Ah Geez - the Hard Drive Does Make a Difference

[sandyk]

Hi Claudius

Yes, it was ages ago when you posted that you also heard those differences, along with a Mastering (?) Engineer who was friend of your brother who is in the same industry.

The differences between those files was far more subtle back then, but other improvements have since been made.

I think you will find it is the same, forever questioning Clay, albeit perhaps a little corrupted these days by a mutual friend of mine from CA !

These days, I do not often UL comparison .wav files, although I do occasionally on request.

Most of the test files that I UL these days are of higher resolution, or better quality CD material where people owning the same material can compare with their own rips , and come to their own conclusions whether I am full of it or not !

If you care to email me, or PM your current email address,

I will be happy to include you on future test material upload advices.

Kind Regards

Alex

 

[Audio_ELF]

Alex,

 

Yes I recall your previous thread and our discussions regarding audio files.

 

Here is a summary of my experiences both on a subjective and objective level...

 

*) At playback, many factors can affect sound quality including (but not exclusively)

FLAC / ALAC vs WAV / AIFF even WAV vs AIFF;

location of the files - internal HDD / SSD / external USB vs FireWire / chipsets;

power fluctuations / PSU quality

And I'm sure many more.

 

*) An uncompressed music file (eg. WAV or AIFF) can be compressed using lossless codec (FLAC or ALAC) and then uncompressed and the final file is identical to the original file. This is PROVED, to a 100% certainty (to my mind), using MD5 checksum.

 

*) A CD can be ripped with a high level of certainty that the rip is accurate. If the checksum of two rips are identical then the files are identical and therefore (allowing for my first observation) the audio playback is identical. This is in the same way that 2 spreadsheets with identical checksum display the same figures.

 

*) An audio file can be moved from location 1 to location 2 to location 3 and onwards again and again with no degradation of quality weather using a $0.39 USB cable, a $3 memory stick, FTP transfer or email. Any problems will be immediately obvious and of great magnitude not "a slight increase in bass" or any such claims.

 

I think this sums it up. If people want to know if different CD readers, storage locations, etc make a difference during ripping I will defend the scientific position on these points despite claims to the contry.

 

As you may remember I did download your files in the past and found no difference; an observation you put down to my system not being resolving enough as I recall.

 

Eloise

 

[sandyk]

"As you may remember I did download your files in the past and found no difference; an observation you put down to my system not being resolving enough as I recall."

Eloise

Yes, I well remember that you thought that a track from a Telarc Audiophile CD sounded to you like an old Beach Boys vinyl recording, and Peter enlightened you as to the source of the posted material.

That is exactly why I no longer NORMALLY provide comparison .wav files to people with less resolving equipment, or those that use routers, external HDD storage or wireless networks.

Best results are obtained using SPDIF (or Firewire) Out from a decent soundcard into a good quality DAC, and preferably via headphones or a good speaker system.Preferably, the files should also be stored on and played from a NON MOVING storage medium such as a SSD.Many other members have already expressed preference for SSD over HDD, even though the file may be played back from system memory !

I have found that USB playback is usually inferior due to SMPS limitations unless something like the HiFace is used in between. The JK modified HiFace with Lithium batteries does an even better job, and there are now several reviews available of it's performance including wideband CRO screen photos of the waveform improvements.

Alex

P.S.

I think that you and I should agree to disagree on this matter, as it is merely rehashing much earlier posts ?

 

 

 

[Audio_ELF]

"Yes, I well remember that you thought that a track from a Telarc Audiophile CD sounded to you like an old Beach Boys vinyl recording, and Peter enlightened you as to the source of the posted material."

 

I think (IIRC) I said that the "audiophile recording" made a great Beach boys track sound like at best lounge music and at worst lift music. That doesn't affect the fact I could tell which track was ripped casually and which was from more careful ripping with your super-duper tweaked CD reader.

 

Beyond that, I'm not going to argue any further about this... Though I will defend my position with science and objective opinion.

 

Eloise

 

PS suggestions of differences in "perfect" rips has nothing to do with the OPs observation that a new / different HDD enclosure affected sound quality during playback!

