Can you hear 16 times the jitter?

[kumakuma]

All of the files provided have been modified to some extent, as they are excerpts from the original track.

As Dennis is worried about clear violations of copyright, perhaps he could have provided links to the COMPLETE tracks on request for personal use ONLY, without naming the files directly. .

 

Exactly!

 

The files have been modified (data removed) and this may explain why some folks hear a difference. It has nothing to do with how electrically quiet or noisy the computer was when the file was saved.

[sandyk]

Exactly!

 

The files have been modified (data removed) and this may explain why some folks hear a difference. It has nothing to do with how electrically quiet or noisy the computer was when the file was saved.

 

There was no actual MUSIC data removed during those conversions, just trimming of the length of the lead in and lead out, either before or after the actual start and finish of the music itself . Look at many recordings in a sound editor and you will see dead silence until shortly before the music starts.

It's like talking to a brick wall, so I will ignore further posts from you that are directed to me in this thread.

[esldude]

I love it! Now we have a case where MEASUREMENTS indicate real differences that most of us can barely hear. And the same people are hearing significant differences between files that measure as "same."

 

Dennis: Were you intentionally stirring up this much controversy? It sounds as though you yourself are a bit confused by these results...

 

Well it is a bit more interesting than the exact same old arguments is it not?

[esldude]

Well I am saying that the simple act of saving an unmodified track to a different storage location as a COPY of the retained original does degrade it's sound ,the actual amount depending on how electrically quiet the PC is, as well as the actual storage location, especially if it's on USB, and I don't give a damn whether you believe me or not.

All of the files provided have been modified to some extent, as they are excerpts from the original track.

As Dennis is worried about clear violations of copyright, perhaps he could have provided links to the COMPLETE tracks on request for personal use ONLY, without naming the files directly.

 

I will not be entering into further discussions with you on this subject.

 

I believe I managed all manipulation with only 4 copy/edit steps. Not because I was worried about degradation of digital copies, but it was an efficient workflow.

 

So Alex how do you square the idea that differences from those copies of digital files is greater than 8 generations of copies through the analog world via DAC/ADC? Each of those had the original 4 copies plus 8 more copies plus analog world degrading. I don't see how that squares up with anything.

[kumakuma]

Deleted

[esldude]

Well I am saying that the simple act of saving an unmodified track to a different storage location as a COPY of the retained original does degrade it's sound ,the actual amount depending on how electrically quiet the PC is, as well as the actual storage location, especially if it's on USB, and I don't give a damn whether you believe me or not.

All of the files provided have been modified to some extent, as they are excerpts from the original track.

As Dennis is worried about clear violations of copyright, perhaps he could have provided links to the COMPLETE tracks on request for personal use ONLY, without naming the files directly.

 

I will not be entering into further discussions with you on this subject.

 

Well I don't think I could do that legally. You might manage it without detection, but in the USA if found out that would be considered a copyright violation. So I am not going there. Otherwise though it would take time I would have happily made entire albums available.

 

Then again the people who actually took part in these listening comparisons was tiny. Some citing not enough time. I doubt downloading hours of music to listen to would have improved participation levels.

[sandyk]

I believe I managed all manipulation with only 4 copy/edit steps. Not because I was worried about degradation of digital copies, but it was an efficient workflow.

 

So Alex how do you square the idea that differences from those copies of digital files is greater than 8 generations of copies through the analog world via DAC/ADC? Each of those had the original 4 copies plus 8 more copies plus analog world degrading. I don't see how that squares up with anything.

 

I decided not to participate after seeing what I considered were already degraded files due to the processing involved, so I can't comment on what others have reported.

It would have been more interesting to have provided original length tracks, with the first a directly ripped copy, and the "copy" simply having been trimmed in the lead in before the actual music started.

Keeping the tracks to distribution via PM request only, without publicly mentioning the name of the artist etc. would have got around your copyright concerns as well as providing a superior listening situation.

I agree also with what another poster said about suitable musical content to provide cues that many of us listen for.

[sandyk]

Well I don't think I could do that legally. You might manage it without detection, but in the USA if found out that would be considered a copyright violation. So I am not going there. Otherwise though it would take time I would have happily made entire albums available.

 

Then again the people who actually took part in these listening comparisons was tiny. Some citing not enough time. I doubt downloading hours of music to listen to would have improved participation levels.

 

Are you 100% certain that 30 seconds from a track is considered fully legal ?

Besides which, providing a single full length track for personal evaluation ONLY may even lead to increased sales .

This has certainly been my experience, with several members, including myself, purchasing the complete album after hearing the sample track.

