AVS/AIX High-Resolution Audio Test: Ready, Set, Go! - Page 7 - AVS Forum | Home Theater Discussions And Reviews
Baselworld is only a few weeks away. Getting the latest news is easy, Click Here for info on how to join the Watchuseek.com newsletter list. Follow our team for updates featuring event coverage, new product unveilings, watch industry news & more!


Forum Jump: 
Reply
 
Thread Tools
post #181 of 222 Old 08-16-2014, 02:04 AM
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,530
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 851 Post(s)
Liked: 1214
Quote:
Originally Posted by SXRDork View Post
I'm an engineer and I believe in the truth of numbers and all that.

There are certain things I don't get about Nyquist, FRF's, Frequency response plots, and the rest that use 44.1kHz as adequate. One of them is the shape of a 20kHz waveform when sampled at 44.1kHz.
What worries you?

A CD player rendition of a 20 KHz sine wave is a thing of beauty. It is the right shape within a tiny fraction of a percent. It has the right amplitude within that same tiny fraction of a percent, and it has the right phase with a decent tolerance as well. It is audibly indistinguishable from the best 20 KHz sine waves that can exist in this world.

Quote:
I predict frequency response functions regularly as an aerospace engineer and I prefer to see at least 5 samples per resonant frequency just to see the waveform in some sort of recognizable shape. Nyquist seems to be the minimum required to mathematically sense frequency content, not correctly represent the waveform.

Can somebody school me on this? What am I missing?
What you're missing is that we don't listen with oscilloscopes or FFTs. We listen with ears. Ears are generally imprecise compared to good test equipment.
arnyk is offline  
Sponsored Links
Advertisement
 
post #182 of 222 Old 08-16-2014, 02:04 AM
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,530
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 851 Post(s)
Liked: 1214
Quote:
Originally Posted by Kele View Post
My amp/pre-amp/source is Stereophile class C. My speakers - vintage JBL reworked. Comparing CD vs SACD or DVD-A is easily heard. While playing MJ's Thriller SACD, my wife came into the room and commented about how good it sounds and she seemingly never cares about such things. I have played Thriller via vinyl, tape, CD and now SACD - the only unsolicited sound quality type comment from wife was with the SACD. The base difference is especially noticeable in that with SACD the base is more defined, digital-base impact, therefore has a higher startle factor. At times it actually sounds like a real drum kit in the adjoining room. The highs are more vivid yet not in a grating way, and ambiance with SACD is also unquestionably better than CD. I've only just recently heard SACD at home. I've had my same stereo components for a long while and know it - especially the speakers. With SACD, my stereo sounds much more capable than I thought. The difference is about like a 320 bit MP3 vs CD. I have no graphs to prove it.
Two words: sighted evaluation.
arnyk is offline  
post #183 of 222 Old 08-16-2014, 02:51 AM
AVS Special Member
 
m. zillch's Avatar
 
Join Date: Oct 2006
Posts: 4,795
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 531 Post(s)
Liked: 358
Quote:
Originally Posted by SXRDork View Post
Nyquist seems to be the minimum required to mathematically sense frequency content, not correctly represent the waveform.

Can somebody school me on this? What am I missing?

Watch Christopher "Monty" Montgomery's videos on reconstruction filters at xiph.org


http://people.xiph.org/~xiphmont/demo/neil-young.html
[This isn't the actual video I mean but should get you in the right ball park and has links]

In A/V reproduction accuracy, there is no concept of "accounting for taste". We don't "pick" the level of bass, etc., any more than we pick the ending of a play. High fidelity means an unmodified, neutral, exact copy (or "reproduction") of the original artist's tonal balance, timing, dynamics, etc..
m. zillch is offline  
post #184 of 222 Old 08-16-2014, 08:27 AM
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,825
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 1334 Post(s)
Liked: 735
Quote:
Originally Posted by antoniobiz1 View Post
(Emphasis mine).
I can't believe you post these results and still claim a significant difference. You were wrong three out of five times during the first part of the test. Going by what you "trained ears" keep saying an mp3 should pierce your eardrums as soon as it starts to play. This is ridiculous.
First, good morning to all . I hope we keep in mind that we are discussing a hobby and not any matters of life and death. So let's keep emotions out of these posts and just talk about the technical topic.

As to your point, your interpretation of what I post is not correct. Both theory and reason I post what I post is not understood.

