Instantaneous Dynamic Range of Human Hearing?

[Dubesaur]

Hello all! There is something I have been wondering for a long time, but never could find an answer to:

I understand that the ear reacts differently in perceived eq at different volumes of sound. I understand this info to be contained on the Fletcher Munson curve. But, I can’t seem to find any information on how the dynamic range of hearing changes to different loudnesses of sounds, and especially what the maximum dbspl can be between two tones and a human still hear both tones (as in the louder one does not completely mask the quieter one). We can hear sounds from -10 to 120-130 dbspl, but certainly not both at the same time.

So, if I am listening to music at 80db, what is the quietest sound I could still hear? And, for other loudnesses, how does that dynamic range change?
Thanks!

 

[Wes]

This is a masking question, and frequency dependent. What freqs. are in the 80 dB music?

 

[Dubesaur]

This is a masking question, and frequency dependent. What freqs. are in the 80 dB music?

The music peaks at 80dB in the sub 200hz range and peaks at maybe 20dB less in the 15khz-20khz range with a relatively linear loss in dB from 100 to 15khz. I’m wondering how low the noise floor would need to be to not hear any noise while the music is playing.

 

[Helicopter]

Maybe 170dB from a pin drop to a 50BMG rifle. I am not sure how well you can hear the 50BMG over something like 375H&H, but I bet you know its louder from the difference in pain, ringing, and instant hearing loss. I have never had a reason to try either without world class hearing protection.

 

[RayDunzl]

I can’t seem to find any information on how the dynamic range of hearing changes to different loudnesses of sounds, and especially what the maximum dbspl can be between two tones and a human still hear both tones (as in the louder one does not completely mask the quieter one). We can hear sounds from -10 to 120-130 dbspl, but certainly not both at the same time.

You could do your own experiments.

Audacity will let you play two tracks at once, tones or music, with whatever attenuation you want on either track, and mute/unmute while playing to see if you can hear it or not.

REW allows you to generate tones with specified harmonic levels, and can turn the harmonics on and off on the fly.

Find out what matters to you in your environment.

 

[Blumlein 88]

I don't remember where, but I've seen some investigation indicating the ear had an instantaneous dynamic range of 60 db. Maybe @j_j knows the answer. My guess being he'll tell us it isn't that simple.

 

[dougi]

Maybe 170dB from a pin drop to a 50BMG rifle. I am not sure how well you can hear the 50BMG over something like 375H&H, but I bet you know its louder from the difference in pain, ringing, and instant hearing loss. I have never had a reason to try either without world class hearing protection.

I remember in my Army Reserve days during training they forgot to give a colleague of mine earplugs when firing the Charlie Gutsache. He suffered immediate and permanent hearing loss. Even with earplugs it was extremely loud and disorienting. (of course my ears still rang for days as well).

 

[Blumlein 88]

Maybe 170dB from a pin drop to a 50BMG rifle. I am not sure how well you can hear the 50BMG over something like 375H&H, but I bet you know its louder from the difference in pain, ringing, and instant hearing loss. I have never had a reason to try either without world class hearing protection.

I had a rifle once with a very, very effective recoil reducer. .308 caliber. I would wear good foam ear inserts and put good ear muffs on over that. It was still a bit painful to shoot it was so loud. I'd warn others at the range and stay as far away as possible. Some scoffed at the idea, and then moved away after a couple shots.

The fellow who sold it to me said he was selling it because it was too loud. After I had used it twice, I sold it the same way. Don't even want to think about what some of guys hear in the military on artillery.

 

[j_j]

I don't remember where, but I've seen some investigation indicating the ear had an instantaneous dynamic range of 60 db. Maybe @j_j knows the answer. My guess being he'll tell us it isn't that simple.

Damn good guess.

Across 20-20K, the ear can, just barely, with the right stimulus, handle as much as 80dB change in slope across the entire frequency range, at a given level. This is referring to slope in short-term spectrum at one time.

HOWEVER, the ear can, at mid frequencies, handle more than 80dB change in level quite handily. That won't change the answer above at all.

60dB is a bit low for simultaneous signals, but frankly isn't far off for almost all average, real signal.

