IPB

Welcome Guest ( Log In | Register )

2 Pages V   1 2 >  
Reply to this topicStart new topic
Is it possible to train yourself to be able to hear higher frequencies
dannytran1191
post Oct 3 2013, 10:19
Post #1





Group: Members
Posts: 3
Joined: 20-August 13
Member No.: 109704



Let's say an 18 year old man cannot hear frequencies above 16kHz. He hasn't suffered any ear damage and his sound system is perfectly capable of reproducing frequencies higher than 16kHz. The only reason why he can't hear them is because he was born unlucky. Can this man train himself to hear frequencies up to 20kHz, or is he stuck with the limits he was born with?
Go to the top of the page
+Quote Post
Roseval
post Oct 3 2013, 10:49
Post #2





Group: Members
Posts: 453
Joined: 26-March 08
Member No.: 52303



The sad news: you canít as it is a limitation of your ears
The good news: very few instruments hit the 16 kHz:
http://www.polkaudio.com/forums/showthread...cal-Instruments
Basically your hearing range donít differ from say an 40 years old
Due to my age, my hearing is gapped at 13 kHz and yes, I still enjoy music tremendously smile.gif


--------------------
TheWellTemperedComputer.com
Go to the top of the page
+Quote Post
probedb
post Oct 3 2013, 11:42
Post #3





Group: Members
Posts: 1122
Joined: 6-September 04
Member No.: 16817



Unless you can regenerate then no wink.gif

I was shocked when I found out I could only hear up to around 15.5KHz but I enjoy music. I've also noticed many IEMs only go up to 18KHz, even the very expensive ones.
Go to the top of the page
+Quote Post
greynol
post Oct 3 2013, 14:51
Post #4





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



HF content (>~16 kHz) is usually masked by lower frequency content in music anyway. People who can hear pure tones at, say 18 kHz usually cannot tell the difference between music with content at 18 kHz and the same music but with the content >16 kHz removed.

Feel free to try it yourself.

This post has been edited by greynol: Oct 3 2013, 14:52


--------------------
Your eyes cannot hear.
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Oct 4 2013, 10:06
Post #5





Group: Members
Posts: 3537
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (dannytran1191 @ Oct 3 2013, 05:19) *
Let's say an 18 year old man cannot hear frequencies above 16kHz. He hasn't suffered any ear damage and his sound system is perfectly capable of reproducing frequencies higher than 16kHz. The only reason why he can't hear them is because he was born unlucky. Can this man train himself to hear frequencies up to 20kHz, or is he stuck with the limits he was born with?


Two answers.

(1) The problem that has been presented appears to be speculative. I know of no evidence that any such person has ever existed. If you are trying to prove that you can fantasize a problem that cannot be solved, then you are frankly wasting our time because it is well known that impossible riddles can be created fairly easily.

(2) Hearing frequencies in a certain frequency range is contingent on the structure of the ear. If an ear lacks normal structures related to hearing in a certain frequency range then the reasonable alternative is to either rebuild the structures or circumvent them by some artificial means.
Go to the top of the page
+Quote Post
Martel
post Oct 5 2013, 07:35
Post #6





Group: Members
Posts: 534
Joined: 31-May 04
From: Czech Rep.
Member No.: 14430



Maybe through a transplant surgery? smile.gif


--------------------
IE4 Rockbox Clip+ AAC@192; HD 668B/HD 518 Xonar DX FB2k FLAC;
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Oct 5 2013, 11:49
Post #7





Group: Members
Posts: 3537
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (Martel @ Oct 5 2013, 02:35) *
Maybe through a transplant surgery? smile.gif


http://en.wikipedia.org/wiki/Cochlear_implant

Right now the implants are only good enough for people with profound ear damage, but you know how technology works.

