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 
Mac
post Dec 17 2002, 22:44
Post #26





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



I found the same. 17khz sounded clearly different, but 17.5khz they sounded identical.

Heh, this also showed me that my left tweeter doesn't go above 14khz! smile.gif How crappy!

(btw, i've voted now at 17khz - very sneeky) tongue.gif

This post has been edited by Mac: Dec 17 2002, 22:45


--------------------
< w o g o n e . c o m / l o l >
Go to the top of the page
+Quote Post
RIV@NVX
post Dec 18 2002, 22:59
Post #27





Group: Members
Posts: 574
Joined: 15-December 02
From: Rijeka, Croatia
Member No.: 4090



What means to ABX a file?
How to do it?
Go to the top of the page
+Quote Post
Guest_SK1_*
post Dec 18 2002, 23:01
Post #28





Guests






Use a program, like WinABX (by KikeG, starter of this poll smile.gif)
clickme
Go to the top of the page
+Quote Post
RIV@NVX
post Dec 18 2002, 23:16
Post #29





Group: Members
Posts: 574
Joined: 15-December 02
From: Rijeka, Croatia
Member No.: 4090



QUOTE (SK1 @ Dec 18 2002 - 02:01 PM)
Use a program, like WinABX (by KikeG, starter of this poll smile.gif)
clickme

OK, will try to hear tomorrow. Will report here, batman wink.gif
Go to the top of the page
+Quote Post
KikeG
post Dec 19 2002, 09:03
Post #30


WinABX developer


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



ABX is a method of testing audible differences, it's a type of DBT (double-blind test). See FAQ thread at the general section.

This post has been edited by KikeG: Dec 19 2002, 16:20
Go to the top of the page
+Quote Post
zerobyte
post Dec 19 2002, 14:49
Post #31





Group: Members
Posts: 23
Joined: 22-November 01
Member No.: 521



the problem is not hearing a 20k or 18k signal, it's differentiating between 18k and 20k, since mechanic resonators (like the human ear) become more broad-banded with higher frequencies.

00h
Go to the top of the page
+Quote Post
KikeG
post Dec 19 2002, 16:04
Post #32


WinABX developer


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



You're right, but many people can hear 18 KHz content and cannot hear 19 KHz content. Also, I think it is interesting to know how high I can hear, because simply I want to know, and because it has some obvious implications on lossy codecs tuning and use.

This post has been edited by KikeG: Dec 19 2002, 16:12
Go to the top of the page
+Quote Post
KikeG
post Dec 19 2002, 16:17
Post #33


WinABX developer


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



I'm seeing in the pool that some people can supposedly hear 19-20 KHz content, but so far only two of them have provided ABX results. Anyone of the others has some ABX results?
Go to the top of the page
+Quote Post
Guest_SK1_*
post Dec 19 2002, 16:56
Post #34





Guests






Hello. I've decided to provide the accurate results. Used KikeG's fine program WinABX. (had a hard time this time smile.gif.. maybe because i have a cold)

CODE
-------------------------------------
WinABX v0.22 (ABA) test report
12/19/2002 17:36:35

A file: E:\ABX\Chenoa_19KHz.wav
B file: E:\ABX\Chenoa_19KHz_lowpass.wav

17:37:05    1/1  p=33.3%
17:37:53    2/2  p=11.1%
17:38:19    3/3  p= 3.7%
17:38:43    4/4  p= 1.2%
17:39:14    5/5  p= 0.4%
17:39:42    6/6  p= 0.1%
-------------------------------------
WinABX v0.22 (ABA) test report
12/19/2002 17:45:06

A file: E:\ABX\Chenoa_20KHz.wav
B file: E:\ABX\Chenoa_20KHz_lowpass.wav

17:46:04    1/1  p=33.3%
17:46:44    2/2  p=11.1%
17:47:23    3/3  p= 3.7%
17:48:01    4/4  p= 1.2%
17:48:58    5/5  p= 0.4%
17:49:48    6/6  p= 0.1%
17:49:48  test finished
Go to the top of the page
+Quote Post
KikeG
post Dec 20 2002, 08:58
Post #35