A lossy audio compressor is constantly analyzing segments of audio (called frames) looking to see how much it can compress it while staying within its bit budget (320 kbps/sec in this instant). As such, the level of distortion varies constantly in sub-second intervals.

Let's look at an extreme but easier to understand example of 5 seconds of complete and total silence (digital "zero"). The entropy coder (back end of the codec) can compress this down to nothing. It wouldn't even need 64 kbps let alone 320 kbps. Ditto for a pure tone. Simple redundancy of data allows it to crunch such content with ease.

At the other extreme but difficult concept to understand are high frequency transients. There is no redundancy per se in a signal that all of a sudden sharply shoots up for a few milliseconds and then goes back down. The lossy encoder's frame now expands before and after the transients most likely. As it attempts to requantize (roughly speaking truncating frequency bands), it will spread that distortion to before and after the transient. What is after the transient will be masked usually. But not nearly as much for what comes before the transient. We call this type of distortion "pre-echo," i.e. echo that happens before the signal itself. As you can imagine, this type of distortion can be quite annoying depending on the amount of it.

What does this mean in this context? There is no such thing as hitting play and instantly being able to recognize the compressed version. Despite my training, there distortions that are inaudible to me as much as it is to you.

What I have shown in the test results is the full process from me starting the test to finding a critical segment that sounded different per above description. I am being transparent here showing you the failures until I found the difference. I could have of course restarted the test when I found the difference and show you perfect scores. But I thought there is education in learning how this type of testing works.

As to why occasionally I fail to tell the difference like that sample in the middle, again, remember this is not life and death discussions . I am listening to these files on my day to day laptop I am using for these posts. I am sitting in our living room while the TV is on and someone else watching it. We also have two dogs that have had a lot of practice barking . So it is not the ideal situation for critical listening.

Also, the way I do these tests is I play A and then B until I can classify in my mind which one sounded which way. As I run the tests, occasionally I confuse A for B because I don't play them again. I play X and Y and based on what I hear, I vote. The smaller the differences, the more I can make mistakes remembering what the difference was through so many trials.

Remember that the standard in such tests is 0.5% probability of chance. That is, we accept some amount of mistakes such as above. Outcomes with 0.5% error are considered proper and "scientific." So there is no requirement for perfect scores.

My results were: Total: 22/26 (0.0%). As you see, I achieved 0% probability of chance. Per above this exceeds the standard. What this means is that there should be no doubt that I found differences and could reliably tell the files apart. To call my results "rediculous" makes absolutely no sense. These are superlative results for double blind tests.

As I have said before, we are not at all used to seeing positive results in double blind test. As such our ability to analyze them is quite low as this discussions shows. Let's put aside emotions and allow the information to add to our audio knowledge and advance our learning. It takes a lot of work to keep answering these challenges (above was Krabapple's claim that I could not tell 320 kbps apart). This is rare kind of data/isngiht that we should celebrate having rather than showing such angst against it.
symphara, ChuckZ and granroth like this.
amirm is offline  
post #185 of 222 Old 08-16-2014, 08:39 AM
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,825
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 1334 Post(s)
Liked: 735
Quote:
Originally Posted by m. zillch View Post
Watch Christopher "Monty" Montgomery's videos on reconstruction filters at xiph.org

http://people.xiph.org/~xiphmont/demo/neil-young.html
[This isn't the actual video I mean but should get you in the right ball park and has links]
This has been discussed to death. See this thread for example: 24/192 Music Downloads and why they make no sense

Monty is a very smart guy and knows a lot about signal processing in the context of lossy audio codecs. But he is not a hardware engineer, nor understand signal processing concepts beyond his core area of expertise. Your foot doctor understands fair bit about cancer treatment but ultimately that is not his area of speciality. Likewise Monty's blog does have a lot of good information (and certainly superior to just about any post I read on forums arguing his point of view), but when it is all said and done, it has issues. We had a follow up to the AVS thread on WBF Forum where Monty himself participated. The discussions got a bit heated as he and I have some history together going back many years as I explain in that thread . But it is worth a read before running off with superficial understanding of the topic and what is written.