[JohnSwenson]

I finally had a little time to listen to these files today.

 

First impression is that it is very difficult to tell any difference between them.The Hyperion Knight file Z has an artifact right at the beginning (a soft high pitched tone) that is not there at the beginning of the other two so I voted for that one as the re(re-re)-recorded file.

 

After reading everything else it might be an artifact of the different cut points, but I think it is longer than the number of samples listed.

 

John S.

[wgscott]

I finally had a little time to listen to these files today.

 

First impression is that it is very difficult to tell any difference between them.The Hyperion Knight file Z has an artifact right at the beginning (a soft high pitched tone) that is not there at the beginning of the other two so I voted for that one as the re(re-re)-recorded file.

 

After reading everything else it might be an artifact of the different cut points, but I think it is longer than the number of samples listed.

 

John S.

 

I just picked the same file for the same reason.

[STC]

Without the XYZ files, could you really identify which one got 8x highly jittered copy? To my ears, all the tracks sounded good enough. Perhaps, esldude should just make 10 different tracks and ask us to guess which got the highest jitter.

[esldude]

Without the XYZ files, could you really identify which one got 8x highly jittered copy? To my ears, all the tracks sounded good enough. Perhaps, esldude should just make 10 different tracks and ask us to guess which got the highest jitter.

Well there is the other poll with the up sampled original and three generations of copies. That one got only one voter I think. So I thought two originals and one odd file out might be easier.

 

The main thing I thought was asking people if they could hear an 8 th gen copy a large majority would have thought it easy. Yet when having the chance to hear it I believe many people were surprised at how close this sounded to the original.

[STC]

Well there is the other poll with the up sampled original and three generations of copies. That one got only one voter I think. So I thought two originals and one odd file out might be easier.

 

The main thing I thought was asking people if they could hear an 8 th gen copy a large majority would have thought it easy. Yet when having the chance to hear it I believe many people were surprised at how close this sounded to the original.

 

Could you please provide me a link? Anyway to be fair to others who could hear the difference, I have not listened to the files using my main system which is supposedly better than the Sennheiser earphone that I used but I doubt it would make any difference. I will give it try this weekend.

[sdolezalek]

The main thing I thought was asking people if they could hear an 8 th gen copy a large majority would have thought it easy. Yet when having the chance to hear it I believe many people were surprised at how close this sounded to the original.

 

I think most voters are trying too hard to find the "odd man out," i.e. which sounds most different rather than really listening to the quality of the sound. Try listening to the decay between notes and whether you ever really hear silence or whether there is always some level of sound. I find this easier to hear on the Hyperion Knight recordings, but I get to the same result on both recordings.

 

Now, that may not actually be hearing jitter, it may just be that the noise floor gets multiplied through that many recordings and therefore you never really get to that point where there is zero energy in your headphone or speaker drivers.

[esldude]

Could you please provide me a link? Anyway to be fair to others who could hear the difference, I have not listened to the files using my main system which is supposedly better than the Sennheiser earphone that I used but I doubt it would make any difference. I will give it try this weekend.

I bumped it to get some more views and hopefully voters. So you should see the generations poll and another where I have the original files and a smartphone recording. Surely people can hear a smartphone.

[PeterSt]

Especially for Mani :

 

Fiona Apple Y1.wav

Fiona Apple Y2.wav

 

(rightclick - Save As to save)

 

Please describe the difference, if any.

 

Mani is fed-up with this (or actually your responses), but did the test and gave his response by email (no, I did not ask his permission to copy his text and no, we did not talk a single word about this thread in private) :

 

Y1 and Y2 sound the same to me. But they sound slightly different to Y – slightly more ‘echoey’ but with the same ‘tone’. To my ears, X still sounds different to all of them – a lighter ‘tone’ and slightly more HF edge to the sound.

 

Y1 and Y2 are exactly the same. That is, Y1 is the trimmed version of X and Y2 is the trimmed version of Y (both front and end trimmed).

 

What I sneaked in here, was testing Mani's own dignity or say good skill of listening. But, this is combined with Alex's ideas about saving files from editors and what not. So this debunks a couple of things :

 

1. Mani hears perfectly right when he says that Y1 and Y2 sounds the same, because they *are* the same.

2. Mani is allowed to hear differences between X and Y because they are different.

3. Alex does not seem to have any base for his arguments, this time.

 

Ad 2.

Yesterday, knowing about Z, I refused to remember or even read well what Mani's verdict was. I took the original test after all. So it is to keep in mind that I did not know about any verdicts between X and Y and I had no clue about perceivable differences myself, knowing that they were the same, once the front and end was trimmed (which I did not do at the time, but I overlayed them and could see it).

