|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
50.159.77.150
In Reply to: RE: The -90 dB argument and the "H" word. posted by Tony Lauck on April 06, 2015 at 12:40:21
"If you say you can't hear something, then that's fine with me. I believe you. I don't know why you say this, it could be you lie, you are willfully deaf (almost the same thing), you are lazy, inexperienced, or untrained or the test setup and equipment is inadequate. I do have a few questions though. If you are not hearing even the grosser forms of differences that are discussed here, why are you wasting your time on this forum? Why do you claim to be a music lover / audiophile? Why should I believe that you are not just an Internet troll?"
So, in summary, the question is: Why would you, the guy who verbally champions logic correctness seemingly every other post, follow logically valid lines of reasoning????
Beats me Tony! I figured maybe you'd want to practice what you preach. Maybe I was wrong.
Follow Ups:
If you think I am being illogical, please be specific, so that I may be able to understand you.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Deeming someone of unknown hearing acuity and experience with identifying audible issues associated with this type of playback unqualified to comment about issues with 16bit 44.1khz's inadequacy by your standards due to your assumption they can't hear what you've presumed should be hearable in this persons hypothetical test system of unknown quality is illogical.
I'm still looking for a properly crafted paragraph with correctly structured sentences. A word salad of run on sentences won't do. Until I can parse your words, I won't be able to analyze your argument.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
LOL The high road it is then I see.
Yes, you are being illogical. Unless there is responsibly gathered statistically significant evidence proving there are those who can positively identify a difference in a system which has been adequately proven to actually resolve at that level, and not just color the signal in an audible way, then the debate is ongoing.
Without statistically significant supporting experimental results suggesting otherwise, your opinions can not be qualified as fact. Not qualifying your opinions for what they really are is being dishonest. Fine if you want to lie to yourself, but post it here and someone is likely to call you on it.
For example: how do you positively know your preference for high rez over redbook files played back on your system isn't actually a preference for some coloration that high rez file playback introduces to your system?
There is no debate possible between people who have a fundamentally different concept of truth, based on fundamentally different epistemology, metaphysics and ontology. You give away this difference through your use of words such as "statistically significant", "adequately proven" and "qualified as fact". Who gets to determine what is "qualified as fact" or not? This is the essence of the matter. My position is simple: I get to determine what I consider to be fact. I don't care what other people think, except as to their ability to aid me in making these determinations. This is especially true where people use "statistics" as the basis of arguments, since the vast majority of statistical arguments are either outright bogus or rely on unstated models of causality. (Of course I am well aware that I could be wrong in my determinations. Such is the nature of personal knowledge.)My preference is for high quality recordings over lower quality recordings. I know these when I hear them. My experience is that the highest quality recordings in my library are all high res. Furthermore, my experience with hundreds of hours of using state of the art conversion tools has convinced me that the finest hi-res recordings can not be transparently down converted into the 44/16 format. In doing these comparisons my motivation was to do the best possible conversion to make my best possible attempt to achieve transparency in down conversion. Note that preference played no part in these experiments, so as to avoid the issue of coloration, especially coloration on recordings vs. coloration during playback, and the possibility of cancelling coloration. In addition, it became clear that different downsampling settings changed what was lost in the downsampling operation in terms of specific sonic qualities that were changed. I was not trying to make the "best" downsampling that everyone would accept, or even the "best" one that I would accept. I was trying to find at least one setting that would provide transparancy. I made my best effort and failed, believing that it was a fundamental problem of "putting 10 pounds into a 5 pound bag" or "squeezing a large balloon into a small suitcase." As the setting space I worked with had three dimensions and millions of possible settings, it could be that I missed the best possible setting. It is possible that better conversion software might become available in the future that provided an expanded choice of settings and more accurate execution thereof. However, I gave up. There is no reason to even consider this debate any more, since the extra cost of higher resolution is pennies per album. (Take note: I am talking about "cost" not "price".)
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Edits: 04/07/15
author compares sample rate conversion capabilities of several softwares. Has me tempted to buy the newest soundforge suite.
Not sure which article you referenced.
I have an old version of Soundforge, 10c. It does not have the latest version of the iZotope 64 bit CRC, which can be found in RX4 or RX4 Advanced. If you are doing cut and paste type editing then I prefer the Soundforge user interface. It can also use iZotope plugins, such as their parametric EQ. However, for doing sample rate conversions or serious repair work I use RX4, because it has a much better user interface for these tasks, including a spectral editor which can be used to remove various noises by selecting time-frequency ranges and editing the plots. You can download a free trial of this software. (Unfortunately, it's not cheap.)
The older version of the iZotope SRC in Soundforge 10c is considerably slower than the newer version and has a more confusing interface for controlling filter parameters. In addition, all the earlier versions before the version in RX4 have a sub-sample time offset in their output. This makes it impossible to get a deep null when doing testing. The new version in RX4 will give a deep null if you feed it a properly bandlimited sample and then resample it (up and down or down and up). While this time alignment has little audible effect when listening to one track it is a big deal if one is doing comparison testing and null testing.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
It is quite true I have defaulted to the methods of data analysis shown to me during my college undergrad classes without ever really considering questioning, with any depth, the validity of the approach. It's always struck me as being a reasonable approach and seemed to work well enough for me. I may have to pick up that book in your link. Thanks for posting it.
Though you may be missing my point. My point was never to debate the merit of hi rez, at least in this thread. I simply wanted to hear how you justify making authoritative statements which might be interpretted by readers as established fact as determined by more traditional data analysis technique. Since I am not aware of conclusive results as taken by more traditional methods, it seemed worth exploring.
Regarding the word "fact". Of course, there is no such thing as a real fact. However, a line must be drawn somewhere within reason if any work is to get done. The approach I'm most familiar with is establishing some limit of uncertainty and working within the limits.
BTW, why would you dawnsample in the first place? Memory is nearly free these days.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: