Poll: Are perfect 1:1 copies possible in the digital domain?

Loud & Clear

OleBoot wrote: »

Certainly in my case - I can't see past my stomach to read the micrometer anyway.

I think I answered the question asked in this thread in a previous post and was going to expand on it, but this thread, if you ignore the butting of heads, seems to me to beg the question of how much you need to know how audio gear works to get the best out of it. Me, I love to know how stuff works. Computers, audio, cars, dishwashers, wheelbarrows, interossiters, positronic brains - whatever I come across. Why? I just get a kick out of when you understand something and the light bulb lights up. I'm not saying I understand things at expert levels of detail, or that I want to - I am very, very far away from being one of those genius people. But I can learn enough to understand if I am capable of fixing my car so as not to pay someone to do it, and enough about my audio hobby to satisfy my curiosity. Dear Abby, am I a geek?

So does knowing a bit about how audio gear works allow me to know how something will sound? Er, kinda, sorta, maybe sometimes, probably not. I'll start with a sort of backwards example. I picked up a CD some time ago when I went to listen to a band. You know, the sort they have in a rack out front that was probably produced by one of these web site copying services that copies them on CD burners than run faster than your average turbocharger. I basically bought it as the band were good, young and local to my area and I wanted to support them. It laid around a while before I gave it a listen, and it turned out to sound not bad. Not, good, but not as bad as one would expect. Sometime later I was in the process of putting all my CDs onto a computer , and this one was in the heap. When I tried to rip it, I got hundreds upon hundreds of read errors. I even tried another CD unit as I didn't believe the CD player could have even played it. So I told the computer to ignore the errors and played it on the computer system. Sounded pretty much the same as it did on the CD player. From what I thought I knew, this should have sounded like cr@p. If it's not bit perfect it will not sound good. jeez.

Have you gone to sleep yet? No? So here's another example for insomniacs. Some time in the last century (still feels strange to be able to say that) I decided to replace the Sony 5.1 channel receiver in my home theater system with separates. My knowledge told me that that a standalone processor would have a better DAC, separate power amps would have dedicated power supplies and be capable of higher current, etc., etc. ,etc. Of course I supplemented this by getting getting other people's opinions, and had friends that had better gear than me that I listened to, but what I thought I knew was a big factor. So I spent some weeks going to hi-fi stores auditioning gear, and finally managed to go back to the store I had selected to buy from on a weekday afternoon. I went in, and there was absolutely nobody around. I let out a "Hellooooo", and a guy appeared from one of the audition rooms. I said I had been in earlier and was looking to replace my home theater receiver. He obviously didn't remember me and that I had listened to separates before, and said that I should really come and listen to something that they had just got in. We went back to the room that he appeared from, where the rest of the shop staff were sitting. And two hours later I left with a 5.1 channel receiver. It's a Denon AVR 4800, which I have to this day.

So I guess my conclusion is that if you want to find out how things work like I do, in terms of audio gear, never ever let the knowledge you have get in the way of listening.

Enjoyable read.

Habanero Monk

DarqueKnight wrote: »

But...your ears will only work properly when you are blindfolded.:sad:

I thought we were talking about 1:1 perfect digital copies? Where you able to generate a MD5 hash on the PDF file?

DarqueKnight

Habanero Monk wrote: »

I thought we were talking about 1:1 perfect digital copies? Where you able to generate a MD5 hash on the PDF file?

PDF file? I thought we were talking about digital audio?

I have already established that there is no such thing as a perfect 1:1 digital audio copy because the transport medium will always introduce some level of noise into the digital signal, and that noise corruption will be stored in the digital file.

The amount of noise corruption can affect the quality of D/A conversion and produce audible differences.

One of the basic principles of communications theory is that an electronic signal will always suffer some type of noise corruption and distortion while passing through a transport medium. The relevant question is whether the noise and distortion obscures information and not whether the received signal is a perfect copy of the transmitted signal.

headrott

I am still trying to figure out why villian (and Habanero Monk) feel that using a computer to "validate" that a file is transferred "perfectly" as a 1:1 bit copy is an actual validation of this? First off, this is not a first person evaluation of the music file. this is using a computer program designed by someone else to evaluate a file, and the computer is doing the evaluating. Certainly not first person.

A computer and the program used to measure that there are no errors can only measure exaclty that. There are no "errors". No one is arguing that a "seemingly" error free digital audio (or any type of file) is passed through from a source to the D/A converter "succesfully". Of course it was transferred "succesfully", that is why you are listening to it and/or viewing it.