 

[sandyk]

" That doesn't affect the fact I could tell which track was ripped casually and which was from more careful ripping with your super-duper tweaked CD reader."

 

Eloise

A Freudian slip ? He, He !

Kind Regards

Alex

P.S.

How's the weather in the U.K. at he moment ?

It's heading for 30C here today, and more tossing and turning at night !

 

 

 

[losingmyreligion]

 

Hi Claudius,

 

yes, I am the other "Clay", formerly known as cfmsp. So, yes, I changed my moniker, apologies for the confusion.

 

clay

 

 

 

[jhwalker]

If this were merely a matter of opinion, it would be easy to "agree to disagree" . . . but it isn't. It is a fact that identical files *are* identical, and no amount of wishful thinking or hopeful listening will banish that fact. I think we all agree that the playback environment can affect the sound, sometimes drastically. But the equipment used to produce the digital rips cannot.

 

I'm old enough to have been an audiophile back in the tape / LP days and there were any number of preparations that could improve the capture of analog sound - but none of that translates to digital. A bit is either on or off, nothing in-between. And identical files are identical, period.

 

Anyway, I'm going to try to stay away from this thread going forward

 

[4est]

"A bit is either on or off, nothing in-between. And identical files are identical, period. "

 

I am not sold on the hard drives making a difference- yet. My experience in bit perfect player programs tells me that this statement is patently false.

 

Furthermore, and excuse my ignorance here, but I am curious as to why no one has questioned the validity of (MD5)check sums being proof positive of identical files. Hasn't the MD5 formula been proven insecure? My very brief wiki read said so, but I do not understand it all.

 

[Audio_ELF]

Forrest / 4est

 

The thing about checksums such as MD5 is that a small change (as may occur when a bit is misread) leads to large change in the resultant checksum. So it's possible two very different files to give the same checksum, but no two very similar files. A great test / example would be to MD5 a photo (preferably RAW or TIFF) then using a graphics programme change a few pixels. The MD5 checksum should e obviously different.

 

What you can't do with MD5 checksum (IIRC) is use it to reconstitute corrupted data. I think this maybe the reference to MD5 being insecure. Edit/Addition: Also MD5 was originally a cryptographic system and this has proven insecure.

 

The utility to create and verify checksums is MD5SUM: a standard CLI command on Mac OS X and Linux but not standard (though available for) Windows.

 

Eloise

 

[4est]

I figured there was a simple thing that I was missing in all of this.

 

[1audio]

The Wiki article has to do with the security of MD5 and the possibility of generating a cooked MD5. For checking an audio file two rips having the same checksum and being different is incredibly unlikely. Its a good way to verify files. I have used it for downloads of DVD size (6 GB) files reliably. The security in this application is pretty irrelevant.

 

[Jeff In San Diego]

First, let me say that regardless of the outcome (if any) of most of the discussions I've read on this forum...the discussions themselves are of remarkably high quality. Nowhere else on the InterTubes have I witnessed a more decorum, IQ (if not SQ), and general good-spirited camaraderie of warring factions. Well done.

 

I've been circling this interesting fight and looking for a place to jump in. For no particular reason other than that I've finished my breakfast and gotten most of it off of my fingers for keyboard safety...here I go.

 

Clay, good man, you were doing just fine right up until you summed it up with this:

"These are likely only some of the aspects of data files (and how we access them) that could conceivably influence the sound produced by one's DAC, despite that (in most cases) file comparisons would elicit the same checksum"

 

Please allow me to translate:

This is a long list of things which seem like they might mess up the data but which have been proven to have no impact on the data whatsoever. We therefore conclude that these feckless minutia are changing the sound but not the data the sound is produced from.

 

No, not so much. An identical stream of 1 and 0 is a logically impenetrable wall. Various types of acoustic and electromagnetic noise can flow around that wall, but bits is bits. Your DAC simply doesn't know whether or not there is a warm cup of expensive coffee sitting on top of the hard drive it's data stream is coming from.

 

I will say though that checksums completely ignore timing of bits. A checksum is a lousy race-official. It dutifully counts the contestants at the finish line, but does nothing to note their arrival times.