Importantly, the recent listening tests results which we are discussing have shown that his theories ultimately do not hold water. My machine had no intermodulation distortion for example so his theories of increased distortion with high resolution music files simply does not hold water.
amirm is offline  
post #186 of 222 Old 08-16-2014, 09:15 AM
AVS Addicted Member
 
NorthSky's Avatar
 
Join Date: Jan 2013
Location: Canada - West Island: Vancouver, South Direction: Go East, dts:x Films: Under the Skin, 3D, Birdman, X_Machina, The Zero Theorem, Interstellar
Posts: 11,656
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 4902 Post(s)
Liked: 2026
Amir, you prefer the AIX hi-res over the CD hi-res?

And two, you prefer the audience positioning over the in the middle-of-the-band one?
NorthSky is offline  
post #187 of 222 Old 08-16-2014, 09:51 AM
Member
 
antoniobiz1's Avatar
 
Join Date: Sep 2007
Posts: 145
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 73 Post(s)
Liked: 80
Quote:
Originally Posted by amirm View Post
First, good morning to all . I hope we keep in mind that we are discussing a hobby and not any matters of life and death. So let's keep emotions out of these posts and just talk about the technical topic.
You're right, manners first. I don't say good morning because I am in Italy. It's 17.45 now, and most of you American guys start replying when it's afternoon here. This is no excuse, though. So good morning to you

No emotions, here, just a civilized discussion (I hope).
Quote:
Originally Posted by amirm View Post
As to your point, your interpretation of what I post is not correct. Both theory and reason I post what I post is not understood.
(Theory explanation removed)
What does this mean in this context? There is no such thing as hitting play and instantly being able to recognize the compressed version. Despite my training, there distortions that are inaudible to me as much as it is to you.
Theory is completely irrelevant, here. The question is just: "Can you hear it?". Rest doesn't matter, at all. And now, for the first time, you are admitting that there's no such thing as hitting play and hearing it. I don't think I'm too far from the truth comparing this to an admission that you would have failed Mayer and Moran. You're saying that if you just play music (which one third of this forum is all about, the other two thirds being video and videogames), you won't hear a difference.

Quote:
Originally Posted by amirm View Post
As to why occasionally I fail to tell the difference like that sample in the middle, again, remember this is not life and death discussions . I am listening to these files on my day to day laptop I am using for these posts. I am sitting in our living room while the TV is on and someone else watching it. We also have two dogs that have had a lot of practice barking . So it is not the ideal situation for critical listening.
You keep repeating this, and I find it kind of offensive. What are you implying? That you don't need to bring your "A game", here? That you're so good that barking dogs and tvs are not enough to prevent you from getting results? I'll mind your business for a second. You're here, day after day, writing lenghty post, replying to anyone who speaks to you. You obviously have time to dedicate to this hobby. Can't you find 15 minutes to properly take the tests?

Quote:
Originally Posted by amirm View Post
Also, the way I do these tests is I play A and then B until I can classify in my mind which one sounded which way. As I run the tests, occasionally I confuse A for B because I don't play them again. I play X and Y and based on what I hear, I vote. The smaller the differences, the more I can make mistakes remembering what the difference was through so many trials.
What has memory to do with the size of differences? One keeps playing A and B until one knows.
Quote:
Originally Posted by amirm View Post
Remember that the standard in such tests is 0.5% probability of chance. That is, we accept some amount of mistakes such as above. Outcomes with 0.5% error are considered proper and "scientific." So there is no requirement for perfect scores.
No perfect scores means you can kinda sorta hear it, maybe, sometimes, playing a 0.9 seconds segment 200 times. Hardly a great headline for an audio magazine or HDTracks or a 2000$ Sony HD player. And this is a hobby, as you said. Why are we switching to strict science all of a sudden?
Quote:
Originally Posted by amirm View Post
My results were: Total: 22/26 (0.0%). As you see, I achieved 0% probability of chance. Per above this exceeds the standard. What this means is that there should be no doubt that I found differences and could reliably tell the files apart. To call my results "rediculous" makes absolutely no sense. These are superlative results for double blind tests.
I said that you missed three times out of five the first part of the test. Again, that means that you, 0.9 seconds segments notwithstanding, were lost in the dark. And this is music you listened to a lot of times, before, because of all the other tests. And high quality music, too. There are a million hires proponents, claiming that compressed audio is pure unlistenable crap, and so far you emerged as the best ears in the world, so far. And you missed three out of five. Please, Amir, be serious. And I didn't misspell "ridiculous".