WinABX developer


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



@SK1: my previous post was aimed to the other people that hasn't provided any ABX results, but good to have more results from you.
Go to the top of the page
+Quote Post
Guest_SK1_*
post Dec 20 2002, 14:05
Post #36





Guests






Yeah i thought so, but felt that anyway it would be more complete smile.gif. Thanks.
Go to the top of the page
+Quote Post
Bedeox
post Dec 21 2002, 13:42
Post #37





Group: Members
Posts: 246
Joined: 20-December 02
From: Quite quiet place in Poland
Member No.: 4181



Yep, ABXed this lately, and it seems I'm able
to hear music up to 18.5

18.5 Hz ABX 8/8
19 Hz ABX 5/8
19 Hz ABX (again) 11/16 (hell, I can SOMETIMES distinguish that...)


--------------------
I've changed only because of myself.
Remember, when you quote me, you're quoting AstralStorm.
(read: this account is dead)
Go to the top of the page
+Quote Post
RIV@NVX
post Dec 21 2002, 17:10
Post #38





Group: Members
Posts: 574
Joined: 15-December 02
From: Rijeka, Croatia
Member No.: 4090



I can hear 17-18 Khz (using ABX 0.23).
Will do more test when I get TerraTec DMX 6fire 24/96... will see if it changes something.

This post has been edited by RIV@NVX: Dec 21 2002, 17:12
Go to the top of the page
+Quote Post
TJA
post Mar 3 2003, 22:05
Post #39





Group: Members
Posts: 84
Joined: 27-February 03
From: OLD Europe
Member No.: 5225



Wow, iīm impressed!

Either that samples are better, more worse or my ears got better B)

The frequency test from pcabx, i could not hear no differences at all!

Here, it looks like this:

CODE
kHz   WinABX
12     10/10
14     10/10
16      9/10
17      5/5    (got lazy, because of WinABX - see below - and because it got hard already!)
17.5    5/5    (yeah! more easy this time - needed a gap, it seems)
18      -/-    (even more lazy)
18.5    9/10
19      9/10   *grin*
20      no way (but i feel that it COULD be possible ... need a better day for that - and relaxed ears)
21      no way
21.5    no way


So, a lowpass of 19 is too low for me (and so, --alt-preset standard!)
19.5 may be, 20 will be enough for me, with 20.5 having a bit more room - so, 20.5 was a good choice, so far smile.gif
Sad, that there is no 19.5 kHz sample sad.gif

Some comments:

That WinABX is, say a FINE programm wink.gif - but i will try another! WinABX is quite uncomfortable ...

I did not completely understand that test, i need to say.
Those samples got faster and faster ... why?!?! (probably just donīt understand enough about thos ethings B) )
I needed to concentrate on two peeks somewhere in the middle!

The basic-samples did not use a lowpass and the ones called *lowpass* where made with a lowpass at the given kHz-number, yes?


Edited to add this:

I tested the 20 kHz sample and noticed the following:

At my first try every time, i have quite a good score to get it right!
But second and subsequent trials i cannot even guess ...
I need to concentrate on that two peeks somewhere in the middle of the sample.
I very strongly have the feeling that my ears reduce sensitivity because of the OTHER sounds around that two peeks. They are to loud compared to that peeks ... does anybody have similar samples to try?


Edited again:

I tried castanets-060_lo18KHz.wav against castanets-060_lo22KHz.wav and had NO chance!
It seems, i need special samples for that - like Chenoa dry.gif

This post has been edited by TJA: Mar 4 2003, 00:51


--------------------
3.90.3 --alt-preset extreme -V0 --lowpass 20.5 -> yeah!
"extremist of extreme", johnV @ Sep 13 2002 - 02:01 PM ;-)
Go to the top of the page
+Quote Post
ger@co
post Mar 3 2003, 22:33
Post #40





Group: Members
Posts: 110
Joined: 20-January 02
From: Canada
Member No.: 1088



I'm impressed too. Impressed by the fact that you would ressurect a dead thread (December 21st, 2002). biggrin.gif

Later.


