Jitter Listening test files

Topic: Jitter Listening test files (Read 24016 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.

Jitter Listening test files

Reply #25 – 2014-12-10 16:27:36

Here is my 0.00625 result and i have to admit it is a pretty big step from 0.0125 but overall it sounds a bit muddy on the guitar strings. I guess the lowest jitter one i have to pass.
HD-590. old X-Fi

foo_abx 2.0 beta 4 report
foobar2000 v1.3.4
2014-12-10 17:18:28

File A: 30 Hz jitter marginal level .00625.flac
SHA1: ee1083c40daa4ad36c07081162f2b3326bfe447c
File B: no jitter.flac
SHA1: 262cd6c4d4c73502a0142f867b00aae013fd13ce

Output:
DS : Primärer Soundtreiber

17:18:28 : Test started.
17:19:49 : 01/01
17:20:18 : 02/02
17:20:43 : 03/03
17:20:56 : 04/04
17:21:11 : 05/05
17:21:29 : 06/06
17:21:53 : 07/07
17:22:11 : 07/08
17:22:32 : 08/09
17:22:47 : 09/10
17:22:47 : Test finished.

----------
Total: 9/10
Probability that you were guessing: 1.1%

-- signature --
8594ec05b059f9ae0e4edf3a3173bdbd8665a65e

Jitter Listening test files

Reply #26 – 2014-12-10 16:44:18

I tried ABX training with that yesterday and was able to get sets of 5 in a row but only with in-between breaks. I found increasing the volume level made it more difficult. I was using IEMs as well, so microphonics and occlusion were not helping.

Jitter Listening test files

Reply #27 – 2014-12-10 16:54:15

Yesterday night i didn't really try lower as 0.0125. This log above was my first try today after some relaxed sofatime. I was surprised myself because the beta foobar plugin shows the result only at the end.

Jitter Listening test files

Reply #28 – 2014-12-10 16:57:18

Even under the best circumstances, I'm pretty confident I wouldn't be able to go much lower than 0.00625.

Jitter Listening test files

Reply #29 – 2014-12-10 17:13:01

Quote from: amirm on 2014-12-09 14:26:52

A receiver needs to lock onto sampling rates from 32 Khz to 192 Khz. If it goes too slow, when you switch sources, you will have to wait many seconds for it to find and lock onto the incoming sample rate. Customers will not accept input switching delays of more than 1 or 2 seconds whereas they are oblivious to jitter as a problem so fast switching wins.

That explains why "jitter" was never mentioned by hi-fi sellers back in the day when consumers only had one digital source (CD). Never ever!

Jitter Listening test files

Reply #30 – 2014-12-10 19:52:33

For fun:

Code: [Select]

foo_abx 2.0 beta 8 report
foobar2000 v1.3.1
2014-12-10 11:41:45

File A: 0 - no jitter.flac
SHA1: ffdad7236acf795cb14a3816abf8c3a0da6b5bf4
File B: 1 - 30 Hz threshold jitter 0.00312.flac
SHA1: 7445aff311ed89263769042c856c089e0f97534d

Output:
DS : Primary Sound Driver, 16-bit
Crossfading: NO

11:41:45 : Test started.
11:45:51 : 01/01
11:46:05 : 01/02
11:46:33 : 02/03
11:46:51 : 02/04
11:47:01 : 03/05
11:47:44 : 04/06
11:47:56 : 04/07
11:48:03 : 05/08
11:48:16 : 06/09
11:48:35 : 07/10
11:48:52 : 08/11
11:49:13 : 08/12
11:49:26 : 08/13
11:49:36 : 09/14
11:49:41 : 10/15
11:49:49 : 11/16
11:49:49 : Test finished.

 ---------- 
Total: 11/16
Probability that you were guessing: 10.5%

 -- signature -- 
5d06b1bdcf6f49cdca391cd1607dd7acb9fddc53

Jitter Listening test files

Reply #31 – 2014-12-10 20:32:37

Nice one greynol!

Is this really what jitter does to audio? Audiophiles report all kinds of weird degrations due to it.
Most of the time it is blamed to sound like unpleasant digitalis scratchy distortion.
With these samples the problem sounds absolutely the way i would blame something to be wrong with analog gear.

Jitter Listening test files

Reply #32 – 2014-12-10 20:45:45

Thanks. If I repeat the test and do slightly better, I'm pretty sure the two combined will have significance.

In response to your earlier posts I mistakenly confused the 0.00625 for the 0.003125 file.

...so let me amend what I said earlier:

Quote from: greynol on 2014-12-10 16:57:18

Even under the best circumstances, I'm ~~pretty~~ confident I wouldn't be able to go ~~much~~ any lower than ~~0.00625~~ 0.003125.

*

In terms of sound quality, I'm really just focusing on clarity in only very short segments of the guitar.

Most audiophiles are full of shit. That's why I call them placebophiles. They'll attribute all sorts techno-babble to the stuff they imagine hearing.

(*) I'm not all that convinced that the fidelity of this Joni Mitchel clip is really all that stupendous, however!