Quote:
Originally Posted by amirm View Post
As I have said before, we are not at all used to seeing positive results in double blind test. As such our ability to analyze them is quite low as this discussions shows. Let's put aside emotions and allow the information to add to our audio knowledge and advance our learning. It takes a lot of work to keep answering these challenges (above was Krabapple's claim that I could not tell 320 kbps apart). This is rare kind of data/isngiht that we should celebrate having rather than showing such angst against it.
What I gather from your results is this: you can't listen to music in your home without being afraid that someone slipped some mp3s in your collection because you wouldn't know. If this is the best world can offer, well, I really find no reason to celebrate.

No angst whatsoever on my part, too. I am disappointed, however. If this is the best world can offer, this is a real, tangible proof that mp3s ARE NOT crap as most audiophiles seem to think, there is NOTHING to gain from Hires audio, and everything about it is a giant marketing scam. Not that I didn't know, but I try to approach things with an open mind, and so far I always doubted it was me and my ears. Apparently it's not.

Anyway, we are discussing things. Nothing personal!
antoniobiz1 is offline  
post #188 of 222 Old 08-16-2014, 11:02 AM
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,825
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 1334 Post(s)
Liked: 735
Quote:
Originally Posted by antoniobiz1 View Post
Theory is completely irrelevant, here. The question is just: "Can you hear it?". Rest doesn't matter, at all.
Of course it does. If a car has a ton of features meant for better handling, you can't say you expect that to show up when you are going straight at 30 kilometers/hour.

The nature of distortions in digital systems is complex. Without understanding them, you simply will not perform a proper test to find them. We use specialized clips and trained listeners to find artifacts routinely in the industry. The notion that any part of a music track picked at random and hear a difference simply doesn't hold up in what we do in real life outside of forums.

This is again the bible of the industry on how to do these tests, from your neck of the woods, International Telecommunication Union (ITU) Recommendation BS1116,

When selecting the programme material, it is important that the attributes which are to be assessed are precisely defined. The responsibility of selecting material shall be delegated to a group of skilled subjects with a basic knowledge of the impairments to be expected.

The clips in this thread were not picked to be revealing based on the distortions that might be there. That I could tell the files apart when I or anyone else with the knowledge of the distortions had nothing to do with the selection of material shows that there is more difference here than we like to accept.

Quote:
And now, for the first time, you are admitting that there's no such thing as hitting play and hearing it. I don't think I'm too far from the truth comparing this to an admission that you would have failed Mayer and Moran. You're saying that if you just play music (which one third of this forum is all about, the other two thirds being video and videogames), you won't hear a difference.
Say that again? You are extrapolating my ability to be able to tell the files in this thread apart, to not being able to tell the difference in another test? What logic are you using for that???

That aside, there were a ton of mistakes in how Meyer and Moran tests were conducted. I have listed them countless times but it seems we love to ignore flaws in tests whose outcome we like.

We can start with the point above. Where do I read in the Meyer and Moran that they had expertise in this matter and used that to select the right results?

Where would I read that they did this based on BS1116: For all test sequences, therefore, the group of skilled subjects shall convene and come to a consensus on the relative sound levels of the individual test excerpts.


Where would I read that they did this based on BS1116 (notes to subjects):The purpose of the training phase is to allow listeners to identify and become familiar with potential distortions and artefacts produced by the systems under test. After training, you should know “what to listen for”. This afternoon, you will be asked to blind grade all the audio.

Where would I read that they did this based on BS1116: Post-screening methods can be roughly separated into at least two classes; one is based on inconsistencies compared with the mean result and another relies on the ability of the subject to make correct identifications.

They used the results of all testers, did they not?

Where would I read that they did this based on BS1116:On the other hand, if these anchors [controls] fail such correct identification by any listeners, then this suggests that either these listeners lacked sufficient expertise, or else that there were sensitivity flaws in the situation, or both. In that case, the apparent transparency of systems cannot be properly interpreted, and the experiment will need to be run again with new listeners to replace the ones who failed this additional test, and with any other changes that may increase experimental sensitivity.

Their testing lacked any kind of control let alone being used to throw out results of the testers who could not pass them. And you see the part about throwing out the results of the controls are not detected by anyone?

Where would I read that they tested the material to see if they even had high frequency content? Or that they were truly > 16 bits of resolution?