Will they have implants so good that people with only moderately damaged hearing get them? Maybe even within our lives.
Go to the top of the page
+Quote Post
Mach-X
post Oct 5 2013, 15:45
Post #8





Group: Members
Posts: 242
Joined: 29-July 12
Member No.: 101859



QUOTE (greynol @ Oct 3 2013, 09:51) *
HF content (>~16 kHz) is usually masked by lower frequency content in music anyway. People who can hear pure tones at, say 18 kHz usually cannot tell the difference between music with content at 18 kHz and the same music but with the content >16 kHz removed.

Feel free to try it yourself.


Indeed. I was rather surprised when, while messing with lame, I downsampled some tracks to 32khz and couldn't hear a difference.
Go to the top of the page
+Quote Post
Rescator
post Oct 6 2013, 08:46
Post #9





Group: Members
Posts: 73
Joined: 13-January 09
From: Trondheim
Member No.: 65515



QUOTE (Mach-X @ Oct 5 2013, 16:45) *
I was rather surprised when, while messing with lame, I downsampled some tracks to 32khz and couldn't hear a difference.


Not that surprising, what is surprising is that you can potentially go even lower without noticing it. I just uploaded two samples that allows a quick way to test this:

http://www.hydrogenaudio.org/forums/index....howtopic=102914
Lowpass Stairstep and inverted Highpass Stairstep sample
A quick way to check your music frequency attention.


--------------------
"Normality exist in the minds of others, not mine!" - Rescator
Go to the top of the page
+Quote Post
Martel
post Oct 7 2013, 07:55
Post #10





Group: Members
Posts: 534
Joined: 31-May 04
From: Czech Rep.
Member No.: 14430



While I generally agree that content over 16kHz (15/14/...) is audible only in some recordings and only specific parts thereof, inaudibility in a single test sample should not be generalized. A carefully chosen sample can yield "opposite" results.

This post has been edited by Martel: Oct 7 2013, 08:04


--------------------
IE4 Rockbox Clip+ AAC@192; HD 668B/HD 518 Xonar DX FB2k FLAC;
Go to the top of the page
+Quote Post
bandpass
post Oct 7 2013, 09:36
Post #11





Group: Members
Posts: 321
Joined: 3-August 08
From: UK
Member No.: 56644



QUOTE (Roseval @ Oct 3 2013, 10:49) *

I don't think those graphs are complete/accurate; percussion especially can produce much higher frequencies than shown.
Of course, perceptual (masking) effects are likely to apply; maybe that's the intention of the referenced graphs, rather than to show the actual frequencies produced.
Go to the top of the page
+Quote Post
Alexey Lukin
post Oct 7 2013, 10:17
Post #12





Group: Members
Posts: 190
Joined: 31-July 08
Member No.: 56508



Every instrument's frequency spectrum is infinite, but these graphs probably show certain percentiles of spectral energy.
Go to the top of the page
+Quote Post
bandpass
post Oct 7 2013, 10:54
Post #13





Group: Members
Posts: 321
Joined: 3-August 08
From: UK
Member No.: 56644



Maybe, but the percentile seems much lower than a typical reader might expect, given that the actual spectra can look like this:

(from a recording of drums here: http://www.lessloss.com/drums-drums-drums-p-203.html )

Some of those hits look to have roughly the same energy above and below 16k.
Go to the top of the page
+Quote Post
Martel
post Oct 7 2013, 13:47
Post #14





Group: Members
Posts: 534
Joined: 31-May 04
From: Czech Rep.
Member No.: 14430



Thanks for the link, the samples show how real drums sound... Unlike 90% of commercial CDs which screw up the sound in some way.


--------------------
IE4 Rockbox Clip+ AAC@192; HD 668B/HD 518 Xonar DX FB2k FLAC;
Go to the top of the page
+Quote Post
pdq
post Oct 7 2013, 14:20
Post #15





Group: Members
Posts: 3305
Joined: 1-September 05
From: SE Pennsylvania
Member No.: 24233



QUOTE (Martel @ Oct 7 2013, 08:47) *
Thanks for the link, the samples show how real drums sound... Unlike 90% of commercial CDs which screw up the sound in some way.