This fact does not make it a "perfect 1:1 bit copy" however, even if a computer program says there are no "errors". This simply means that the minimum level of passing the signal was met. The minimum error threashold is NOT a perfect 1:1 bit copy as there are noise and other "contaminants" introduced to the audio and/or video, etc. file(s). The program used to measure a "perfect 1:1 bit copy" that is error free is not/ cannot looking for a true noise free file (which would make it "perfect"), it is looking for the minimum amount of voltage pulse to successfully produce 1's and 0's (bits) and that is all.

Habanero Monk

DarqueKnight wrote: »

PDF file? I thought we were talking about digital audio?

I have already established that there is no such thing as a perfect 1:1 digital audio copy because the transport medium will always introduce some level of noise into the digital signal, and that noise corruption will be stored in the digital file.

The amount of noise corruption can affect the quality of D/A conversion and produce audible differences.

One of the basic principles of communications theory is that an electronic signal will always suffer some type of noise corruption and distortion while passing through a transport medium. The relevant question is whether the noise and distortion obscures information and not whether the received signal is a perfect copy of the transmitted signal.

The title of the thread is: Are perfect 1:1 digital copies possible?

It's just an exercise in establishing if it's mathematically possible to, in the digital domain, to have a perfect copy of the source file.

It could be a PDF, it could be a .TIF, .WAV, .MP4. I used PDF just for the smaller size of the file would make it easier to implement.

Is a file with vector based data in it a problem?

I am interested in if, using your basic principals of communications theory, we see:

1. The PDF transferred downloaded and MD5 hash ran does the hash stay the same? If different does the file still open?

2. If different and still opens can there be a visual tell in the PDF?

No you haven't established that noise can be stored in the file. But it's a good question: Can we simulate that? I was interested in one of the papers I linked to about the CRC errors introduced and curious if that simple meant retry until complete? And if it completed I wonder if the hash would be the same?

So are you of the belief that two files can come out mathematically/computationally the same but sound, look, read, whatever different?

Habanero Monk

ZLTFUL wrote: »

"Come on guys! VALIDATE ME!"

It's easy enough to participate. I don't see why posting a file with an MD5 hash and having the equivocators give it a shot is problematic?

DarqueKnight

Habanero Monk wrote: »

It's just an exercise in establishing if it's mathematically possible to, in the digital domain, to have a perfect copy of the source file?

You first need to arrive at a good understanding of noise and distortion effects on signal integrity. Once you do that, it will become obvious to you that the concept of a "perfect" copy of any transported signal is pure fantasy.

headrott

DarqueKnight wrote: »

You first need to arrive at a good understanding of noise and distortion effects on signal integrity. Once you do that, it will become obvious to you that the concept of a "perfect" copy of any transported signal is pure fantasy.

I would say that villian and Habanero Monk need to start an even more basic level of realizing that digital signals are not impervious to the laws of physics.

Habanero Monk

DarqueKnight wrote: »

You first need to arrive at a good understanding of noise and distortion effects on signal integrity. Once you do that, it will become obvious to you that the concept of a "perfect" copy of any transported signal is pure fantasy.

I was reading up on some serial encoding schemes where it was +5- +12 and then -5 - -12. Anything in between constituted noise and therefore discarded. So that was a 10v sweep for noise or other error.

How do you reason with a file that has the same MD5 hash at both the source and destination?

Habanero Monk

DarqueKnight wrote: »

You first need to arrive at a good understanding of noise and distortion effects on signal integrity. Once you do that, it will become obvious to you that the concept of a "perfect" copy of any transported signal is pure fantasy.

We aren't transporting a signal as you imply. We are however using a signal to transport some data.

The thread title is "Are digital 1:1 copies possible". Not: "Are various signals always 100% pristine?"

The difference is not semantic.

headrott

Incorrect. The title is "Are perfect 1:1 digital copies possible?". That is symantics.

Edit: But you are changing the argument.

Habanero Monk

headrott wrote: »

Incorrect. The title is "Are perfect 1:1 digital copies possible?". That is symantics.

Edit: But you are changing the argument.

Perfect should have been included. Not changing the debate.

So when you have a file that is mathematically correct both at source and destination on a computer is that file a perfect 1:1 copy?

SCompRacer

DarqueKnight wrote: »

You first need to arrive at a good understanding of noise and distortion effects on signal integrity. Once you do that, it will become obvious to you that the concept of a "perfect" copy of any transported signal is pure fantasy.