Jitter Listening test files

Reply #33 – 2014-12-10 21:43:08

I thought I understood what jitter was, but I really need someone to dumb this down for me. What do the 30 Hz and the 0.003125 to 0.1 figures represent?

Were the files produced by varying the timing of the samples, i.e. each sample was taken not precisely 1/44100th of a second apart? Or is it a simulation of the effects of that? Or something else?

Jitter Listening test files

Reply #34 – 2014-12-11 14:02:49

Quote from: mjb2006 on 2014-12-10 21:43:08

I thought I understood what jitter was, but I really need someone to dumb this down for me. What do the 30 Hz and the 0.003125 to 0.1 figures represent?

Jitter is FM distortion, which interestingly enough is usually far more of a problem in the analog domain than the digital domain. In the analog domain it is called flutter and wow and most of the ways to address it are mechanical. For example high quality analog tape recorders have scrape flutter filters which are massy mechanical rollers that are placed in contact with the tape and tend to stabilize its speed over the heads.

In the analog domain the leading causes of FM distortion are analog media such as the LP and tape both of which are pretty horrendous if you look at the comparable numbers. The adoption of digital media vastly reduced this problem.

Speakers can also be strong sources of analog FM distortion. Multi-way speakers can strongly reduce this problem. The adoption of subwoofers has probably reduced the incidence of FM distortion due to speakers to a great degree. Attempting to get strong deep bass out of small speaker drivers with just one or two frequency bands favors the creation of more FM distortion.

FM multipath can also introduce jitter. Jitter from this source can be very audible but is usually more sporadic in its effects and is generally far worse for mobile receivers.

Multipath can also cause jitter in digital TV which will effect both the audio and the video. Modern OTA digital TV receivers now contain special circuits to minimize it. It is probable that the classic Benjamin and Gannon AES paper about jitter was at least partially stimulated by concerns about audible jitter due to digital TV reception.

Quote

Were the files produced by varying the timing of the samples, i.e. each sample was taken not precisely 1/44100th of a second apart? Or is it a simulation of the effects of that? Or something else?

The samples I provide are produced in the digital domain by varying the timing of the data contained in the samples by means of complex DSP algorithms that I am not privy to. I know they work based on exhaustive technical tests. For example 30 Hz jitter introduces a cyclic change in the frequency of the recorded signal over the span of more than 100 samples. The clock rate or inter-sample timing was never actually changed.

Jitter Listening test files

Reply #35 – 2014-12-11 14:19:31

Quote from: Wombat on 2014-12-10 20:32:37

Is this really what jitter does to audio?

Jitter is the same as tremelo and thus has two orthogonal dimensions - frequency of modulation and magnitude or depth of modulation. This is 30 Hz jitter. Real world audio system jitter can vary in frequency from less than 1/3 Hz (LP rotation) to over 500 Hz, so this is just one kind of jitter of a great many.

I have prepared a few samples with very large amounts of jitter at 200 Hz and 500 Hz. You can download them from here:

Dropbox link to 200-500 Hz jitter files

I need to know whether other people can hear the jitter in these files well enough for them to be used as training files.

Quote

Audiophiles report all kinds of weird degrations due to it.

I can only echo Greynol's comments about audiophile=placebophile. My experience and the scientific literature suggest that most audiophiles are describing events that exist only in their imaginations when they talk about digital dither in good audio gear.

Quote

Most of the time it is blamed to sound like unpleasant digitalis scratchy distortion.

IME most so-called digitalis is actually due to poorly produced recordings and poor speakers and room acoustics. IOW it has nothing to do with the intherent properties of digital audio as is currently implemented.

Quote

With these samples the problem sounds absolutely the way i would blame something to be wrong with analog gear.

As I point out in another post I just made to this thread, problems in the analog domain are the sources of most audible jitter.

I find it ironic that jitter was not treated very seriously by audiophiles until it was mostly reduced by vast amounts with the introduction of the CD.

Jitter Listening test files

Reply #36 – 2014-12-11 14:37:57

I find this frequency of jitter most noticeable in the tone at the end, where it is fairly easy to spot. I didn't find that with the 30Hz jitter.

Jitter Listening test files

Reply #37 – 2014-12-11 14:40:16

Quote from: Arnold B. Krueger on 2014-12-11 14:19:31

Jitter is the same as tremelo and thus has two orthogonal dimensions - frequency of modulation and magnitude or depth of modulation.

I always thought tremolo was chopping/amplitude modulation, and vibrato was FM.

Jitter Listening test files

Reply #38 – 2014-12-11 15:24:39

Quote from: julf on 2014-12-11 14:40:16

Quote from: Arnold B. Krueger on 2014-12-11 14:19:31
Jitter is the same as tremelo and thus has two orthogonal dimensions - frequency of modulation and magnitude or depth of modulation.

I always thought tremolo was chopping/amplitude modulation, and vibrato was FM.

That is correct. Jitter is Vibrato or FM. Typo! Tremolo is AM.

Jitter Listening test files

Reply #39 – 2014-12-11 18:05:59

Now I'm confused about an electric guitar's "whammy" bar because it's sometimes called tremolo.

Jitter Listening test files

Reply #40 – 2014-12-11 18:09:53

The whammy bar alters the pitch, so the name tremolo is misapplied.

