IPB

Welcome Guest ( Log In | Register )

3 Pages V   1 2 3 >  
Reply to this topicStart new topic
How high can you hear (with music & lowpass), ABX required
How high can you hear?
You cannot see the results of the poll until you have voted. Please login and cast your vote to see the results of this poll.
Total Votes: 145
Guests cannot vote 
KikeG
post Dec 16 2002, 12:26
Post #1


WinABX developer


Group: Developer
Posts: 1578
Joined: 1-October 01
Member No.: 137



I've uploaded some test samples at http://www.kikeg.arrakis.es/lowpass in order to test how high can people hear with a clip of 'special' but real (pop) music.

These clips have a very strong high-frequency impulsive peak, at the frequency specified for every sample, and there is a lowpassed clip for each one with just that frequencies an above removed. All clips are resampled versions of the same clip, (where the strong peak is originally at 16300 KHz) , so that for every sample the peak is at a different frequency.

In order to know if you can hear the strong impulsive peak for every sample, you must be able to ABX the test sample against the lowpassed version of it. ABX is required in order to have meaningful results.

I have available just the clips shown here because I have no more space in my web page.

The 12 KHz clip should be easily ABXable for nearly anyone, specially if they are not very old. On casual listening (without concentrating) and using good soundcard & headphones (Audiophile 2496, Sennheiser HD 580) I can hear a difference on the lowpassed 17.5 KHz clip. Concentrating I could ABX the 18 KHz file, with results of 12/14 (p=0.6 %).

This should be a more realistic test of how high can people hear in extreme cases of somewhat 'real' music.

This post has been edited by KikeG: Dec 16 2002, 16:42
Go to the top of the page
+Quote Post
Guest_SK1_*
post Dec 16 2002, 13:48
Post #2





Guests






To me this is even funny.. The difference is SO VERY obvious to me with all the samples except the 20khz one. ABX'd in seconds 8/8 each time. With the 20khz one though it's harder, but 10/10. The major difference is noticable in the ringing sound. It's easy for me, ah i feel good now, thanks for this test KikeG smile.gif.
Go to the top of the page
+Quote Post
daniel
post Dec 16 2002, 14:36
Post #3





Group: Members
Posts: 86
Joined: 16-January 02
Member No.: 1047



:'( :'( 14khz abx:12/12
Will rest my ears for few days and try later.

This post has been edited by daniel: Dec 16 2002, 15:38
Go to the top of the page
+Quote Post
caligae
post Dec 16 2002, 14:52
Post #4





Group: Members
Posts: 186
Joined: 23-January 02
Member No.: 1132



QUOTE (KikeG @ Dec 16 2002 - 01:26 PM)
This should be a more realistic test of how high can people hear in extreme cases of somewhat 'real' music.

I don't get this. Why should resampling provide a more realistic scenario?

Wouldn't it be more real-world like if you just lowpass the same sample at different frequencies?

Anyway. I could abx up to 18kHz. Using "normal" samples on ff123.net my cutoff is way earlier. To test how high you can hear, you can alway use some sinus waveform.
Go to the top of the page
+Quote Post
KikeG
post Dec 16 2002, 16:04
Post #5


WinABX developer


Group: Developer
Posts: 1578
Joined: 1-October 01
Member No.: 137



@caligae:

I have used resampling simply to generate musical signals with a clear, strong peak at a particular frequency, whilst not being simply pure tones (sinus waveforms as you say). Using music instead of tones, masking effects are more taken into account, effects that with a single isolated tone aren't. Also, I hope that aliases caused by poor resampling in some soundcards are less audible than with pure tones, because it is more probable they get masked.

With pure tones, I can easily hear up to 18.5 KHz if I crank up the volume. With these test signals, maybe if I cranked up the volume and put lots of effort I could ABX the 18.5 KHz sample, but at normal conditions I can't. Even at those extreme conditions I'm not sure I could ABX them, due to ear internal masking.

Edit: actually, this test tries to emulate 'worst-case' musical signals. Whilst the original sample has this strong 'dominating' peak at 16 KHz all over the whole song, I don't think it can be common music with of this kind of peaks over 17 KHz. I believe that with regular music (even the original non-resampled file of this test), a lowpass of 17 or 17.5 KHz would be very difficult to hear.