Isn't the term for not achieving a perfect or 1:1 copy called Generation Loss? That is IF the equipment is not performing properly, you will not achieve a perfect copy? IF it is, you will?

villian

F1nut wrote: »

You should know, your expertise in both is unmatched even if you can't spell one of them correctly.

The keyboard I was using has a few sticky keys, and not the kind you get from pressing "Shift" repeatedly. Trust me, I know how to spell

DarqueKnight wrote: »

PDF file? I thought we were talking about digital audio?

Does it matter? Does an ethernet cable, computer, or decoder "See" the 1's and 0's of a PDF file differently than the 1's and 0's of a digital audio file?

Come on DK, this is elementary digital communications stuff we're talking about. You should know, you're the expert here..right?

Dr. Raife F. Smith II, Professor, Electrical Engineering
Ph.D. Electrical Engineering, Tulane University

Dr. Smith's teaching and research interests are in the areas of Communications Signals and Systems, Broadband Telecommunications Network Design and Optimization, and Stochastic Modeling. Dr. Smith's consulting interests are in enterprise network and public switched network modeling, design and optimization.

headrott wrote: »

I am still trying to figure out why villian (and Habanero Monk) feel that using a computer to "validate" that a file is transferred "perfectly" as a 1:1 bit copy is an actual validation of this? First off, this is not a first person evaluation of the music file. this is using a computer program designed by someone else to evaluate a file, and the computer is doing the evaluating. Certainly not first person.

Exactly, because the computer is validating the 1's and 0's, not the sound. There is NO sound in a file stored on a computer. It's just a bunch of 1's and 0's. A signaling method to represent "Data"..another non-tangible thing.."Stored" on a computer. The computer program evaluating and checking the integrity of the file thus has but one thing to evaluate: 1's and 0's. If the 1's and 0's are the same as the source, then it's impossible for the "Sound" to be different when output.

The "Sound" is not literally stored on the computer, it's a bunch of 1's and 0's that will later be decoded as a sound. I'm not really sure how else to put this, but you're the one who claims to understand how computers and digital storage works..so I'll let you figure that out.

headrott wrote: »

A computer and the program used to measure that there are no errors can only measure exaclty that. There are no "errors". No one is arguing that a "seemingly" error free digital audio (or any type of file) is passed through from a source to the D/A converter "succesfully". Of course it was transferred "succesfully", that is why you are listening to it and/or viewing it.

This fact does not make it a "perfect 1:1 bit copy" however, even if a computer program says there are no "errors". This simply means that the minimum level of passing the signal was met. The minimum error threashold is NOT a perfect 1:1 bit copy as there are noise and other "contaminants" introduced to the audio and/or video, etc. file(s). The program used to measure a "perfect 1:1 bit copy" that is error free is not/ cannot looking for a true noise free file (which would make it "perfect"), it is looking for the minimum amount of voltage pulse to successfully produce 1's and 0's (bits) and that is all.

Can you provide an example or explain to me how noise, or any of the other things you claim are introduced into copied files are stored within or alongside those 1's and 0's..let alone how they are reproduced? Computer's only see 1's and 0's, so where are the 1's and 0's of the noise and other abberations that you allude to?

The computer does the evaluating because the computer has the ability to see and interpret 1's and 0's. Can you see 1's and 0's? Furthermore the computer isn't just "Checking for errors.." it's checking each and every 1 and 0 against the source 1 and 0. If all those come back lined up in the exact same way that they were in the source file, then there is no error. That's how digital works. There's nothing else hidden in those 1's or 0's, and there's no real "Error" to look for. It has nothing to do with voltage pulses and what not. The computer could care less about any voltage variation in the cabling that got it the 1's and 0's. It's only language is 1's and 0's, not voltage, not noise, not fire, not dragon, not amperage, not phase loop linear dielectric dection chromolyium bendium polymer. 1 and 0. It's that simple. I promise.

You're clearly still stuck thinking about digital technology in an analog way when they are simply not comparable. They are two totally different technologies with completely different workings. It's like comparing a Hippo to a Skilcraft Pen. They just aren't the same, nor do they share a single thing.

Habanero Monk wrote: »

So are you of the belief that two files can come out mathematically/computationally the same but sound, look, read, whatever different?