 

All relative to each other and supposed you play from X via Y to Z :

 

X. Sounds nice. Nothing annoys really.

Y. Is firmer; has more slam. This is especially audible at the beginning. I couldn't avoid it once it was in my head.

Z. Is more sss like in the highs. The glass (?) in second 8 seems to have a lower pitch.

 

To me Z was the worse, but possibly because I knew. However, seeing what has happened to the file (in comparison to X/Y) it looks like a miracle that it resembles the original. No way I would have expected it to be that good, and probably when I would have listened to this as just a track on a Thursday night, I wouldn't have thought/said that something was wrong with it.

 

For me it is true that the difference between X and Y is larger than X and Z. But this is because of that profound more "slam" in Y. So while Z seems more tinny because of natural reasons (it is molested), X is also more tinny, but most probably perceived like that because Y is the outsider (it seems that the more slam is not what was intended).

 

When I later made the Y1 and Y2 as described, I could not perceive a difference between the two, but kept on finding that they both sounded different than all of the others and Y remained to have that slam (it would therefore be my choice as the best one).

 

Keep in mind, Mani and I did not spend a word about this in private; I didn't even respond to his email ...

 

Now it is all up to you to decide for who is crazy or deaf or hoping for outcome against all odds.

Audio remains to be interesting.

 

Peter

[sandyk]

I have a couple of questions for Dennis

 

1. Did you separately rip both "identical" versions ?

If the 2nd one is a copy of the original rip, then it's another processing stage/ pass through the PC.

 

2.Did you trim both copies ?

If so, why? Trimming just the 2nd one would be a more logical move, giving you non identical check sums .

If the 2nd one was copied from the original one , and only the derived copy was trimmed, then the "copy" has been through 2 additional processing stages which exposed it to further RF/EMI and the likelihood of audible degradation.

Alex

[goldsdad]

I finally had a little time to listen to these files today.

 

First impression is that it is very difficult to tell any difference between them.The Hyperion Knight file Z has an artifact right at the beginning (a soft high pitched tone) that is not there at the beginning of the other two so I voted for that one as the re(re-re)-recorded file.

 

After reading everything else it might be an artifact of the different cut points, but I think it is longer than the number of samples listed.

 

John S.

 

 

It is just a result of different cut points. The offsets in the HK set are different from the previously mentioned offsets in the FA set.

 

File Z of the HK set is offset from the X and Y (which are offset from each other by only about 100 samples) by starting 0.4 seconds earlier in the recording, and that period contains the clearly audible tone.

[goldsdad]

I have a couple of questions for Dennis

 

1. Did you separately rip both "identical" versions ?

If the 2nd one is a copy of the original rip, then it's another processing stage/ pass through the PC.

 

2.Did you trim both copies ?

If so, why? Trimming just the 2nd one would be a more logical move, giving you non identical check sums .

If the 2nd one was copied from the original one , and only the derived copy was trimmed, then the "copy" has been through 2 additional processing stages which exposed it to further RF/EMI and the likelihood of audible degradation.

Alex

 

Grasping at straws, Alex?

[esldude]

I have a couple of questions for Dennis

 

1. Did you separately rip both "identical" versions ?

If the 2nd one is a copy of the original rip, then it's another processing stage/ pass through the PC.

 

2.Did you trim both copies ?

If so, why? Trimming just the 2nd one would be a more logical move, giving you non identical check sums .

If the 2nd one was copied from the original one , and only the derived copy was trimmed, then the "copy" has been through 2 additional processing stages which exposed it to further RF/EMI and the likelihood of audible degradation.

Alex

I opened the original 30 second segment into an editor. At that point it included 20 seconds of a 440 hz tone used to set recording levels with 3 seconds of silence before the music. 3 seconds of siIence after and then the Knight track. I duplicated the track on the editing software. I selected and deleted the tone off both tracks at the same time. Deleted the Knight track from both. Then trimmed the ends. Then exported the two tracks naming each with different names.

 

With your concerns it would appear there's no way for any two tracks to ever sound the same no matter what your workflow is. I don't believe that is an issue. Nor have I ever heard an issue from such causes.

[sandyk]

Grasping at straws, Alex?

 

No. I was able to hear clear differences between corrected copies of 24/192 DVDs from Barry Diament where the header info was removed to make it playable with both cPlay , and my Oppo 103. Sound Forge 9 stated it had incorrect (too much) header information, and then gave me the option to save it again without the header info ,which then made it playable in the Oppo 103.