Where would I read that they enabled this ability as recommended in BS1116:
The pre-selection of suitable test excerpts for the critical evaluation of the performance of reference two-channel downmix should be based on the reproduction of two-channel down-mixed programme material.

See? You are clearly allowed to pick subsections of tracks. Indeed MPEG reference audio test clips are all short, roughly 30 seconds each. There is no requirement nor is that good practice to force people to listen to the whole track and tell the difference in the whole track.

The purpose of such testing in the industry is not to make fool of a group of people has Meyer and Moran DIY intended or its similar use in these forums. We do these listening tests because we are truly interested in finding any and all flaws to fix them. We don't put our head in the sand and ignore every proper protocol for a proper test as to fool ourselves into thinking there is no difference.

Quote:
You keep repeating this, and I find it kind of offensive. What are you implying? That you don't need to bring your "A game", here? That you're so good that barking dogs and tvs are not enough to prevent you from getting results? I'll mind your business for a second. You're here, day after day, writing lenghty post, replying to anyone who speaks to you. You obviously have time to dedicate to this hobby. Can't you find 15 minutes to properly take the tests?
This must be a cultural thing as I can't for the life of me figure out why me telling you my test conditions being casual is offensive. No, I don't bring my "A game" here, nor was it necessary to do so. I told the files apart down to 0% probability of chance using a stock laptop and set of headphones/in-ear monitors. Why would I go and lock myself in a dedicated room and perform the tests??? They results can't get better than 0% probably of chance.

Remember, you challenged me on a single miss in the middle of correct results and I explained that this could happen. And that it would even happen in best testing scenarios because we allow 0.5% probability of error. Subjective listening tests are not foolproof. There is no reason to take offense.

Just because I have time to post these things, doesn't mean I enjoy doing audio tests. Vast majority of vocal members here would who write incredible missives and thousands of posts refused to run and report these tests. Why not go after them to bring their "A game" and run more tests? I have done my thing for the kingdom .

Last edited by amirm; 08-16-2014 at 11:09 AM. Reason: Can't write this morning :).
amirm is offline  
post #189 of 222 Old 08-16-2014, 01:37 PM
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,825
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 1334 Post(s)
Liked: 735
Quote:
Originally Posted by antoniobiz1 View Post
No perfect scores means you can kinda sorta hear it....
No, it doesn't mean that. What I heard i heard with 100% certainty or the probability of guessing wouldn't be zero.

If you mean I can select a random segment and immediately be able to tell I can't do that. And I explained why. Once again, digital systems have dynamic distortion, not static. The performance is variable and can change from 0.1 second to another 0.1 second. Fact that some or even most of a track is transparent is neither here, nor there. If we can identify any part of the track that is not transparent, then the argument that 320 kbps MP3 is as good as the original is falsified.

Quote:
, maybe, sometimes, playing a 0.9 seconds segment 200 times. Hardly a great headline for an audio magazine or HDTracks or a 2000$ Sony HD player. And this is a hobby, as you said. Why are we switching to strict science all of a sudden?
Because that is what is required when differences get small. I don't make the rules. I follow best practices and can't put it aside as to let someone win a verbal argument online.

Quote:
I said that you missed three times out of five the first part of the test. Again, that means that you, 0.9 seconds segments notwithstanding, were lost in the dark. And this is music you listened to a lot of times, before, because of all the other tests. And high quality music, too. There are a million hires proponents, claiming that compressed audio is pure unlistenable crap, and so far you emerged as the best ears in the world, so far. And you missed three out of five. Please, Amir, be serious. And I didn't misspell "ridiculous".
You didn't misspell that but you are asking me to throw away all the science and engineering behind this field and then follow you along. I can't do that. I know how lossy audio compression works and as a result, I know that it can generate transparent audio and audible artifacts at the same time, in the same track. I have not said that the entire track will be degraded.

I also have not said it sounds like "crap." Whoever those millions of people are you need to go and argue with them, and not ask me to defend them. To my ears, when I hear compression artifacts, it is super, super annoying. More annoying than a bad speaker. More annoying than even FM radio. It is like nails scratching on a blackboard. It is not to millions of listeners and that is fine. We are not talking about them here.

