[infinitesymphony]

Quote:

Originally Posted by markl /img/forum/go_quote.gif You guys agree a digital bit is a digital bit, yes? He's transferred the *exact same information* twice (albeit from different sources), and has ended up with (guess what) two identical copies of the same data! How on earth could they sound different?

He swapped the power cords on both his DAC and PC; the DAC was outputting analog information and the PC was receiving it and converting it back to digital. You seem to be implying that he basically ripped the tracks from the CD, which wasn't the case. Also, the waveforms are not identical...

 

[nick_charles]

Quote:

Originally Posted by markl /img/forum/go_quote.gif You guys agree a digital bit is a digital bit, yes? He's transferred the *exact same information* twice (albeit from different sources), and has ended up with (guess what) two identical copies of the same data! How on earth could they sound different?

No, you are missing the point here. The point was to see if there was a diffrence in the analog outputs from the DAC using different cables, not the digital bistream.

The outputs from the DAC using two different cables were then recaptured to wavs, the wavs are proxies for the analog outputs.

 

[Sovkiller]

Quote:

Originally Posted by markl /img/forum/go_quote.gif If he wanted to do something scientifically valid that could "solve" the problem, he would take a statistically significant sample size of true, acknowledged "golden ears" (like an army of Michael Fremers and such), use *their* systems that they know like the back of the their hand, and then conduct your double blind test by swapping out *their* power cord of choice (the sound of which they know like the back of their hand), and swap it out with a cheap stock cord. Then report your results. That would be a good start. But this test?...

Do you have any link to that test to see the results as well...it will be really interesting...

Quote:

You guys agree a digital bit is a digital bit, yes? He's transferred the *exact same information* twice (albeit from different sources), and has ended up with (guess what) two identical copies of the same data! How on earth could they sound different?

Well not according to the cable adorers, they say that power cords make a difference (and regardless of where it is use) or sort of, for better or worst, right? That is what we have heard here, people lsitening diferences between two power cords in DACs, CD players, amps, (TVs, even irons an microwaves!!!

)

That is not the first time we heard of those right, then, why the hell if that can be heard, it could not be measured??? I do not get this logic, is your hearing better than any instrument? Sorry but I doubt that...otherwise we do not need them...

If there is a heard difference, this difference by all means could be, should be, and must be, (and if not now, in the future) measured and quantified in one way or another...audio this is not a faith, it is part of an exact science...

Why not leaving margin for other opinions as well, a wise advise, what about if in 20 years from now science prove you all wrong, with objective measurments, that thaere is no differnece, and make you all try and fail all imaginable tests? what are you going to say? Will you still insisit in that the earth is still flat and on top of 4 elephants!!!!

 

[nick_charles]

Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif He swapped the power cords on both his DAC and PC; the DAC was outputting analog information and the PC was receiving it and converting it back to digital. You seem to be implying that he basically ripped the tracks from the CD, which wasn't the case. Also, the waveforms are not identical...

Even though I was unable to get two perfectly aligned samples even my crude tests show quite definite (small) variations in the wave forms from two attempts.

 

[nick_charles]

Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif He swapped the power cords on both his DAC and PC; the DAC was outputting analog information and the PC was receiving it and converting it back to digital. You seem to be implying that he basically ripped the tracks from the CD, which wasn't the case. Also, the waveforms are not identical...

On both ?, that isnt so good. If that is true then there are two variables DAC cable and ADC cable, that makes it a poorer design. It is always best to just change one variable at a time before you start getting into interaction effects.

 

[bigshot]

Quote:

Originally Posted by nick_charles /img/forum/go_quote.gif Even though I was unable to get two perfectly aligned samples even my crude tests show quite definite (small) variations in the wave forms from two attempts.

You might want to capture a couple of times with the same setup to see if the difference in waveform is just capture variability.

See ya
Steve

 

[markl]

Quote:

The point was to see if there was a diffrence in the analog outputs from the DAC using different cables, not the digital bistream.

This must have been buried deep in his post, because he talks about "data throughput from each cable". I've scanned the post again, and I still don't see where he says he's comparing the analog output. He talks a lot about the poor quality of the stock cable but does not discuss his methodology in a an easily readible way.

I have no time, but even if he is, this is absolutely riddled with problems:

1. You are comparing a copy of a copy on a different machine, you are evaluating the perrformance of *your* set-up, not the original. And what about the original set-up? Is it so hot to begin with? There will be degradation of the signal made during the copying process. How good is the source of the transfer, how good is the recipient of the transfer? If you are taking a picture of a picture, if your lens is not absolutely *perfect* (and what is), you will have artifacts, blurring, that extra layer of grain you get from a print of a print, etc. What DAC is he using to make the transfer or is it a computer? Is that buried in there somewhere? Does this source even reveal differences in things like cables in the first place? What interconnects were used to make the transfer? Are they also of low quality, further degrading sound?