Do you think it might be possible to ABX against a 44/16 version of this clip?
Go to the top of the page
+Quote Post
spoon
post Oct 7 2013, 14:32
Post #16


dBpowerAMP developer


Group: Developer (Donating)
Posts: 2725
Joined: 24-March 02
Member No.: 1615



We know that age reduces ability to hear high frequencies, however I wonder if there exist a % of people who can hear much higher frequencies than average, perhaps based on race (nature does like to change the variables based on environment).

Has any such study been done to identify?

This post has been edited by spoon: Oct 7 2013, 14:33


--------------------
Spoon http://www.dbpoweramp.com
Go to the top of the page
+Quote Post
greynol
post Oct 7 2013, 15:04
Post #17





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



QUOTE (Martel @ Oct 6 2013, 23:55) *
A carefully chosen sample can yield "opposite" results.

Good luck finding such a sample. My guess is that it'll be hanging out with a unicorn at the end of a rainbow.

Has anyone put that drum sample through a good low pass filter? Use just a filter, not a lossy codec. It should only attenuate high frequencies without causing artifacts such as ringing.

This post has been edited by greynol: Oct 7 2013, 15:13


--------------------
Your eyes cannot hear.
Go to the top of the page
+Quote Post
Martel
post Oct 7 2013, 15:20
Post #18





Group: Members
Posts: 534
Joined: 31-May 04
From: Czech Rep.
Member No.: 14430



By "opposite" (intentionally in quotes) I meant a sample where there's little masking to hide the loss of frequencies above 16kHz for someone who can naturally hear above 16kHz.

And no, I don't think it is possible to ABX 96/24 vs. 44/16 for vast majority of adult humans. I managed to ABX a 16kHz lowpass on a harpsichord sample (somewhere in the uploads section) which was extremely difficult. ABXing a 22kHz lowpass is just futile.


--------------------
IE4 Rockbox Clip+ AAC@192; HD 668B/HD 518 Xonar DX FB2k FLAC;
Go to the top of the page
+Quote Post
greynol
post Oct 7 2013, 15:29
Post #19





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



Yes, and I said "usually" twice (NB a more proper usage of quotes).

Regarding positive ABX results (aka "opposite"), I imagine deviating from a flat playback response can change the outcome (in either direction). Unfortunately this may not be so easily controlled.

This post has been edited by greynol: Oct 7 2013, 17:47


--------------------
Your eyes cannot hear.
Go to the top of the page
+Quote Post
Rescator
post Oct 12 2013, 01:38
Post #20





Group: Members
Posts: 73
Joined: 13-January 09
From: Trondheim
Member No.: 65515



QUOTE (greynol @ Oct 7 2013, 16:04) *
QUOTE (Martel @ Oct 6 2013, 23:55) *
A carefully chosen sample can yield "opposite" results.

Good luck finding such a sample. My guess is that it'll be hanging out with a unicorn at the end of a rainbow.

Has anyone put that drum sample through a good low pass filter? Use just a filter, not a lossy codec. It should only attenuate high frequencies without causing artifacts such as ringing.


No idea if the marquee of old Audition 1.5 in sprectral view is good or not.
But I noticed the drum sample LessLoss-Individual_drum_presentation--drum_set-16_bit-44.1_kHz-LL.wav had noise in the highs, and another faint one further down.

I did a marquee of the lower part of the audio spectrum and deleted the audio (using the undo feature allows quick high or low pass testing this way), I kept doing this from the top and down until eventually I heard something that resembled music in the highpass area. But at this point I was at 100% system volume and down to around 10KHz frequency range. Then I undid my last change and instead of a deleting the lower 10KHz I did the opposite, I deleted anything above 10KHz. I also took a full frequency and a 10KHz lowpass copy of the left channel and put them into a left and right and I found not audible difference in sound.

