Here's a thought experiment.
Assume I have two files A and B, both bit identical, but with audible differences.
I print both of them out in hex form and compare the printouts to be sure they're the same.
I open a hex editor on my computer and type in the hex values from the printout by hand.
I listen to the resulting file.
Which of the two original files, A or B, will the newly-typed file sound like?
Assume I have two files A and B, both bit identical, but with audible differences.
I print both of them out in hex form and compare the printouts to be sure they're the same.
I open a hex editor on my computer and type in the hex values from the printout by hand.
I listen to the resulting file.
Which of the two original files, A or B, will the newly-typed file sound like?
Yes, it agrees with your post.
Haven't really proved anything. But it's a nice test. I don't think it's news, tho. It's all been written about many times before. Just the first time I've had problems with a brand new CD.
Does this really translate into errors or more noise in a CD player at 1X speed? I have not seen any measurements to support that, tho it is plausible.
That would depend on whether you used a computer with a noisy power supply, or typed the files onto a HDD or flash drive, I suppose.
Worth noting the language, viz: "It appears that the files may have suffered some degradation while being moved around between HDDs during renaming."
So what were these degradations? How did they "appear"? Why the equivocal nature of the comment?
And what, pray, would be the "usual alteration" made for audio or multimedia? In fact, what was the specification of the computer at all? And why would that spec lead anyone to a conclusion that it may induce some degradation in the simple act of moving and renaming a file?
Forgive me if this is all asked and answered, but ploughing through a seemingly endless stream of patently illogical argument and/or invective was more than I could cope with.
Hi MarkT,
It depends which printout you used. 😀
Have you thought of using a binary editor?
regards
Actually, I think the equivocal part is spot on. He has no evidence, didn't really check things over, and is being absolutely honest about that. I wish he (and other posters) were more equivocal about other things asserted as bald facts, but hey, this is a good start.
Parkinson's induced jitter may be very audible. Your "thought experiment" requires understanding of what is a file and being able to follow basic logic. Be reasonable, we are just diyers blessed with good hearing.
You have to have a very steady hand when issuing the copy command, else you'll get the old wow and flutter in your file.
The Story of the Thread So Far
Time for a little summary.
There are several subplots here.
The Claims
The Counter Arguments
I believe that sums up the arguments so far. If I have missed anything important or misstated anything, please let me know.
Pano.
Time for a little summary.
There are several subplots here.
The Claims
- Different CD media can sound different even if the files contained on the CD are the same. Differences in reflective index, sharpness or depth of the pits, etc. may make some CDs harder to read than others. This may result in jitter or noise from the transport. Analog quality is damaged.
- Files carefully ripped from an inferior quality CD and written to a better quality CD-R or CD-RW will sound better than the original when played back on the same transport. The new CD being easier for the transport to read, will result in fewer errors and lower noise.
- Solid State drives sound better than moving drives for computer playback. Similar to argument #1. Noise created by the spinning hard drive and its motor can cause jitter, errors and noise. These can find their way into the final analog signal and degrade the sound.
- Files ripped to a solid state device sound better than those ripped to a hard drive. Similar to numbers 1,2,3 above. Lower system noise during the rip results in better files. A USB device powered by a linear psu is said to be best. This advantage can be heard even when the file is transfered to another USB drive.
- Checksums do not reveal all the differences in the files. Listening tests indicate differences from very subtle to obvious between files with the same checksum.
- System noise can find its way into the audio files during a rip or file transfer. Lower noise systems result in better audio files and rips.
The Counter Arguments
- Arguments 1 and 2 above are plausible, but have not been proven. It may be possible for the strain of reading a difficult CD to result in audible change. No definitive proof has come forward so far.
- Bits are bits. It is impossible for two identical files to sound different on the same playback device. Any differences heard are imagined. The files are exactly identical.
- No theory has been advanced as to where sonic difference may lie in two audio files that have the same checksum or same exact content. There is nowhere known for the extra information to reside.
- Files read off HDD or solid state drives are buffered into RAM first. Thus any advantage of the solid state device is eliminated by the buffer.
- Files of many megabytes or even terabytes can be transfered successfully from drive to drive, even across vast networks, without a single error.
- Normal system noise does not affect the content of audio files.
I believe that sums up the arguments so far. If I have missed anything important or misstated anything, please let me know.
Pano.
Good summary, I'm most interested in No.5 especially the case where the files are bit by bit the same. I wonder if Shannon would even bother LHAO.
Yes, I agree, a good summary.
To have any hope of resolving this stuff one of two things needs to happen.
Either:
(1) there is an agreement between both sides of the argument on a test protocol that will satisfy both sides that the claims of a difference in perceived sound are true (as opposed to just a claim). I can't see that happening because those who believe in controlled listening tests are unlikely to get the other side to do such tests. So there seems to be no bridge there.
(2) some engineering explanation for the claimed difference, based on known mechanical or logical principles, can be provided. None has been provided so far. One may be provided. But I think the chances are low.
Last edited:
Thinking out loud, the way the file is stored might influence quality. Suppose i have a mediaplayer with hdd inside. Suppose i rip a track from an usb stick. Suppose i rip the same track from a cd some time later. Same file, same checksum (when copied to a computer and compared). Now what if there is a difference (first one better), voodoo?
What if i defragmented the hdd between storing the first and the second file and what if the mediaplayer is really crappy and the hdd reading of data (heads moving around) influences the sound..?
But this is a form of playback differences, similar to good cd vs bad cd.
Note that the above is just thinking out loud. Not fact and i don't think this occurs in reality (and in my case there is ethernet between hdd and mediaplayer). Just to say that differences could be explained technically, even if not obvious. Two files cannot be stored physically on the same location on a hdd, stick or memory.