I'll tell you one thing. Reading this thread today is a completely different experience than yesterday. The screen smells quite bad, whereas yesterday it had this nice flowery aroma. I also don't get the same smell when I load the page late at night, or at a friends. I'm definately with DK on this one. Just because the 1's and 0's are the same doesn't mean that the web page won't sound, look, read, or even smell differently than it did the last time it was loaded. My nose is the proof. I doubt it has anything to do with the coffee I spilled on the keyboard. It's the bits.

nooshinjohn

villian wrote: »

I'll tell you one thing. Reading this thread today is a completely different experience than yesterday. The screen smells quite bad, whereas yesterday it had this nice flowery aroma. I also don't get the same smell when I load the page late at night, or at a friends. I'm definately with DK on this one. Just because the 1's and 0's are the same doesn't mean that the web page won't sound, look, read, or even smell differently than it did the last time it was loaded. My nose is the proof. I doubt it has anything to do with the coffee I spilled on the keyboard. It's the bits.

Then perhaps it's time to flush this **** and make a new thread with some of your fresh BS...

Habanero Monk

ZLTFUL wrote: »

"Come on guys! VALIDATE ME!"

ZLTFUL wrote: »

Translation: I have to be right at all costs so somebody needs to validate me! :rolleyes:

Don't need anyone's validation. Just your $1600 :biggrin:

Habanero Monk

ZLTFUL wrote: »

I don't need validation...I just need validation.

If you don't have validation disorder, why are you all freaked out because people aren't validating your "test"?

It's an easy test. Why are people freaked out about not trying it and posting their result?

Habanero Monk

DSkip wrote: »

Because we don't care? Any weight this debate has was lost about 30 pages ago.

You can only lead a horse to water. You can't make it think.

tonyb

Isn't that the problem in a nutshell ?

Monk wants to "think", while everyone else wants to "listen" ?

F1nut

tonyb wrote: »

Isn't that the problem in a nutshell ?

Monk wants to "think", while everyone else wants to "listen" ?

You're so close.....Monk wants everyone to think and do as he does, like the bozo at 1600.

Habanero Monk

It took me a whopping 5 minutes to build an example that can be verified by any member here. May be DK can ask someone in the Math department how it's after all possible.

I find it strange that for all the people wanting to have their voice heard in this thread that they can't take the time, that a well formed posts takes, and verify independently the PDF file and see the MD5 hash matches up to others.

It's a perfect 1:1 digital copy.

Stupidity is ignorance willfully left uncorrected. The fundamental misunderstanding in something so trivial has left me wondering what other things you take for granted that you are also foundationally incorrect on.

F1nut

^Need more proof that my comments in post #159 were spot on?^

Nah, didn't think so.

headrott

Habanero Monk wrote: »

Perfect should have been included. Not changing the debate.

So when you have a file that is mathematically correct both at source and destination on a computer is that file a perfect 1:1 copy?

I've already explained multiple times that until you can "see" deeper than mathematically "accurate" 1's and 0's being transferred you will not undertand. Care to look deeper and actually learn something? OR you can continue believing that there are actually digital transfers that are completely impervious to the laws of physics. That's up to you.

headrott

villian wrote: »

Exactly, because the computer is validating the 1's and 0's, not the sound. There is NO sound in a file stored on a computer. It's just a bunch of 1's and 0's. A signaling method to represent "Data"..another non-tangible thing.."Stored" on a computer. The computer program evaluating and checking the integrity of the file thus has but one thing to evaluate: 1's and 0's. If the 1's and 0's are the same as the source, then it's impossible for the "Sound" to be different when output.

The "Sound" is not literally stored on the computer, it's a bunch of 1's and 0's that will later be decoded as a sound. I'm not really sure how else to put this, but you're the one who claims to understand how computers and digital storage works..so I'll let you figure that out.


Can you provide an example or explain to me how noise, or any of the other things you claim are introduced into copied files are stored within or alongside those 1's and 0's..let alone how they are reproduced? Computer's only see 1's and 0's, so where are the 1's and 0's of the noise and other abberations that you allude to?

The computer does the evaluating because the computer has the ability to see and interpret 1's and 0's. Can you see 1's and 0's? Furthermore the computer isn't just "Checking for errors.." it's checking each and every 1 and 0 against the source 1 and 0. If all those come back lined up in the exact same way that they were in the source file, then there is no error. That's how digital works. There's nothing else hidden in those 1's or 0's, and there's no real "Error" to look for. It has nothing to do with voltage pulses and what not. The computer could care less about any voltage variation in the cabling that got it the 1's and 0's. It's only language is 1's and 0's, not voltage, not noise, not fire, not dragon, not amperage, not phase loop linear dielectric dection chromolyium bendium polymer. 1 and 0. It's that simple. I promise.