Jitter Listening test files

Reply #41 – 2014-12-11 18:40:16

When digital jitter-phobes say they need a pro/audiophile-grade DAC and interconnects because the jitter in consumer-grade gear makes everything sound worse than analog, is this FM/vibrato effect the thing they're worried about? (regardless of whether they actually hear it) ... Or does that kind of jitter, when taken to extremes like in these audio files, have a different sound?

Jitter Listening test files

Reply #42 – 2014-12-11 18:45:33

You'll have to ask them, but make sure you bring your waders.

Jitter Listening test files

Reply #43 – 2014-12-11 19:28:33

Quote from: Arnold B. Krueger on 2014-12-11 14:19:31

I have prepared a few samples with very large amounts of jitter at 200 Hz and 500 Hz. You can download them from here:

Dropbox link to 200-500 Hz jitter files

I need to know whether other people can hear the jitter in these files well enough for them to be used as training files.

I had trouble hearing the problem in the music in the 200 Hz version at first. The jitter was very audible in the ending tone though. After a little while the dirty sound was audible during the guitars too.
The 500 Hz version was audible in its entirety without effort.

Jitter Listening test files

Reply #44 – 2014-12-11 19:53:34

Quote from: Arnold B. Krueger on 2014-12-11 14:19:31

I have prepared a few samples with very large amounts of jitter at 200 Hz and 500 Hz. You can download them from here:

Dropbox link to 200-500 Hz jitter files

I need to know whether other people can hear the jitter in these files well enough for them to be used as training files.

Thanks Arny. Especialy the exaggerated 500Hz sample has indeed an ugly metallic signature to me just what i expect from jitter. The 200Hz sample is not annoying but also clearly audible.

Jitter Listening test files

Reply #45 – 2014-12-12 00:26:40

Quote from: Arnold B. Krueger on 2014-06-20 01:02:59

30 Hz max jitter 0.1.flac
30 Hz Severe Jitter 0.05.flac
30 Hz jitter strong level .025.flac
30 Hz noticable jitter 0.0125.flac
30 Hz jitter marginal level .00625.flac
30 Hz threshold jitter 0.00312.flac

No one to date has submitted a successful ABX report for the "threshold" file, but some have succeeded with the "marginal" file. I thought it worthwhile to report that the subjective disturbance (at least for my ears) does not follow a linear type of relationship along the lines of a doubling or halving of the effect when moving from one jitter level in the above files to the next. Rather there is a much bigger change. For example, in going from listening to the "noticeable" file to the "strong" file, the subjective impact is many times more pronounced for the "strong" file.

Quote from: Arnold B. Krueger on 2014-12-11 14:19:31

I have prepared a few samples with very large amounts of jitter at 200 Hz and 500 Hz. You can download them from here:

Dropbox link to 200-500 Hz jitter files

I need to know whether other people can hear the jitter in these files well enough for them to be used as training files.

I think the effect on the guitar would be obvious enough for people with good hearing, but not over-obvious, for use as training files.

I note that introducing a stable jitter at one frequency need not sound unpleasant when applied to some instruments (though this can be a question of personal preference). For my ears, the guitar sounds a bit like a harpsichord in these files. A more complex, and arguably more interesting, sound. The jitter doesn't seem to improve the vocal in any way however!

Quote from: greynol on 2014-12-10 20:45:45

I'm not all that convinced that the fidelity of this Joni Mitchel clip is really all that stupendous, however!

I agree.

I think the simplicity of plucked/strummed guitar strings, as we have in this Joni Mitchel clip, is good for exposing frequency variations. Another possibility would have been piano. I recall with domestic open-reel tape recorders, and cassette tape recorders, that flutter could often be heard readily with the sound of the piano.

Jitter Listening test files

Reply #46 – 2014-12-12 06:16:44

So I get bragging rights if I pass*?

If I have time, I'll do it.

(*) the downside is that the information will likely be misused either out of ignorance or to serve some agenda.

Jitter Listening test files

Reply #47 – 2014-12-14 02:37:40

How about make a signal that consists of equal parts of 18kHz and 19kHz signal and use that for jitter detection?

Jitter Listening test files

Reply #48 – 2014-12-14 12:40:57

Quote from: Woodinville on 2014-12-14 02:37:40

How about make a signal that consists of equal parts of 18kHz and 19kHz signal and use that for jitter detection?

Tried it.

18 KHz and 19 KHz are inaudible to most people including me.

Adding frequency modulation creates sidebands that are tightly clustered around 18 KHz and 19 KHz, and they are also inaudible to most people.

It would be nice to have a test signal that:

(1) Is audible to most people

(2) Is similar to a musical sound

Jitter Listening test files

Reply #49 – 2014-12-14 14:32:22

It would be nice to have test signals where the jitter:

(1) Mimics the modulation found in typical situations, including the current placebophile paranoia: HDMI jitter.

(2) Is at extremely exaggerated levels in order to be audible to most people.

The original content should contain music that that was optimally recorded and produced. If a synthesized signsl makes the jitter easier to detect, it should be included as well. It may need to be more complex than a single tone.