 

[solderdude]

I once tested this and came to about 70dB dynamic range and like to give it 10dB 'headroom' so would say 80dB is safe to assume is enough for normal to loud listening levels.

I have a file that plays music with peaks at almost 0dBFS peaks that are attenuated by specific amounts of attenuation all the way down to 90dB (in 10 dB steps from -50 to -90.
Experiment with that and you may find that when listening pretty loud a sudden 70 or 80dB drop of level is enough for you to hear silence.

I'ts 8.5MB so too big as an attachment. I could file-mail it and place a link here.

 

[KSTR]

The overall dynamic range of the ear is so large only because there is a compresser built in: https://en.wikipedia.org/wiki/Acoustic_reflex
So, for two stimuli at the same time (so a fixed amount of compression), apart from masking issues, the dynamic range is much smaller. As soon as there is enough of a time gap between stimuli, the compressor can regulate and make the apparent range greater.

 

[Speedskater]

Isn't there a test track that plays music at normal volume, then drops the level in 10 or 20dB increments?

 

[j_j]

The overall dynamic range of the ear is so large only because there is a compresser built in: https://en.wikipedia.org/wiki/Acoustic_reflex
So, for two stimuli at the same time (so a fixed amount of compression), apart from masking issues, the dynamic range is much smaller. As soon as there is enough of a time gap between stimuli, the compressor can regulate and make the apparent range greater.

https://www.aes-media.org/sections/pnw/pnwrecaps/2019/apr2019/

You can't leave out the cochlear filtering, either.

 

[KSTR]

^ Just went through the slides, interesting stuff

btw, @j_j , with you being an expert in the field, what do you think of the Heerens & DeRu physical theory of hearing?

 

[danadam]

audiocheck.net has test files with pink noise alternating between full scale and N dB below full scale: https://www.audiocheck.net/testtones_dynamic.php

 

[j_j]

^ Just went through the slides, interesting stuff

btw, @j_j , with you being an expert in the field, what do you think of the Heerens & DeRu physical theory of hearing?

It would take a while to go through it, but there are a number of theories of cochlear dynamics that do not incur the objections I see at the start. They discuss a single, particular theory, and their description as a travelling wave, or "ripples in the pond" do not really express any of the more modern theories I know of. Certainly there is no "surface" inside the cochlea in the sense of a pond, and so the idea of travelling waves in that sense just doesn't really fit anything much.

Zwislocki's theory is where I would head for a good physics discussion. It avoids the amplifier hypothesis, explains compression, and matches the actual anatomy of the organ of Corti.

 

[KSTR]

@j_j, I've now watched the whole video of your AES presentation twice (with a sleep break in between), this is brilliant stuff! In those two hours I've probably learned more about our hearing than in my whole life before that.

As for the Heerens&DeRu paper, well, even if they are wrong/incorrect/outdated I'm still puzzled by the experiments they propose most of which I've recreated myself (they also offer a software tone generator to set up the examples). Especially the last set (experiments 3.8.x, and highlighting 3.8.9) I found very enlighting and their proposed differentiating and squaring process (distortion right at the "input") seems to offer a model/explanation why we perceive these test signals the way we do. Of course I'm not knowledgable enough to see if any other theories and models would also give an explanation/prediction that fits.

 

[j_j]

@j_j, I've now watched the whole video of your AES presentation twice (with a sleep break in between), this is brilliant stuff! In those two hours I've probably learned more about our hearing than in my whole life before that.

As for the Heerens&DeRu paper, well, even if they are wrong/incorrect/outdated I'm still puzzled by the experiments they propose most of which I've recreated myself (they also offer a software tone generator to set up the examples). Especially the last set (experiments 3.8.x, and highlighting 3.8.9) I found very enlighting and their proposed differentiating and squaring process (distortion right at the "input") seems to offer a model/explanation why we perceive these test signals the way we do. Of course I'm not knowledgable enough to see if any other theories and models would also give an explanation/prediction that fits.

It's really hard to know WHERE an effect comes from, from "entering the cochlea" to "central nervous system".

 

[Ron Texas]

Shotguns are really loud.

 

[j_j]

Shotguns are really loud.

Indeed, and at the muzzle, far, far higher intensity than the level where air simply ceases to be a linear medium.