You're clearly still stuck thinking about digital technology in an analog way when they are simply not comparable. They are two totally different technologies with completely different workings. It's like comparing a Hippo to a Skilcraft Pen. They just aren't the same, nor do they share a single thing.

It's already been explained multiple times and yet you refuse to accept and/or understand it. I cannot help you with that.

Why do you believe there is any transfer of anything (using the technology we posess) that is not subject to degredation? That is an irrational thought.

Habanero Monk

headrott wrote: »

I've already explained multiple times that until you can "see" deeper than mathematically "accurate" 1's and 0's being transferred

Deeper than mathematics? Are you serious? Mathematics is the language of our universe.

What laws of physics are changing a file that computationally, mathematically, equals?

You haven't explained anything. At least I have a provided a simple proof for verifying a file is indeed the same.

headrott

Habanero Monk wrote: »

Deeper than mathematics? Are you serious? Mathematics is the language of our universe.

What laws of physics are changing a file that computationally, mathematically, equals?

You haven't explained anything. At least I have a provided a simple proof for verifying a file is indeed the same.

If you stop twisting the meanings of what is said (to suit your agenda), you may understand.

villian

Will somebody please explain to headrott how digital data works? Specifically the absolute value of a 0 and a 1.

Habanero Monk

headrott wrote: »

If you stop twisting the meanings of what is said (to suit your agenda), you may understand.

So far there is zero evidence (yes proof) that a file that mathematically checks out the same at both source and destination is anything other than the same file.

So far, in this thread, I'm the only one to provide a proof.

Habanero Monk

ZLTFUL wrote: »

For someone who doesn't need validation, you sure are seeking validation...
http://www.avsforum.com/forum/173-2-channel-audio/1586722-perfect-1-1-digital-copies-possible.html

I started that over there to see how the percentages would differ. That's why I kept the poll responses the same. You're welcome to post over there also.

I have a feeling the response to the poll is going to be vastly different there.

You confused it for curiosity.

Habanero Monk

ZLTFUL wrote: »

Your problem is...well, yours and villains is that you posed a computer question to an audio forum.
And not only that, but a vague one at that.

Theoretically, yes, digital copies should be 1:1 bit perfect copies of each other. But that is assuming a perfect transfer to a perfect media from a perfect media.
But it is naive to think that everything in the chain is going to work perfectly.
Your example itself is flawed because the size of the file reduces the chance of an error occurring...a bit being "flipped" if you will.
The larger a file size gets, the more likely you are to encounter this.

A CD is a digital media correct? And the files stored on a CD are digital files correct? And the pits on a CD represent a 1 or a 0 based on their size correct? But yet, a 1 pit can be a range of sizes as can a 0 pit. These pits aren't perfectly sized exactly the same. Their size can be within a range that defines it as either a 1 or a 0. That alone shows an example of a digital copy *working* correctly but the possibility being there for a non-perfect copy of that digital file.
Data CDs do indeed have more rigorous error checking on them. However what's interesting is how audio CD error checking works - if the disc isn't being read properly it will obviously try and correct it, and if it fails to read the missing data properly it "bridges" the gap by interpolating what should be there (unless it's really missed a lot, in which case it jumps or whatever). So what happens is that the fine detail in the music is reduced. If your CD ROM drive is doing this when you read a CD before copying, the copy will obviously not be as good as the original, even though everything has been digital throughout. The difference will only be apparent when you play it on a better CD player of course, but still it knocks the "perfect digital copy" argument on the head.

We won't even get into the shortfalls of magnetic media like tape. It presents its own massive issues in possible degradation of the file.

And how about the law of thermodynamics? It's impossible for files not to degrade over time without intervention because of thermodynamics.

At best, with sufficient redundancy and error correction algorithms, and good equipment maintenance, you can drop the probability of error ludicrously low but still not absolute 100%.

The Computer Music Center at Columbia University did a great paper (trying to track it down) on Computers and Music. They state that while digital copies *should be* perfect, there are times when they simply weren't. That for some unexplainable reason (ghost in the machine stuff) the copy was audibly different even though, for all intents and purposes, it was a "perfect" digital copy of the original.

Before you posted such a lengthy reply did you consider that it doesn't preclude a perfect, mathematically, identical file?

It's the classic F.U.D tactic: Fear, Uncertainty, and Doubt.

I've never said you would always get a perfect copy. I never precluded exigent circumstances. I don't operate with a sky is falling approach.

Hard drives break. Tape deteriorates.

Now here is another question: How is it in a 3 drive RAID 5 you can lose 1/3 of your drives and still not miss a single byte or bit of data?