This post has been edited by KikeG: Dec 16 2002, 16:25
Go to the top of the page
+Quote Post
KikeG
post Dec 16 2002, 16:16
Post #6


WinABX developer


Group: Developer
Posts: 1578
Joined: 1-October 01
Member No.: 137



@SK1

Man, you have bat ears. If anyone could host more samples, I could make available similar 20.5 KHz, 21 KHz and 21.5 KHz samples in order to fully test your ears.

However, the 21.5 KHz peak would be on the limit of possibilities of reconstruction of 44.1 KHz DACs. Mmm... I can upsample them to 48 KHz. That would be also good for the other files, but maybe this way they all won't fit in my web space... I have to check it out.

What is your equipment? I read it, but I can't remember now.
Go to the top of the page
+Quote Post
Pio2001
post Dec 16 2002, 17:34
Post #7


Moderator


Group: Super Moderator
Posts: 3936
Joined: 29-September 01
Member No.: 73



Same as with the sweep tone, resampling in most soundcards will get any file ABX able even at 21 kHz lowpass, because it will change the low frequency content. Use non resampling soundcards or burn CDs.
Go to the top of the page
+Quote Post
ErikS
post Dec 16 2002, 17:35
Post #8





Group: Members
Posts: 757
Joined: 8-October 01
Member No.: 247



18.5 (4/4 ABA)

I was surprised... 16-17 kHz lowpass used to be inaudible before. Is it because the special clip now?

This post has been edited by ErikS: Dec 16 2002, 17:36
Go to the top of the page
+Quote Post
Guest_SK1_*
post Dec 16 2002, 17:54
Post #9





Guests






I don't think i have bat ears smile.gif. I'm 19.5, so i guess i have good hearing fit for this age. Maybe better than of most but i'm really not sure. However, i was really bummed when i called my brother to try hear if there's a difference between the 20khz samples, because it seemed it was so much easier for him, damn! smile.gif He's 16 years old. Didn't ABX, he's not into that, never did any listening test either smile.gif i just had to try though, asked him what's "sharper" or stronger, he listened a few times and then answered, and then each time answered correctly.
Actually, i think i should have written it was -much- harder to notice the difference with the 20khz samples. With everything below it's..relaxingly easy for me but 20khz is totally different. I think the only difference i can hear is that the "riiing" is louder, stronger, wouldn't say i really notice a difference in sharpness like with 19khz.
21 khz and 21.5 khz, wooo, i think that's impossible for me smile.gif i'll never hear a difference, no way. With sweeps (not "real music") it's possible maybe, but naa, no WAY with music smile.gif.

I did this test with speakers that had their tweeters replaced by Boss TW-17car tweeters i believe smile.gif, the sound isn't really great..but for highs i have all i need (hmm i don't think those tweeters can reach 21khz anyway smile.gif their specs say 20khz, so probably somewhat higher but not 21, anyway who cares smile.gif such sharpness blows your mind..)

edit: hey damn this post has a lot of smilies! biggrin.gif biggrin.gif

This post has been edited by SK1: Dec 16 2002, 17:55
Go to the top of the page
+Quote Post
caligae
post Dec 16 2002, 19:48
Post #10





Group: Members
Posts: 186
Joined: 23-January 02
Member No.: 1132



QUOTE (KikeG @ Dec 16 2002 - 05:04 PM)
Also, I hope that aliases caused by poor resampling in some soundcards are less audible than with pure tones, because it is more probable they get masked.

I was fooled by my soundcard some time ago using waveforms. I made a 20 second sinus going from 20Hz up to 22000Hz. Listened to it and i could hear the whole range without any problems.

After i upgraded my soundcard i noticed that those very high tones i heard were not the sound it was supposed to be at all but the soundcard making werid noises.

With this sample this issue might not be as bad since there is some music included beside the high frequency tone.
Go to the top of the page
+Quote Post
SometimesWarrior
post Dec 17 2002, 00:15
Post #11





Group: Members
Posts: 671
Joined: 21-November 01
From: California, US
Member No.: 514



QUOTE (Pio2001 @ Dec 16 2002 - 08:34 AM)
Same as with the sweep tone, resampling in most soundcards will get any file ABX able even at 21 kHz lowpass, because it will change the low frequency content. Use non resampling soundcards or burn CDs.