2. Now, subjects are listening to the copy of a copy on their *computers*, which are poor sources compared to the high-end stand-alone machines. Are they using cheap computer speakers to try to hear the differences? I can barely use my stock Dell speakers to listen to Youtube clips.

3. Who are these subjects? Audiophiles or other skeptics who can't hear cable differences to begin with?

4. The test actually did reveal some (small) variances between the two samples (oops, how embarrassing). But there are some saying that this could be due to simple variances that occur with the machinery from copy to copy. OK, so it is acknowledged you are using flawed gear that can't consistently provide the same result to measure something as small as cable differences. This is a big problem with this test.

This "test" again makes the same mistake that the skeptics continue to make WRT to cables-- these differences are *small*. Now you are taking a recording of the original event to evaluate an original, possibly using crummy eqipoment to do it with. How do these differences even have a chance of showing up?


Rats, out of time, will play more tomorrow.

 

[unoab]

well if we really want to kick this study into high gear, I think we should bust out a true analysis program like matlab.

Use your soundcard to play a sound file out and loop it back into the computer, now import both sound files into matlab (ideally this would be done IN matlab using device drivers with synchronus playback/recording). Do a fourier transform and divide them to find out the transfer function of the computer+soundcard, now we are ready to test power cables.

Use the soundcard out to play a test signal, like you did before, through any number of components using whatever power cord you want. Record the output back into the computer, do another fourier transform in matlab, and divide by the transfer function of the computer you determined before. Now you will be left with a sound clip that is just affected by the particular component you ran the sound through, and the power cord it is using (well and the interconnects, but you can daisy chain them in the first step so they are included in the response of the computer).

From here we could either audition these sound clips to see if there is a difference. Or the other option is to go back into matlab, go back into the fourier domain and divide the newly created output file by the original test file that we played. We are now left with the response of the component it was passed trough and the effects the power cable has. Both of these responses can be plotted and compared since the differences will be purely from the power cable.

This is probably the best way to go with this experiment, completely objective, and all parts of the chain are taken into account and are removed from the final two files. But there will always be those that say something something is wrong....

Also, if you think a power cord improved your system are you not saying something greater about the power supply built into your components? I find esoteric powercords an insult to power supply engineers...

 

[Sovkiller]

May I ask which more credible physical methodology is under the saying I hear this or that, and what makes it more valid than this experiment? Do you know how easy is to fool your perceptions, even more while they are sometimes a month or two appart?

Oh yes, I forgot, there is one, I take out this cable and put the other one in, but that is the same for both cases...

Quote:

Originally Posted by bigshot /img/forum/go_quote.gif You might want to capture a couple of times with the same setup to see if the difference in waveform is just capture variability. See ya Steve

That is the first thing to determine, as I stated before, if the system at least is accurate with the same track, if not we are wasting out time since square one...

 

[nick_charles]

Quote:

Originally Posted by bigshot /img/forum/go_quote.gif You might want to capture a couple of times with the same setup to see if the difference in waveform is just capture variability. See ya Steve

It was - 3 attempts all three slightly different, we are talking very very small variations and not visible until you zoom in to absurd levels i.e 0.0001s or even 0.00005s. would this swamp variations in cables - dunno ?

 

[JohnFerrier]

I agree with markl, except that it doesn't take a team of Michael Fremers, it takes only one. Akin to what Hirsch posted in 2004: Quote:

Originally Posted by Hirsch /img/forum/go_quote.gif If one, and only one, person can be shown to be able to reliably detect cable differences in a properly controlled setting, meeting stringent statistical criteria, then the question is answered. The differences exist.

However, it seems to me, that people need to think more clearly about human perceptions. Neural networks interconnect informaton. The brain, "the most complex thing in the known universe", is "massively parallel." How do you seperate the interconnections? If you don't fine, but then I don't think we are talking about "the real thing".

(Regarding the "test" in this tread. I don't know. If the cables measure difference resistances, do you want this test to reveal differences in resistance? If it can't, then is the test equipment resolving enough?)


.

 

[unoab]

Quote:

Originally Posted by JohnFerrier /img/forum/go_quote.gif (Regarding the "test" in this tread. I don't know. If the cables measure difference resistances, do you want this test to reveal differences in resistance? If it can't, then is the test equipment resolving enough?)

Different resistances? I can tell you as a fact they are different resistances, it is 10 vs. 14 gauge there will be a difference (you can find charts online). But whether that is going to be an audible factor I can definitely tell you it will not be, the only thing resistance would do is lower the voltage at the other end of the cable based on the load being drawn. If you want to take into account inductance and capacitance too, then we are talking about a transmission line here, and also a filter, so we go into different problems. And a multimeter would be used to measure resistance, not a soundcard.