Peter (the surfing alien) wrote for me a special little Command Line program that did the same, however Peter's software corrected version clearly sounded better than the SF9 version despite both having identical check sums.

Apparently the original problem arose with the software used to convert the original .aiff files to .wav files.

[esldude]

Mani is fed-up with this (or actually your responses), but did the test and gave his response by email (no, I did not ask his permission to copy his text and no, we did not talk a single word about this thread in private) :

 

Y1 and Y2 sound the same to me. But they sound slightly different to Y – slightly more ‘echoey’ but with the same ‘tone’. To my ears, X still sounds different to all of them – a lighter ‘tone’ and slightly more HF edge to the sound.

 

Y1 and Y2 are exactly the same. That is, Y1 is the trimmed version of X and Y2 is the trimmed version of Y (both front and end trimmed).

 

What I sneaked in here, was testing Mani's own dignity or say good skill of listening. But, this is combined with Alex's ideas about saving files from editors and what not. So this debunks a couple of things :

 

1. Mani hears perfectly right when he says that Y1 and Y2 sounds the same, because they *are* the same.

2. Mani is allowed to hear differences between X and Y because they are different.

3. Alex does not seem to have any base for his arguments, this time.

 

Ad 2.

Yesterday, knowing about Z, I refused to remember or even read well what Mani's verdict was. I took the original test after all. So it is to keep in mind that I did not know about any verdicts between X and Y and I had no clue about perceivable differences myself, knowing that they were the same, once the front and end was trimmed (which I did not do at the time, but I overlayed them and could see it).

 

All relative to each other and supposed you play from X via Y to Z :

 

X. Sounds nice. Nothing annoys really.

Y. Is firmer; has more slam. This is especially audible at the beginning. I couldn't avoid it once it was in my head.

Z. Is more sss like in the highs. The glass (?) in second 8 seems to have a lower pitch.

 

To me Z was the worse, but possibly because I knew. However, seeing what has happened to the file (in comparison to X/Y) it looks like a miracle that it resembles the original. No way I would have expected it to be that good, and probably when I would have listened to this as just a track on a Thursday night, I wouldn't have thought/said that something was wrong with it.

 

For me it is true that the difference between X and Y is larger than X and Z. But this is because of that profound more "slam" in Y. So while Z seems more tinny because of natural reasons (it is molested), X is also more tinny, but most probably perceived like that because Y is the outsider (it seems that the more slam is not what was intended).

 

When I later made the Y1 and Y2 as described, I could not perceive a difference between the two, but kept on finding that they both sounded different than all of the others and Y remained to have that slam (it would therefore be my choice as the best one).

 

Keep in mind, Mani and I did not spend a word about this in private; I didn't even respond to his email ...

 

Now it is all up to you to decide for who is crazy or deaf or hoping for outcome against all odds.

Audio remains to be interesting.

 

Peter

Well I am sorry if Mani is upset with my responses. I don't hear what you two are describing. I find the explanations inadequate to explain a larger difference between trimmed identical files than either vs an 8 th gen copy. So I just can't get there from here in my head or my ears.

[sandyk]

FWIW, if you search hack through the forum you should also find reports regarding a similar weird phenomena from ex member Steve Nugent from Empirical Audio.

Clay (losing my religion) put me in touch with him about it quite a few years ago, but I have forgotten the details..

 

I just found it again.

 

http://www.computeraudiophile.com/f8-general-forum/file-format-2861/index2.html see post 33.

[esldude]

I don't know the specifics of how they listened. Let me give a gross example of what causes the same thing. You have two files playing concurrently with the ability to instantly switch as in say foobar abx. There is a one second offset however. You have a tell for which is which. Your mind will often then hear them as different even if they aren't. It will latch onto this or that aspect as different.

 

I have experienced this at much smaller fractions of a second. I don't know if 100 or 200 samples offset is enough, but maybe. When I have done this it was listening to portions of a larger sample. When I straightened up the timing without actually changing the file at all those 'differences' disappeared.

[YashN]

The main thing I thought was asking people if they could hear an 8 th gen copy a large majority would have thought it easy. Yet when having the chance to hear it I believe many people were surprised at how close this sounded to the original.

 

Was (were) the control file or files tampered with?

 

One mediocre piece of music is far from enough to be able to reach conclusions, at least for me.

 

I was thinking of making a DB of test music tracks for listening impressions after tweaks or equipment changes, and now I think that same thing can be used for such tests, which would provide a larger basis for comparisons.

 

I know PeterSt uses several tracks for different characteristics to listen for over several days.