 

Paraphrasing some piano virtuoso from a past millennia who's name escapes me:

"Anyone can play the notes. It is the space between the notes...ahhhh, that is where the music lives."

 

So I believe I can summarize this thread from a scientific perspective thus:

Three possible explanations were given for sound differences due to the use of an SSD drive instead of physically spinning magnetic media --

1) noise (either acoustic or electromagnetic)

2) timing of the data at the input to the DAC

3) changes to the binary data delivered to the DAC

 

Number 3 can now be removed from this list as it has been categorically disproven to a mathematical certainty. Not all bits are the same...but bits that are the same... ARE THE SAME!

 

thanks for listening,

jp

 

 

 

 

[losingmyreligion]

"Please allow me to translate:

This is a long list of things which seem like they might mess up the data but which have been proven to have no impact on the data whatsoever. We therefore conclude that these feckless minutia are changing the sound but not the data the sound is produced from."

 

Just to be clear, nowhere did I posit that the "long list of things" MIGHT change the data, but rather that each are aspects of the data that might (and have been reported to by multiple sources) impact the sonics during playback ... via the mechanisms you refer to as:

"1) noise (either acoustic or electromagnetic)

2) timing of the data at the input to the DAC."

 

 

"No, not so much. An identical stream of 1 and 0 is a logically impenetrable wall. Various types of acoustic and electromagnetic noise can flow around that wall, but bits is bits. Your DAC simply doesn't know whether or not there is a warm cup of expensive coffee sitting on top of the hard drive it's data stream is coming from."

 

By responding to (your translation of) my comments with the words "No, not so much...", and with your use of "feckless minutiae" above, you seem to be suggesting here that the factors I listed in my post can not actually impact sonics?

 

If not, then I can't fathom what you found so disagreeable as to call me out with "you were doing just fine right up until..."

 

Oh wait, maybe its just that you were looking for a fight (or in your own words - "circling this interesting fight and looking for a place to jump in" ;0

 

clay

 

PS, the pianist whose name you've forgotten is Artur Schnabel.

 

 

 

 

 

[Jeff In San Diego]

Was looking for a "fight", yes (I'd like to buy an argument...)

 

I would not go so far as to say that none of the 10 items you mentioned could possibly impact sonics...but I would say that none of them will change the values of the bits that arrive at the DAC. If any of them impact sonics, they are doing it not by changing the bits, but in some other way by having in impact on "noise" (I include power supply imperfections in this) or by changing the timing of the arrival of the data in a way that a given DAC is unable to make up for (perhaps even an asynchronous DAC could be effected by large enough changes in data flow rate?). So my claim is that your list is "feckless" with respect to the data arriving at the DAC, not necessarily the sonics of the whole system.

 

Since you've taken my well intentioned jab with good humor, let me follow with another paraphrasing of another quote who's author is unknown to me:

"There are three kinds of people in this world -- those that understand binary, and the other people".

 

cheers,

jp

 

[ziggyzack]

You mean ten

 

[Audio_ELF]

I think the quote is...

"There are 10 kinds of people... Those who understand binary and the rest."

 

To misquote shows you to be one of the others! :-)

 

Eloise

 

[losingmyreligion]

 

"I would not go so far as to say that none of the 10 items you mentioned could possibly impact sonics...but I would say that none of them will change the values of the bits that arrive at the DAC."

 

I can readily agree that none of them likely change the values of the bits before they are read by a software player.

 

I'll likewise agree that it is rather unlikely that any of them change the values of the bits that arrive at the DAC.

 

I do remain open-minded to pretty much any possibility however.

 

In the spirit of sharing quotations, I'll offer this one relevant to my last point:

 

"In the beginner's mind, there are many possibilities,

in the expert's, there are few"

 

- Shunryu Suzuki (Zen Roshi)

 

 

FWIW, the latter (i.e. the expert mind) is not better than the former to a Zen teacher.

 

 

cheers,

clay

 

 

 

[Jeff In San Diego]

I'll counter with another quote:

note to Eloise -- This time I'm quoting myself, so I know I'm getting it right.

 

"Science can only progress by admitting all possibilities...and focusing on the probabilities."

 

- Jeff In San Diego (ComputerAudiophile Newbie)

 

cheers,

jp