Shouldn't the audio files made available for downloading be resampled to 48kHz (for owners of resampling soundcards)? Otherwise the test isn't about hearing high frequencies in real music, it's about hearing soundcard aliasing in real music.
Go to the top of the page
+Quote Post
Pio2001
post Dec 17 2002, 00:46
Post #12


Moderator


Group: Super Moderator
Posts: 3936
Joined: 29-September 01
Member No.: 73



The problem might be the same at 48 kHz. Someone with an SB live should test it.
But two traps must be avoided :
1) Resampling the sweep from 44.1 kHz to 48 kHz can itself introduce aliasing, even with antialias.
2) Sweep generators in sound editors can themselves generate aliases

Performing a spectrum analysis on the 48 kHz sweep before testing is necessary to ensure that the sweep is pure.
Go to the top of the page
+Quote Post
rehgf
post Dec 17 2002, 01:02
Post #13





Group: Members
Posts: 51
Joined: 21-April 02
From: Finland
Member No.: 1836



One of those waiting for a (good) 48 kHz version is me. I hear above 20 kHz but can't trust my SB Live soundcard with 44.1 kHz samples.
Go to the top of the page
+Quote Post
Mac
post Dec 17 2002, 01:28
Post #14





Group: Members
Posts: 650
Joined: 28-July 02
From: B'ham UK
Member No.: 2828



Ugh, the polling software confused me! I click on view results to see how other people did, and now it won't let me vote, I cast a null one. Move to phpBB!!!!

Could one of the moderators vote for me please? The highest I could hear was 17kHz.


Great idea for a test btw, I really like the idea of resampling to gradually go higher or lower smile.gif


--------------------
< w o g o n e . c o m / l o l >
Go to the top of the page
+Quote Post
mithrandir
post Dec 17 2002, 02:54
Post #15





Group: Members
Posts: 669
Joined: 15-January 02
From: SE Pennsylvania
Member No.: 1032



The 17KHz lowpass file sounds the same as the "original". The 16KHz samples sound different but it takes some concentration.
Go to the top of the page
+Quote Post
JohnV
post Dec 17 2002, 03:52
Post #16





Group: Developer
Posts: 2797
Joined: 22-September 01
Member No.: 6



QUOTE (Mac @ Dec 17 2002 - 02:28 AM)
Ugh, the polling software confused me!  I click on view results to see how other people did, and now it won't let me vote, I cast a null one.  Move to phpBB!!!!

You voted a null vote in order to see the results.. You can't first check the results, and then give your vote...that could lead to remarkable bias in some polls.
Null vote is a vote..

Invision Board is very much better than phpBB. wink.gif


--------------------
Juha Laaksonheimo
Go to the top of the page
+Quote Post
KikeG
post Dec 17 2002, 09:16
Post #17


WinABX developer


Group: Developer
Posts: 1578
Joined: 1-October 01
Member No.: 137



Uploading 48 KHz files, old files not available at the moment...

Edit: upload not immediate, ftp problems in this moment.

Edit: fixed, see my next post.

This post has been edited by KikeG: Dec 17 2002, 17:18
Go to the top of the page
+Quote Post
SometimesWarrior
post Dec 17 2002, 10:25
Post #18





Group: Members
Posts: 671
Joined: 21-November 01
From: California, US
Member No.: 514



QUOTE (JohnV @ Dec 16 2002 - 06:52 PM)
QUOTE (Mac @ Dec 17 2002 - 02:28 AM)
Ugh, the polling software confused me!  I click on view results to see how other people did, and now it won't let me vote, I cast a null one.  Move to phpBB!!!!

You voted a null vote in order to see the results.. You can't first check the results, and then give your vote...that could lead to remarkable bias in some polls.
Null vote is a vote..

Invision Board is very much better than phpBB. wink.gif

If you really want to "sneak a peek" at the results, log off and view the thread anonymously. My browser has a nasty habit of forgetting that I logged into HA my last visit (I screwed up the cookie settings), so I end up seeing the results every time before I vote, which pretty much nullifies the significance that the poll had before I voted. smile.gif

I'm glad that KikeG is making the effort to distinguish between high-frequency hearing with sine sweeps and with "real" music, because it's something I haven't really bothered testing extensively. Looking forward to the 48kHz samples... (as if I didn't know what SSRC was, but I don't have any spectrum analysis tools to make sure I don't screw something up)
Go to the top of the page
+Quote Post
daniel
post Dec 17 2002, 10:51
Post #19





Group: Members
Posts: 86
Joined: 16-January 02
Member No.: 1047



smile.gif Yesterday I couldnt distinguish 16khz lowpass. Heh.
Today: 16khz 16/18
17.5khz 4/4
Havent voted yet, becouse hearing improves with age (w00t)

This post has been edited by daniel: Dec 17 2002, 10:51
Go to the top of the page
+Quote Post
KikeG
post Dec 17 2002, 16:58
Post #20


WinABX developer


Group: Developer
Posts: 1578
Joined: 1-October 01
Member No.: 137



I've finished uploading the files resampled to 48 KHz. I've uploaded additional 21 KHz and 21.5 KHz files to test some people hearing limits.

