• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

White noise improves hearing

pozz

Слава Україні
Forum Donor
Editor
Joined
May 21, 2019
Messages
4,036
Likes
6,825
So a type of spectral unmasking related to the "cocktail party effect"? This suggests a lot.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
Stochastic resonance has been studied in numerous fields. https://en.wikipedia.org/wiki/Stochastic_resonance

Funny, I've known this for decades, even before my first job after college, but not under "stochastic resonance". And in many applications, from audio to light transmission. I did not take the stochastic processes course in college (was on the list, but Life and Work intervened) so have had minimal exposure to them through other classes and various work assignments.
 

scott wurcer

Major Contributor
Audio Luminary
Technical Expert
Joined
Apr 24, 2019
Messages
1,501
Likes
2,821
It has of course been offered as a factor in the appeal of vinyl.
 

Eirikur

Senior Member
Joined
Jul 9, 2019
Messages
318
Likes
509
That is a massively interesting read. The experiments cannot easily be repeated on ourselves, let's stick to subjective for this:
A custom-made stainless-steel head-restraint post was fixed on the bone [of the mouse] on top of the left hemisphere, and used to head-fix the animals. Using a scalpel, a craniotomy was performed just above the auditory cortex.

Surprisingly high levels of noise were induced, compared to the actual signal
Pure tones [...] were played [...] at 60 dB SPL [...]
For the WN experiments [...] 50 dB SPL (bandwidth of 1 to 64kHz) was played in addition to the tones

In the past I performed an experiment on myself, playing pink noise in the background (at work) at a level just above my hearing threshold. Subsequently I needed to turn down the volume a few times; somehow the steady hiss increased acuity and forced me to redefine "background".
I also noticed that turning the noise down below conscious hearing level didn't make it imperceptible. It did become unpleasant, perhaps because my brain couldn't link the (sound)pressure to something audible anymore?

PS: pink noise is much more agreeable to me than white noise, apparently I like curves (* insert obscene picture supplied by Thomas here *)
 
Last edited:

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
Most folk find audio pink noise (equal energy per octave, downward slope with frequency) more pleasing the white nose (equal energy per decade, flat slope).
 

scott wurcer

Major Contributor
Audio Luminary
Technical Expert
Joined
Apr 24, 2019
Messages
1,501
Likes
2,821
Most folk find audio pink noise (equal energy per octave, downward slope with frequency) more pleasing the white nose (equal energy per decade, flat slope).

The concept of noise and pleasing does not compute with most folks. I won't belabor my soft spot for Japanoise music, though I have not studied the spectra of Masonna or Merzbow.

Funny thing, people have put a lot of time and energy into mathematically analyzing Jackson Pollock's drip paintings to prevent fakes (one take on them is pleasing visual noise). I would think DNA testing of the cigarette butts dropped in them would seal the deal.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
Sounds like dithering.

Related but not quite equal... Dither (noise decorrelation) can have a variety of different noise profiles and usually refers to decorrelating quantization (ADC/DAC) spurs so (e.g.) the noise floor from your DAC sounds more "white" or "analog". Dither can be considered a specific application of noise added to a signal. The end effect is essentially the same.

This is my off-the-cuff thought based on my knowledge base; others may have different/better descriptions.
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
Related but not quite equal... Dither (noise decorrelation) can have a variety of different noise profiles and usually refers to decorrelating quantization (ADC/DAC) spurs so (e.g.) the noise floor from your DAC sounds more "white" or "analog". Dither can be considered a specific application of noise added to a signal. The end effect is essentially the same.

This is my off-the-cuff thought based on my knowledge base; others may have different/better descriptions.
I thought more about dither as helping audio signals to jump over a threshold. Like in ADCs where dithering helps signals below the LSB to be digitized. Similar to knocking with your finger on a barometer to help the needle to break free and move.
 
Last edited:

Eirikur

Senior Member
Joined
Jul 9, 2019
Messages
318
Likes
509
I thought more about dither as helping audio signals to jump over a threshold. Like dithering in ADCs where dithering helps signals below the LSB to be digitized. Similar to knocking with your finger on a barometer to help the needle to break free and move.
I don't think so in their experiment, the injected WN has a hefty volume compared to the signal, only 10dB SPL lower!
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
I thought more about dither as helping audio signals to jump over a threshold. Like in ADCs where dithering helps signals below the LSB to be digitized. Similar to knocking with your finger on a barometer to help the needle to break free and move.

No, it is more like moving the threshold around so it is more random, making it easier to hear patterns (like music). If I had to use an analogy, think of a rock sitting in the grass just below the tops of the leaves. Dither is like a random breeze that whips the grass around but the rock stays put and is easier to see.

HTH - Don
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
No, it is more like moving the threshold around so it is more random, making it easier to hear patterns (like music). If I had to use an analogy, think of a rock sitting in the grass just below the tops of the leaves. Dither is like a random breeze that whips the grass around but the rock stays put and is easier to see.

HTH - Don
I see your point but in case of an ADC the dither is added to the digitized signal, not to the reference voltage of the ADC - or is it? I know there is a technique where you add dither before the ADC and subtract it afterwards to conteract the noise added by the dither.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
I see your point but in case of an ADC the dither is added to the digitized signal, not to the reference voltage of the ADC - or is it? I know there is a technique where you add dither before the ADC and subtract it afterwards to conteract the noise added by the dither.

I have always added noise before the ADC (in the analog domain) but it can also be added after quantization. But if you want to randomize the quantization of the sampled signal it is usually best to do it before it is quantized. Dither can also be added in the digital (before) or analog (after) domain of the DAC, naturally. There are trades in resolution (amplitude and time) in applying dither in the analog or digital domain for an ADC or a DAC. I am not sure this is the place to delve that deeply, and it has been some years since I really dug into the analyses. There are many schemes for noise decorrelation, and often (at least IME) the choice depends upon the application and ADC/DAC architecture.

I do not recall ever adding noise to the reference voltage of an ADC; I summed it with the signal itself. Mucking with the reference can be dangerous and cause other issues. That said, I have designed and used multiplying DACs that intentionally modulate the reference.
 
Top Bottom