder if it is true or not… hmm…

(In a quiet room… most of us cannot even hear our heart beat…)

That can be deceptive. We are much more sensitive in the 3-5 khz range. Room noise is also much lower in that range. Not uncommon for it to only be 10 db above the noise threshold in that 3-5 khz area in a room with normal noise levels. I find noise levels often look like a mix of pink and brown noise. Here is a plot of a brown/pink mix where I pegged the result with 105 db SPL in room as the max and where the noise would be if it were at 35 db SPL. The FFT splits it into 64 bands while our ear might do more like 32 bands. As you can see where our hearing is sensitive things are just a few db over what would be 0 db SPL (-105 db on the graph).

 

[Frank2]

Let's look at an analogy:
Is it possible that particles interfering with the fotons traveling from your TV screen to your eyes will affect image quality? Well yes, if your room is filled with smoke or water vapor this will affect image quality. But would you believe that a cold room yields a noticable better image quality than a warm room? After all, more heat means that particles in the air are more in motion which increases the probability of interference. On top of that there will be more warm body radiation which has a very small component in the visible spectrum at room temperature.

The effect is there until it isn't noticable anymore. There will always be some noise in any signal but at some point it will have 0 effect on your perception and on your speakers. Unless you strongly believe it has some effect. I bet I can convince some people to turn down their room temperature when watching movies with the argumentation above

 

[the_lumberjack]

Thank you everyone for your answers. Very enlightening.

 

[paulraphael]

That can be deceptive. We are much more sensitive in the 3-5 khz range. Room noise is also much lower in that range. Not uncommon for it to only be 10 db above the noise threshold in that 3-5 khz area in a room with normal noise levels. I find noise levels often look like a mix of pink and brown noise. Here is a plot of a brown/pink mix where I pegged the result with 105 db SPL in room as the max and where the noise would be if it were at 35 db SPL. The FFT splits it into 64 bands while our ear might do more like 32 bands. As you can see where our hearing is sensitive things are just a few db over what would be 0 db SPL (-105 db on the graph).

View attachment 405839

This. I was going to post something similar. An SPL meter always shows every room I've measured to have a 20db to 40db noise floor, but a spectrum analyzer shows it's mostly in the bass. So not only is it in a range where we don't hear so well, but it's not going mask detail at 2khz, for example.

Nevertheless, I doubt I've been in a listening room that would reveal the difference between, say, a -80db noise floor and a -100db noise floor.

Unrelatedly, I've sometimes found noise to contribute to a sense of space and liveness, especially on some older recordings. I have some old Charles Mingus recordings, where when I put them on, as soon as the noise of the recording begins, I get this illusion of the space opening up and me being present in it. I'm sure some of this noise is ambient sounds of the space in the recording, but most of is just noise. Probably a crappy 1960s tape machine, and a tape that's been abused for half a century. Anyone else ever get an impression like this? It makes me wonder how much the noise measurement really matters to listening enjoyment.

 

[Speedskater]

John Atkinson at Stereophile and others have speculated that a small amount of background noise can add sparkle or brightness.

I think that what the listener perceives with this cable is that at low levels, the sound is fattened and made more coherent-sounding by the dominant second-harmonic distortion. In addition, the presence of background noise cannot be dismissed, as there is some evidence that introducing small amounts of random noise results in a sound that is preferred by listeners. At higher signal levels, transients are accompanied by bursts of higher harmonics. However, these subside as quickly as they appeared. The overall effect is to render the system sound as being more vivid,
John Atkinson
August 2005

 

[antcollinet]

as there is some evidence that introducing small amounts of random noise results in a sound that is preferred by listeners.

1 - It'd be great if that evidence were actually presented, rather than hinted at. Assuming of course there is evidence, as opposed to a few old guys speculating in a forum somewhere.

2 - Even if so, in order for that to be true, the noise would have to be audible. If it is too low to be detectable by the human ear, it cannot create a preference. And for me at least every time (every single time) I've experience audible noise - it just sounds like noise: hiss, hum, pops or crackles. Perhaps even a dog barking. Sometimes it can be ignored when the music is playing - but it never somehow magically creates a perception of better sound.

 

[MaxwellsEq]

Hi @the_lumberjack . To answer the question "Question about audible differences: does a lower noise floor improve the sound?"

Simply, yes.

A very high noise floor masks the sound. Imagine listening to a quiet acoustic ballad next to a jet engine. Then the same ballad next to a washing machine on spin. Then the same ballad in a padded cell. As the noise floor gets lower, fine detail becomes more obvious, quiet sections become more intelligible and dynamic shifts are more apparent.

 

[antcollinet]

Hi @the_lumberjack . To answer the question "Question about audible differences: does a lower noise floor improve the sound?"

Simply, yes.

A very high noise floor masks the sound. Imagine listening to a quiet acoustic ballad next to a jet engine. Then the same ballad next to a washing machine on spin. Then the same ballad in a padded cell. As the noise floor gets lower, fine detail becomes more obvious, quiet sections become more intelligible and dynamic shifts are more apparent.

Bear in mind OP is asking about noise below the level of audibility causing audible impacts through other means. IE inaudible noise causing audible artifacts by somehow disrupting the operation of the reproduction gear.

To which the answer is a resounding "no it bloody well doesn't"