The files are now in .ape (Monkey's Audio) format, because the compression for some of the files is much better that with FLAC.

About the resampling, it was performed too with CEP 1.2a, quality 120 and post-filtering. Possible aliases due to resampling are explained at the web page:

"Update: actual files were resampled again from 44.1 KHz 16 bit to 48 KHz 32 bit and then re-dithered to 16 bit using same procedure, in order to avoid any possible bad soundcard resampling. Added noise due to this last step is mimimal.

Note on resampling effects: frequency aliases that appear due to the resampling process appear from 19 KHz to 22 KHz and are caused always from a frequency 4 KHz below (15 KHz to 18 KHz). They have a max. amplitude of around -95 dB (at 22 KHz) down the aliased frequencies. All other aliases are equal or below quantization noise. So, this aliases can be considered inaudible and of no importance in this test. ".

About this aliases, I've seen from time to time cd tracks that have a much louder in comparison constant tone at 16 KHz and above, probably caused by CRT displays (TVs, computer monitors) at the recording studios, that are unnoticed when listening, they are only noticeable on FFTs.

Anyway, it would be interesting for the people who has taken the test, to repeat it again and see if their results differ much.

This post has been edited by KikeG: Dec 17 2002, 17:23
Go to the top of the page
+Quote Post
ff123
post Dec 17 2002, 17:10
Post #21


ABC/HR developer, ff123.net admin


Group: Developer (Donating)
Posts: 1396
Joined: 24-September 01
Member No.: 12



QUOTE (KikeG @ Dec 17 2002 - 07:58 AM)
I've finished uploading the files resampled to 48 KHz. I've uploaded additional 21 KHz and 21.5 KHz files to test some people hearing limits.

So far, I've only tried the 14 kHz lowpassed (48 kHz version), with 8/8 for ABX. It was surprisingly (at least I was surprised) easy to hear. I'll try others later.

My soundcard (M-Audio Audiophile 2496) plays at the native sample rate.

ff123

Edit: Ah, 16 kHz is beyond me.

This post has been edited by ff123: Dec 17 2002, 17:26
Go to the top of the page
+Quote Post
Q!
post Dec 17 2002, 18:13
Post #22





Group: Members
Posts: 98
Joined: 26-August 02
From: PL
Member No.: 3199



18kHz

Funny, I can abx the 18kHz lowpassed easily, but 18.5 sounds the same as the original.

This post has been edited by Q!: Dec 17 2002, 18:15


--------------------
http://qisgod.host.sk
Go to the top of the page
+Quote Post
Guest_SK1_*
post Dec 17 2002, 20:20
Post #23





Guests






Hi again.
Well, i tried and i tried smile.gif.
I can hear the difference between all the samples up to 20khz (including), but with 21 (and obviously 21.5) i can't distiguish the two at all...no difference arrrg! no matter how much i try. (maybe it's the tweeters, specs say 20khz, but i totally doubt it).
Go to the top of the page
+Quote Post
Mr. Superbad
post Dec 17 2002, 22:16
Post #24





Group: Members
Posts: 18
Joined: 27-November 02
Member No.: 3893



I had to focus on the area from about 1-2 seconds... there's two "dings" where I can tell a difference up to 20khz.

I'm trying 21khz now... but it's already much, much harder.


Edit:

I got 4/4 on the 20hz sample, then went back an hour later and got 10/10. I can usually breeze through the first couple trials but then have to take a slight break between each try after that, or else I can't tell a difference.

This post has been edited by Mr. Superbad: Dec 18 2002, 00:13
Go to the top of the page
+Quote Post
CRasch
post Dec 17 2002, 22:21
Post #25





Group: Members
Posts: 12
Joined: 6-November 02
From: California
Member No.: 3708



I did a hearing test and I could hear past the 20kHz range. But I can barely tell the diffrence when a using a lowpass filter between a 18, 19 and 20kHz for most music, but I can tell on the diffrence between the 17 and 18 kHz on most new wave and techno music.
Go to the top of the page
+Quote Post

3 Pages V   1 2 3 >
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: 1st September 2014 - 19:56