 

[JohnFerrier]

Quote:

Originally Posted by unoab /img/forum/go_quote.gif Different resistances? I can tell you as a fact they are different resistances, it is 10 vs. 14 gauge there will be a difference (you can find charts online). But whether that is going to be an audible factor I can definitely tell you it will not be, the only thing resistance would do is lower the voltage at the other end of the cable based on the load being drawn. If you want to take into account inductance and capacitance too, then we are talking about a transmission line here, and also a filter, so we go into different problems. And a multimeter would be used to measure resistance, not a soundcard.

Sure. I, like markl, don't understand how sound cards can be used to reveal audible differences (especially, since I question that it would reveal resistance differences, then you have capacitance, inductance, shielding, orientation to other equipment, etc).

 

[nick_charles]

Quote:

Originally Posted by markl /img/forum/go_quote.gif 1. You are comparing a copy of a copy on a different machine,

Do you mean the wav file on the website ? It is a digital copy I could download it 100 times and it will be exactly the same each time.


Quote:

you are evaluating the perrformance of *your* set-up, not the original. And what about the original set-up? Is it so hot to begin with?

The evaluation employs a Pioneer Elite DV-45A as a transport and Yulong DAH1 outboard DAC combination, connected with a Toshiba pure-glass TOSLink Cable. This is driven into a PC (using one or the other of the two cables under consideration), which has been custom-built for this evaluation. The computer sports an Intel D915GA series mainboard with the latest BIOS update (0482), a dual-core 2.8 GHz Pentium processor, one-gigabyte of Kingston brand DDR-400 RAM memory and a Seagate 80 Gigabyte, SATA hard disk drive. All other drives have been disconnected and cables removed. The mainboard hosts only one PCI card; an ASUS Xonar audio interface

My own set-up is not quite so good , but pretty decent and I have a good external DAC , amp and headphones, I bypass the DAC in my USB soundcard so I only use the USB to S/PDIF function and I have bit-perfect output.

Quote:

There will be degradation of the signal made during the copying process.

Do you mean the digitization process ?

Quote:

How good is the source of the transfer, how good is the recipient of the transfer?

see above

Quote:

If you are taking a picture of a picture, if your lens is not absolutely *perfect* (and what is), you will have artifacts, blurring, that extra layer of grain you get from a print of a print, etc.

For a digital copy this isnt an issue. I cannot help feeling you are mixing metaphors here.

Quote:

What DAC is he using to make the transfer or is it a computer? Is that buried in there somewhere? Does this source even reveal differences in things like cables in the first place? What interconnects were used to make the transfer? Are they also of low quality, further degrading sound?

See above

Quote:

2. Now, subjects are listening to the copy of a copy on their *computers*, which are poor sources compared to the high-end stand-alone machines. Are they using cheap computer speakers to try to hear the differences? I can barely use my stock Dell speakers to listen to Youtube clips.

For me an Edirol external sound receiver simply taking a bit-perfect USB stream and outputting S/PDIF into an Entech DAC then an M^3 amp and Sennheiser HD580 or Audio-Technica ATH-A700 headphones. Not stellar but surely good enough ?

Quote:

3. Who are these subjects? Audiophiles or other skeptics who can't hear cable differences to begin with?

Folks here, several posted earlier, what does skepticism have to do with it ?.

Are you suggesting that those who cannot hear differences become skeptics, or that skepticism somehow makes you unable to detect differences, or that some people decide to be skeptical and thus some organic change occurs in their auditory systems, does that then mean that you have to actively believe in the differences before you can hear them ?

Quote:

4. The test actually did reveal some (small) variances between the two samples (oops, how embarrassing). But there are some saying that this could be due to simple variances that occur with the machinery from copy to copy. OK, so it is acknowledged you are using flawed gear that can't consistently provide the same result to measure something as small as cable differences. This is a big problem with this test.

When I did the digitization of the same CD track 3 times with the same set up I did get minor variations in the wavs, I posted some images and numbers and the variations were miniscule - seriously near 44.1K sample level. For a serious test you would want to repeat the process 20 times or so to get an average picture, I do not quibble with that and hopefully the Op will be moved to do this.

Please dont use the word copy when you mean digitization it is hard to get what you mean and it is misleading.

Quote:

This "test" again makes the same mistake that the skeptics continue to make WRT to cables-- these differences are *small*. Now you are taking a recording of the original event to evaluate an original, possibly using crummy eqipoment to do it with. How do these differences even have a chance of showing up?