--------------------
"Did you just say he contacts you through a bird? Did I just hear you say that?" Sonny Valerio (Cliff Gorman). Ghost Dog: The Way of the Samurai.
Go to the top of the page
+Quote Post
TJA
post Mar 3 2003, 23:12
Post #41





Group: Members
Posts: 84
Joined: 27-February 03
From: OLD Europe
Member No.: 5225



Hi Ger@co,
to be honest, i donīt understand your concept of "old". smile.gif

This is a valid thread (or topic), not locked, not closed and not deleted. For me, and any other person reading this the first time, it is as good as new ...
Also, the last reply - from 12. Dec 2002 - is only three month away!
What, if the next reader will stumble upon it in 10 years? For him, it will be quite an interesting and valid thread to reply to!

As i think about that, it seems, you tracked the thread, did you? wink.gif
Or did you look every other day manually? ohmy.gif

Anyway, i would like to get 19.5 and 20.5 kHz samples sad.gif


--------------------
3.90.3 --alt-preset extreme -V0 --lowpass 20.5 -> yeah!
"extremist of extreme", johnV @ Sep 13 2002 - 02:01 PM ;-)
Go to the top of the page
+Quote Post
KikeG
post Mar 4 2003, 10:04
Post #42


WinABX developer


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



TJA, I think indeed this test is quite useful no matter how old is it.

About your comments, you seem to have quite good hearing. The fact that you can hear "higher" here as opposed to PCABX samples is because the special clips I used. They are also speeded up for this reason, read the thread and the page explanations and I think you will understand why.

Hearing high frequencies is in practice limited my ear masking, low frequencies mask easily higher frequencies, and since most of the musical content is on low and medium frequencies, ultra-high frequencies get usually masked. With the "Chenoa" (name of a spanish singer) samples, the high frequency content is relatively unusual (bot not so much with modern music) both in amplitude and in frequency of the peaks. Resampling (speeding up) the samples I move the frequency of the peak, up in the frequency scale.

Note also that this is a near-worst-case test, most non-speeded up music doesn't have such high peaks at isolated frequencies over 16-17 KHz.

QUOTE
Anyway, i would like to get 19.5 and 20.5 kHz samples


I think I have them, but I ran out of web space, I guess I could send them to you via PM.

About your tests, don't stress too much, take your time to do them, and rest when needed.

This post has been edited by KikeG: Mar 4 2003, 10:08
Go to the top of the page
+Quote Post
TJA
post Mar 4 2003, 18:53
Post #43





Group: Members
Posts: 84
Joined: 27-February 03
From: OLD Europe
Member No.: 5225



QUOTE (KikeG @ Mar 4 2003 - 11:04 AM)
QUOTE

Anyway, i would like to get 19.5 and 20.5 kHz samples


I think I have them, but I ran out of web space, I guess I could send them to you via PM.

Thanx for the explanation, Kikeg!

I would like to get that samples and have setup an ftp-mirror for such samples.

Here you could put your additional samples to:

anonymous ftp (user "ftp") to 131.220.120.101:/devel/ftp/incoming

Here i will store all samples:

anonymous ftp (user "ftp") to 131.220.120.101:/devel/ftp/pub

Currently, it looks like this:

wav/kikeg
wav/pcabx
wav/mixed

later there will also be wav/sqam (see below)