Quote:
What I gather from your results is this: you can't listen to music in your home without being afraid that someone slipped some mp3s in your collection because you wouldn't know. If this is the best world can offer, well, I really find no reason to celebrate.
No, what you should gather is that if the server dropped the food on the floor and picked it back up and put it on a plate, I would not eat it . I want food that has not gone through that "experience."

I don't know why we have so much trouble with this simple logic: I want the music as it was originally produced and mastered. I don't want it resampled to 16 bits. I don't want it resampled to 44.1 Khz. And I certainly don't want it converted to 320 kbps MP3. I can do any of these conversions myself as can countless other audiophiles.

There is no way you can tell me that having the distributor or talent do the conversion is good for me. If you did, it is the same as saying that you will eat the food that had fallen on the floor just the same as one that has not. If so, then your standard is different than mine. I am not here to tell you what to do. I am here to tell you my standards for audio fidelity and they are admittedly high. I want to know I can get the best version there is.

Quote:
No angst whatsoever on my part, too. I am disappointed, however. If this is the best world can offer, this is a real, tangible proof that mp3s ARE NOT crap as most audiophiles seem to think, there is NOTHING to gain from Hires audio, and everything about it is a giant marketing scam. Not that I didn't know, but I try to approach things with an open mind, and so far I always doubted it was me and my ears. Apparently it's not.

Anyway, we are discussing things. Nothing personal!
I appreciate that. But please appreciate that you are confusing me with a different crowd. I am an engineer and use proper listening tests to form my opinions. I have not praised high resolution audio for the sake of high resolution audio. I have repeatedly said what I have said here in red. No one, no one has come up with a single argument against that. Yet the arguments go on and on and on. What do you guys get out of spending time writing these posts? Why is it so important to push the results of these tests under the rug? To what end? To convince distributors to only give us compressed music???
amirm is offline  
post #190 of 222 Old 08-16-2014, 08:38 PM
Member
 
SXRDork's Avatar
 
Join Date: Feb 2007
Location: Broomfield, CO
Posts: 123
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 28
Quote:
Originally Posted by arnyk View Post
What worries you?

A CD player rendition of a 20 KHz sine wave is a thing of beauty. It is the right shape within a tiny fraction of a percent. It has the right amplitude within that same tiny fraction of a percent, and it has the right phase with a decent tolerance as well. It is audibly indistinguishable from the best 20 KHz sine waves that can exist in this world.



What you're missing is that we don't listen with oscilloscopes or FFTs. We listen with ears. Ears are generally imprecise compared to good test equipment.
What I was missing was the implementation of the low-pass filter and the subsequent interpolation of the signal. Basically, a quick read of the Nyquist-Shannon sampling theorem (Wikipedia) cleared things up for me. I wasn't considering those two operations and that is why I was so confused.

SXRDork: SXRD is a technology. My name is Vann. Usernames are for life...
SXRDork is offline  
post #191 of 222 Old 08-17-2014, 02:09 AM
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,530
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 851 Post(s)
Liked: 1214
Quote:
Originally Posted by antoniobiz1 View Post

Theory is completely irrelevant, here. The question is just: "Can you hear it?". Rest doesn't matter, at all. And now, for the first time, you are admitting that there's no such thing as hitting play and hearing it. I don't think I'm too far from the truth comparing this to an admission that you would have failed Mayer and Moran. You're saying that if you just play music (which one third of this forum is all about, the other two thirds being video and videogames), you won't hear a difference.
Good question, as the saying goes. ;-)

Long ago I realized that ABX can be an overly sensitive test in the sense that seems to be illustrated above.

ABX can make a big thing out of differences that the casual, one-time or infrequent listener won't notice.

Frankly, the market that ABX was designed for is people who obsess about perfection in audio. That would characterize most if not all of the 6 people in the original ABX development team.

There is a saying - "An ear worm" that describes a sound that gets trapped in the mind of many of us and for some time we will hear it in our minds again and again.

http://en.wikipedia.org/wiki/Earworm

"An earworm, sometimes known as a brainworm,[1] is a catchy piece of music that continually repeats through a person's mind after it is no longer playing.[2] Phrases used to describe an earworm include musical imagery repetition, involuntary musical imagery, and stuck song syndrome.[1][3][4] The word earworm is a calque from the German Ohrwurm[5] and was, according to Oliver Sacks, first used in the 1980s.[1]"

According to research by James Kellaris, 98% of individuals experience earworms. Women and men experience the phenomenon equally often, but earworms tend to last longer for women and irritate them more.[14] Kellaris produced statistics suggesting that songs with lyrics may account for 73.7% of earworms, whereas instrumental music may cause only 7.7%.[15]

...