As above the recording kit does not look crummy, though even my laptop without my new and improved sound recording get-up could show differences between the analog outputs of two different CD players easily and measurably and my new set-up is quantifiable superior, it measures far better.

Just how small are these cable differences, I can resolve down to the sample level, if they are smaller than that then it doesnt really seem worth the bother

 

[markl]

Quote:

Do you mean the wav file on the website ? It is a digital copy I could download it 100 times and it will be exactly the same each time.

You are listening to a recording of an event (the output of the analog signal), that's copy#1. You are also listening to the recreation of that event on *your* gear (it has to go through your analog outputs, too, to get to your ear), that is copy #2.

Quote:

The evaluation employs a Pioneer Elite DV-45A as a transport and Yulong DAH1 outboard DAC combination, connected with a Toshiba pure-glass TOSLink Cable

You have 3 components here that require power cabling-- the transport, the DAC, and the PC recording their output. Where was the "good" cord placed? Why there and not somewhere else? Would it not make more sense to replace *all* these cables, rather than diluting the effect of the one cord you are trying to measure?

Quote:

Do you mean the digitization process ?

Yes, a digital picture of your backyard is not the same as the view from your window. Now we are introducing all the artifacts that digital recording entails. We are limited to the 16/44.1 resolution of the event, which is not the 100% resolution of actually listening to the actual analog output yourself. Ask a mastering engineer like Steve Hoffman, he will gladly tell you (what we all know) that digital has a "sound". This sound deviates from the absolutely perfect and is a coloration. You are adding this coloration to the sound when you record the output.


Quote:

For a digital copy this isnt an issue. I cannot help feeling you are mixing metaphors here.

You informed me the OP was not simply transferring digital data, he recorded the analog output of his system for this test. There will absolutely be a difference between the actual analog output and the digital recording of it. Also, you are then passing this along to be processed by your own PC, sending that signal back out through the noisy, low-quality analog output section of your sound card. Then you pass that signal on again into your headphone amp for even further distance from the source signal and more degradation. No one should labor under the illusion that they are hearing *exactly* what they would hear by listening to the OP's actual system. Not even close.


Quote:

For me an Edirol external sound receiver simply taking a bit-perfect USB stream and outputting S/PDIF into an Entech DAC then an M^3 amp and Sennheiser HD580 or Audio-Technica ATH-A700 headphones. Not stellar but surely good enough ?

Not picking on you, you have what you have, but Entech, is that that ancient Monster Cable litle DAC? Or am I thinking of something else? Not familiar with your headamp.

Quote:

Folks here, several posted earlier, what does skepticism have to do with it ?

Placebo cuts both ways. If I expect a certain result (or non result) I will likely get it.

Quote:

Please dont use the word copy when you mean digitization it is hard to get what you mean and it is misleading.

Sorry, but you are making a digital recording of sound, namely the output of a couple RCA jacks. It's not that different than taking a mike and making a digital recording of the ambient sounds in your room or the traffic outside. It may sound pretty good, but it's still a recording/copy. the quality of that recording will depend largely on your recording device. How good do you think the computer soundcard's input is? Studio grade? How much time did the designers spend on this feature very few people will use? It may get overloaded during busy/loud passages and distort. Maybe it lacks full dynamic range and compresses the the sound. Maybe it's made of cheap parts that add grain to the sound? Who knows.

What you are doing in effect is "mastering" an analog recording. Big mastering studios that do this professionally to wring every last drop of sound out of an analog source have hundreds of thousands of dollars worth of better equipment than your humble sound card.

If you don't think the quality of digital transfers can vary, compare an antique CD from 1984 to one made today (that has not been compressed to death). The old ADC chips and associated gear were vastly inferior to what we have today. And what the OP is using in his computer is vastly inferior to what is needed to do it as well as modern technology can (which still ain't "perfect".

Quote:

Just how small are these cable differences, I can resolve down to the sample level, if they are smaller than that then it doesnt really seem worth the bother

Please tell me what you are attempting to measure and how do you quantify it and what machines are you using to measure it? Please show me the machine that measures soundstage, air around instruments, the firmness of the bass, speed of attack, decay of notes, etc.

You will tell me you are looking at either or both of the two measurable things:

1. The output/peak levels (volume). Well, if a power cord changed peak levels or made certain passages louder and other quieter, that would indeed be a remarkable finding.

2. The frequency response. All those measurements can tell you is, is there a signal present at 1,000 Khz? Is there a signal at 125Hz? Etc. Etc. It tells you absolutely zero, about how those frequencies sound, and how the whole gestalt of frequencies in that area sound together playing actual music.


Man, it's my free night and I keep getting taken away from my chance to do some actual music listening. Bye!