Iīm currently creating the same structure for Flac, WavPack and Monkeyīs Audio.
This will take some time to reach the ftp-server as i am at home and only can use DSL at 128kbps, but itīs moving already! tongue.gif

flac/kikeg
flac/sqam
flac/mixed

Comment: Flac is making problems for me ... it complans about some subchannels?!?! blink.gif

options: -P 4096 -b 4608 -m -l 12 -e -q 0 -r 0,6
piano1_1644_MASK06.wav: 94% complete, ratio=0,489piano1_1644_MASK06.wav: WARNING
: skipping unknown sub-chunk 'LIST'

wv/kikeg
wv/sqam
wv/mixed

ape/kikeg
ape/sqam
ape/mixed

Additionally, i plan to include the PCABX test-samples in the same structure as well (will take some time, as the directories made problems to my script - need to look at that):

flac/pcabx
wv/pcabx
ape/pcabx

I offer this place for any exchange of test-samples, if need is be! biggrin.gif

Spread the Word B)

Just put it into "/incoming" and i will move it to an appropriate location ...

Edited: ftp-mirror is set up as written above - only pcabx in compressed versions is missing

This post has been edited by TJA: Mar 4 2003, 20:49


--------------------
3.90.3 --alt-preset extreme -V0 --lowpass 20.5 -> yeah!
"extremist of extreme", johnV @ Sep 13 2002 - 02:01 PM ;-)
Go to the top of the page
+Quote Post
DonP
post Mar 4 2003, 21:32
Post #44





Group: Members (Donating)
Posts: 1471
Joined: 11-February 03
From: Vermont
Member No.: 4955



VERY interesting test! The first sample I tried (17.5) I didn't hear any difference between A and B until I concentrated on the bells.
After that, up to 18.5 I could guess pretty reliably by just playing X OR Y (like 10/12 reliable) and then playing the other
one to check.

By 21 khz I had to play X and Y multiple times to be sure, and occasionally check against A and B.

I didn't see how to get a text report from the program, so I just copied the scores off the screen..

12/12 right up to 21 kHz.

Particulars:
Age: 47
Phones: Sony V6
Soundcard: Intel I810 on motherboard

Even though I could detect the difference, I would be hard pressed to
say that I'd have reduced enjoyment listening to 17.5 khz limited music.
I would have to try with (good) speakers to see if imaging is affected,
which I don't percieve as strongly with phones.

Side note: 15 years or so ago (when my hearing was presumably better)
I was at an "audiophile" friend's house checking out his system (and our ears)
with a test tone CD. At 18 kHz I could only hear it in a very narrow angle
from the center axis of his tweeter, maybe +/- 5 degrees. Beware that effect
if you try this test with speakers.
Go to the top of the page
+Quote Post
KikeG
post Mar 5 2003, 12:08
Post #45


WinABX developer


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



QUOTE (TJA @ Mar 4 2003 - 06:53 PM)
Comment: Flac is making problems for me ... it complans about some subchannels?!?!  blink.gif

It's not sub-channels, but RIFF sub-chunks. A sub-chunk is part of a wave file that contains some type on information or comments about the data type. The warning can probably be ignored, just make sure that FLAC compression works ok.

About your ftp: nice! I'll pass my remaining samples when I have some time, I don't have access to them right now.
Go to the top of the page
+Quote Post
KikeG
post Mar 5 2003, 12:13
Post #46


WinABX developer


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



QUOTE (DonP @ Mar 4 2003 - 09:32 PM)
I didn't see how to get a text report from the program, so I just copied the scores off the screen..

There's a log file at the program location. However, no need to copy the whole log file, just final results, included p value, is fine.

QUOTE
12/12 right up to 21 kHz.

Particulars:
Age: 47
Phones: Sony V6
Soundcard: Intel I810 on motherboard


Interesting results... However, it's *really* strange that at 47 you are able to hear so easily 21 KHz signals... I'd try to repeat the test using some good quality sound card, instead of an embedded one. There's a possybility that your card is folding down into the lower audio spectrum those very high frequencies. One way of checking this somehow, is paying attention to the kind of differences heard. I believe that very high frequencies are not heard, but "felt" inside your head, the higher the frequency, the more this happens.