I also tested what would happen if I converted this odd paired channel comparison from 44KHz to 22KHz sampling rate (remember the right channel was turned into a duplicate of the left but with a 10KHz lowpass),
i set the quality of the conversion to minimum (which was 30 whatever that means), and I turned off pre/post filtering as well. It sounded just the same, and why should it not. 22KHz can reproduce 10KHz without issue.
I then converted from 22KHz to 44KHz and this time with 999 quality and pre/post filtering. and then I inverted the samples and blended them with the previous 44KHz, and the result was amusing, the "full" frequency channel did not null out, in fact it had sound from 22KHz down to 8KHz. While the lowpassed channel was almost nulled, it still had some audio content that was barely audible. Conclusion is that lowpassed content gets modified less by samplerate converstion (at least from 44KHz to 22Khz sample rate and back).

This only confirmed what I heard in my two test samples mentioned in this thread. If you toss away what is not identifiable as music/rythm/pattern/whatever (i.e. what just sounds like high freq noise), then nobody will notice it's gone. (some people may be able to ABX it, but who normally listens to music while ABXing? I certainly don't, Id' rather hear the next song instead).
As I said In that post, I'll say again, I'm surprised by how much you can toss away without noticing. For example the audio above 10KHz was only audible to me as high freq (and a tad painfull) hiss or constant tone similar, and this was at 100% system volume with headphones, I usually listen with volume at 20-25% and I consider that loud enough, even for stuff replaygained with a gain adjustment to make it hit -23dBFS reference instead.

It is difficult to ABX stuff when you lowpass stuff. It is much easier to do a highpass instead (and no need for ABX).
If I can't hear anything resembling rythm/music/beat/voice/whatever at full system volume with a highpass with a xxxxxxx Hz cutoff, then I know that I wont' loose anything if I do a lowpass with a cutoff at the same, + whatever margin needed for the lowpass rolloff so a highpass at 10KHz may need the lowpass at 11KHz if the rolloff is 1KHz for example..

But to get fully back on topic, no you can't train yourself to hear high frequencies.
You can train yourself to notice hgigh frequencies your hars can hear but that your mind just ignored.
And if you do a highpass and then keep moving the cutoff lower and lower you will soon find where the "music" starts. Then just increase the cutoff a little and raise the volume carefully. Do you hear anything? Does it sound like music at all?
In my case with my test sample and that drum sample listened to the answer is no (for me), and I highly doubt I have the discipline to train myself to ABX those high pitched "whines" (not sure what else to call them).

If I can't hear anything of audible interest using a highpass and max (or very high) volume, I do not even bother doing a ABX, it saves me a helluva lot of time.
I have not yet tried playing a sample back at half speed to see how those high frequencies I cant' really hear actually sound like, but again, I normally never listen to music at half speed so...


--------------------
"Normality exist in the minds of others, not mine!" - Rescator
Go to the top of the page
+Quote Post
Rescator
post Oct 12 2013, 01:53
Post #21





Group: Members
Posts: 73
Joined: 13-January 09
From: Trondheim
Member No.: 65515



QUOTE (dannytran1191 @ Oct 3 2013, 11:19) *
Let's say an 18 year old man cannot hear frequencies above 16kHz. He hasn't suffered any ear damage and his sound system is perfectly capable of reproducing frequencies higher than 16kHz. The only reason why he can't hear them is because he was born unlucky. Can this man train himself to hear frequencies up to 20kHz, or is he stuck with the limits he was born with?


Since my previous comment went off to the side quite a bit I'll try and help you out by saying the following:

Your question should probably be re-asked (maybe in a new thread) as:
"Is it possible to train yourself to unmask high frequencies at will?"

The human brain is pretty good at ignoring repetitive sounds/noise (and there is a lot of high frequency noise in our daily lives).
Someone with good hearing range (that goes into the really high freq areas) could be able to unmask mentally in a controlled/quiet environment.

So Yes! I believe that would be possible.

And Greynol (or some of the other oldtimers here) may correct me on this, but I believe this is actually what experienced ABXers on HA have actually done.