Dont believe 6. But it could in theory maybe influence the way the file is stored (?). Rip a crappy cd and somehow that results in the file being stored in small fragments on a hdd (ripping takes ages, small fragments all over the place. Crap software without buffers etc etc) compared with ripping from a "perfect" cd, so the file is stored "nicely" on the hdd (meaning less fragmented). And the same playback influence as above. Again, not describing reality, just trying to find a technical argument for 6.
Equal files are not stored equally on a physical device (impossible with two files).
Again, don't believe 6 as such. Suppose somebody would give me a "good" and a "bad" file. Store the good one and play it and then store the bad one at exactly the same location (overwrite bit by bit with the same data, so at the same location on hdd, memory card). Then there will be no difference when playing the bad one (apart from other variables then good/bad file).
Last edited:
My problem is that people couch their claims in a way to force first principles to be wrong. Even guys like Charles Hansen (who should know better) do that at times. Claims that there is "information" other than 1's and 0's in a digital file are absurd. The constant attack on the people and ideas that built this entire industry is infantile and an embarrassment.
ohh ffs. A listening test is not a valid test of the file only - you are not listening to the bitstream.
Identical is identical - which part of this is hard to understand peps?
The case for claiming that the file storage makes the difference is a weak as a claim that a book's language changes sublty but undeniably if read from a Kindle instead of from paper.
Its patent bollocks.
I will grant that the experience of reading from a bound 1st edition can be aesthetically more pleasing than that of reading from the somewhat impersonal screen of an e-book. But the MESSAGE REMAINS THE SAME.
Are there valid technical reasons that a file retreived from a solid state device might sound different from that retreived from an optical or magnetic device - yep for sure. But that can only be about the process of retrieval, not the file data.
As the guys sang so well, the song remains the same.
Identical is identical - which part of this is hard to understand peps?
The case for claiming that the file storage makes the difference is a weak as a claim that a book's language changes sublty but undeniably if read from a Kindle instead of from paper.
Its patent bollocks.
I will grant that the experience of reading from a bound 1st edition can be aesthetically more pleasing than that of reading from the somewhat impersonal screen of an e-book. But the MESSAGE REMAINS THE SAME.
Are there valid technical reasons that a file retreived from a solid state device might sound different from that retreived from an optical or magnetic device - yep for sure. But that can only be about the process of retrieval, not the file data.
As the guys sang so well, the song remains the same.
I have read through this thread, but I wonder why I have bothered. Perhaps I was hoping for science to prevail, but so far it has not. I suspect the basic problem is that many people are not aware of just how powerful the placebo effect is, even when you know how powerful it is. Bad temper on both sides has not helped, although I realise how frustrating it can be to explain something to someone who either lacks the foundation to understand what is being said or has already decided to reject certain types of evidence.
Let us suppose I carried out the experiment, with a suitable protocol, and found that I could tell the difference. I would conclude that I was mistaken, or that it was a statistical fluke no matter how many trials were carried out. This is because when it comes to audio I do believe in science. There are unknowns and mysteries, but this is not one of them. Some may accuse me of adopting a 'faith' position, intending this to be a criticism, but that would only show that they don't understand what faith is. It is not a denial of evidence, but based on evidence (even though the evidence may be incomplete). I know enough about audio to know that if two identical files sound different when played then it is the playing which introduces the difference. To argue otherwise is to claim that black is white.
Let us suppose I carried out the experiment, with a suitable protocol, and found that I could tell the difference. I would conclude that I was mistaken, or that it was a statistical fluke no matter how many trials were carried out. This is because when it comes to audio I do believe in science. There are unknowns and mysteries, but this is not one of them. Some may accuse me of adopting a 'faith' position, intending this to be a criticism, but that would only show that they don't understand what faith is. It is not a denial of evidence, but based on evidence (even though the evidence may be incomplete). I know enough about audio to know that if two identical files sound different when played then it is the playing which introduces the difference. To argue otherwise is to claim that black is white.
Mike,
I think you've done an excellent job of summarizing the main points.
Please realize that given a number of issues where there is, at this time, a lack of absolute (or any) certainty as to the causes, that there may be differences in opinions. I believe that most will understand that with any group of engineers, etc., there will be divergent opinions on the mechanisms involved, or even in some cases, their relative importance.
I have attempted to point out, as has SandyK, what seem to be areas and even techniques, that can be tried by members of the forum to ascertain the validity of certain claims that have been made or at least indicated to be of some import.
What I have stated in this thread, are some of those things that our people have done that seem to make a sonic difference (or improvement) in our listening tests, that may be of interest to those that are pursuing improved sound.
I am not selling anything, or recommending products and was solely motivated to share some of what our people have been doing that seemed to, or might, yield positive results. However, to be accused of being some sort of charlatan, or attempting to instill "dangerous" ideas (as one person put it), is certainly not my idea of what I had hoped would be an interesting topic for discussion.
The fact that some people have made accusations that somehow there is a blatant refusal to perform ABX test is ludicrous to say the least. I believe that there are few here that are more in favor of controlled tests than I am, and we've done a number of these in the past. They are, however, hard to get organized and to exercise the degree of rigor that should be maintained. I might also mention that, many of our club members find ABX tests to be about as interesting as watching paint dry, so it becomes quite difficult to organize a meeting for just that purpose.
Best Regards,
Terry
" .... that seem to make a sonic difference ...."
Seem to. Yet no allowance or acceptance is made for the obvious possibility that there may be no difference apart from subjectivism!
Seem to. Yet no allowance or acceptance is made for the obvious possibility that there may be no difference apart from subjectivism!
- Status
- Not open for further replies.