Edit: strictly talking, nobody can make sure that this is not happening with other peoples' setups, including amp and headphones/speakers. The better the audio setup, the less likely this is happening.

This post has been edited by KikeG: Mar 5 2003, 12:24
Go to the top of the page
+Quote Post
tigre
post Mar 5 2003, 12:46
Post #47


Moderator


Group: Members
Posts: 1434
Joined: 26-November 02
Member No.: 3890



QUOTE (KikeG @ Mar 5 2003 - 03:13 AM)
One way of checking this somehow, is paying attention to the kind of differences heard.

Another way is to create sine tones with changing pitch (e.g. sweeps) with some wave editor at the sampling rate you're going to use for abxing. Then listen to it through your soundcard. If you watch the spectral view while listening and you hear decreasing pitch at a point where frecquency increases in spectral view (or the other way round) you should change something, e.g. use other equipment, other drivers, even lowering system/wave volume could help in some cases. To improve security, you should burn the test sample to audio CD and listen to it on different (hopfully decent) CD-Players.

QUOTE
I believe that very high frequencies are not heard, but "felt" inside your head, the higher the frequency, the more this happens.


It's similar for me. Until 17.5kHz cutoff it sounds different to me (brightness of the "bing"s), at higher cutoffs I could abx (tried up to 18.5kHz so far) because it felt like there was some kind of increased presure in my ear causing some kind of dizzy feeling in my head. (No, I didn't listen at 130dB. wink.gif )


--------------------
Let's suppose that rain washes out a picnic. Who is feeling negative? The rain? Or YOU? What's causing the negative feeling? The rain or your reaction? - Anthony De Mello
Go to the top of the page
+Quote Post
KikeG
post Mar 5 2003, 12:51
Post #48


WinABX developer


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



QUOTE (tigre @ Mar 5 2003 - 12:46 PM)
Another way is to create sine tones...

Yes, this is a more objective way of checking it. The best method is to record the sweep output of the card with a different, good card, and analyze the recording.

This post has been edited by KikeG: Mar 5 2003, 12:52
Go to the top of the page
+Quote Post
DonP
post Mar 5 2003, 14:08
Post #49





Group: Members (Donating)
Posts: 1471
Joined: 11-February 03
From: Vermont
Member No.: 4955



QUOTE (KikeG @ Mar 5 2003 - 03:13 AM)
Interesting results... However, it's *really* strange that at 47 you are able to hear so easily 21 KHz signals...

I was surprised too. As to the quality of sound I was keyed on it was always that the lowpassed
bell rings sounded a little duller, with the difference getting more subtle as the cutoff went up..

20 and 21 required a lot of concentration. If there were a fold-down or mixing product providing a
"cheat", then I would expect the spurious tone would get lower as the frequency went up, and
distinguishing the samples would get easier or remain the same. If the spurious low frequency
is only in the unfiltered wave, then would that still be the one to sound brighter?

Is it possible that the low pass filtering induced a temporal shift near the cutoff frequency?
That wouldn't show up in a spectrum analysis of the filtered wave.

Since the CD player isn't so good for A/B switching, maybe I can get the kids to
put together a random (but write it down!) sequence of the samples in 1 wave file
and then burn it. then I can try to determine the sequence by listening.

For what it's worth, in high school there was a 30 khz transducer in the physics lab. I couldn't
hear it as a tone, but I could always tell when it was on. I've always been the type to take
ear plugs to rock concerts.
Go to the top of the page
+Quote Post
KikeG
post Mar 5 2003, 14:41
Post #50


WinABX developer


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



QUOTE (DonP @ Mar 5 2003 - 02:08 PM)
Is it possible that the low pass filtering induced a temporal shift near the cutoff frequency?
That wouldn't show up in a spectrum analysis of the filtered wave.

According to SpectraLab FFT magnitude & phase measurements, frequency response mismatch is below 0.1 dB up to 20900 Hz, phase response mismatch below 1 degree up to 21100 Hz.
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: 23rd July 2014 - 10:31