They have learned to focus their hearing/attention.

This post has been edited by Rescator: Oct 12 2013, 01:56


--------------------
"Normality exist in the minds of others, not mine!" - Rescator
Go to the top of the page
+Quote Post
greynol
post Oct 12 2013, 04:02
Post #22





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



QUOTE (Rescator @ Oct 11 2013, 17:38) *
It is difficult to ABX stuff when you lowpass stuff. It is much easier to do a highpass instead (and no need for ABX).
If I can't hear anything resembling rythm/music/beat/voice/whatever at full system volume with a highpass with a xxxxxxx Hz cutoff, then I know that I wont' loose anything if I do a lowpass with a cutoff at the same, + whatever margin needed for the lowpass rolloff so a highpass at 10KHz may need the lowpass at 11KHz if the rolloff is 1KHz for example..

This isn't what I was hoping for since it doesn't illustrate the point that masking can keep you from hearing frequencies that would otherwise be audible, but thanks for doing this anyway. I'm surprised by your results.


--------------------
Your eyes cannot hear.
Go to the top of the page
+Quote Post
Rescator
post Oct 16 2013, 03:06
Post #23





Group: Members
Posts: 73
Joined: 13-January 09
From: Trondheim
Member No.: 65515



QUOTE (greynol @ Oct 12 2013, 05:02) *
I'm surprised by your results.

As was I. Considering that a friend who's educating himself to be an expert in hearing did a test of my hearing, and other than a tinnitus induced "dip" at a certain frequency on my right ear (which I've been aware of for years obviously), my hearing seemed to be average/normal otherwise.

So I was surprised I could do such a severe lowpass and not miss anything musicwise, (the highpass correlated with the lowpass, the threshold of where I noticed a difference (without ABS) was roughly at the same point/frequency region).
Then again, many are fine with MP3 and AAC and Vorbis and Opus quality even at lower bitrates, and they too toss away a lot of high frequencies.

If I'll purely speculate here, it is that the ear/mind just zooms in on the frequencies we want to hear.

I guess to angle back at the original posts question...
Theoretically people may have a difference in how wide or narrow their hearing frequency focus band or HFFB (I just made that term up right now).
Now if that changes as you get older (along with your hearing) or if you can train to expand it (train yourself to focus on wideband) I have no idea.

ABXers if I'll guess is very narrow band, they try to listen wide at first but then usually focus on a certain range.
Maybe a (what would it be called, double or tripple ABX blind test?) would be of some scientific interest.

I.e. the test subject does not know what to listen for in their ABX and they do not know the audio material from previously (and thus no idea what to look for).
And then they can try to figure out which is the one with something missing vs the "original".
To use a weird term, virgin ABX samples. but good luck with that, it certainly could not be a snippet of any previously released music (as some testers might have heard it before).

But I'm getting sidetracked again.
And I'm basically just saying the same thing as earlier...that:
One can not train to hear better, but one can train to listen better.


--------------------
"Normality exist in the minds of others, not mine!" - Rescator
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Oct 30 2013, 18:42
Post #24





Group: Members
Posts: 3537
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (Alexey Lukin @ Oct 7 2013, 05:17) *
Every instrument's frequency spectrum is infinite, but these graphs probably show certain percentiles of spectral energy.


Infinite spectrum implies infinite energy...
Go to the top of the page
+Quote Post
kdo
post Oct 30 2013, 22:18
Post #25





Group: Members (Donating)
Posts: 304
Joined: 18-April 02
From: Russia
Member No.: 1812



QUOTE (Arnold B. Krueger @ Oct 30 2013, 20:42) *
Infinite spectrum implies infinite energy...

That's just wrong.

Energy (total) is finite when spectral energy density function has a convergent integral (which usually is, in real world).
What infinite spectrum does imply is infinitesimal (negligibly small) energy density at higher frequencies.
Go to the top of the page
+Quote Post

2 Pages V   1 2 >
Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 20th April 2014 - 15:44