In a 2006 book by Daniel Levitin entitled This Is Your Brain on Music: The Science of a Human Obsession, he states that research has shown musicians and people with obsessive–compulsive disorder (OCD) are more likely to suffer from earworm attacks. An attack usually involves a small portion of a song, a hook, equal to or less than the capacity of one's auditory short-term memory. Levitin reports that capacity as usually 15 to 30 seconds. Simple tunes are more likely to get stuck than complex pieces of music. He also mentions that in some situations, OCD medications have been known to minimize the effects.[8] In 2010, published data in the British Journal of Psychology directly addressed the subject, and its results support earlier claims that earworms are usually 15 to 30 seconds in length.[10]

...

Scientists at Western Washington University found that engaging the working memory in moderately difficult tasks (such as anagrams, Sudoku puzzles, or reading a novel) was an effective way of stopping earworms and of reducing their recurrence.[16][17] Another publication points out that melodic music has a tendency to demonstrate repeating rhythm which may lead to endless repetition, unless a climax can be achieved to break the cycle.[18]

...

Jean Harris, who murdered Dr. Herman Tarnower, was obsessed by the song "Put the Blame on Mame", which she first heard in the film Gilda. She would recall this regularly for over 33 years and could hold a conversation while playing it in her mind.[19]
"

So don't underestimate ear worms, they may have even gotten some people killed (just raising questions about those who say that this is not an issue of life and death). ;-)

The point is that once a momentary distortion is identified to the individual, it may become a kind of a ear worm.

Furthermore IME once a person learns how to hear an audible artifact, it often becomes far more audible. I might be suggesting that earworms have a borderline state where individuals become hypersensitive to certain sounds or distortions of sounds. They may hear them at lower levels and they may be more distressed by them than usual.

For example I spent about 12 years doing live sound in a poorly designed room that was prone to acoustic feedback. With all available technology in play, the choice was often whether a performer would be loud enough to be clearly heard but feedback was a probable risk, or whether the performer would not be clearly heard at all. People who work in those kind of environments say that they can hear feedback building long before it becomes clearly audible and if they know what to do about it, it will never be heard by almost all of the audience.

And, in that context, the hypersensitivity of ABX would appear to have some value, even great value.

It is especially true that if the distortion or artifact can be completely eliminated or vastly reduced at a reasonable cost, this begs the question, why not just do it?
arnyk is offline  
post #192 of 222 Old 08-17-2014, 07:27 AM
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,825
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 1334 Post(s)
Liked: 735
Quote:
Originally Posted by arnyk View Post
It is especially true that if the distortion or artifact can be completely eliminated or vastly reduced at a reasonable cost, this begs the question, why not just do it?
Yes, yes, yes! Why not indeed? We can get the original masters prior to conversion to 16/44.1 with even less than "reasonable cost." That track already exists so it doesn't need to be produced. Indeed it takes work to create the 16/44.1. The bandwidth costs are negligible compared to the royalties of the track. And storage of said file locally costs next to nothing. And just about any playback hardware can play these files.

So let's use this opportunity to come together as Arny has and embrace the concept of getting original high resolution masters delivered to us before the CD starts to die off and we are stuck with MP3s.
stereoeditor likes this.
amirm is offline  
post #193 of 222 Old 08-17-2014, 10:17 AM
AVS Special Member
 
m. zillch's Avatar
 
Join Date: Oct 2006
Posts: 4,795
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 531 Post(s)
Liked: 358
Quote:
Originally Posted by arnyk View Post
ABX can make a big thing out of differences that the casual, one-time or infrequent listener won't notice...
And, in that context, the hypersensitivity of ABX would appear to have some value, even great value.
I'm curious Arny, under your definition of "gaming" the system [I literally had to look up that use of the word "game", as a verb], when I repeated the short section of the song Mosaic and could pick up on that lower pitched "CHA" sound vs. a higher pitched "CHIP" sound at the guitar riff's conclusion [proving the ~10 millisecond delay problem has audible consequences], is that "gaming" or cheating the ABX system in your opinion?

Question to all: Other than a passing "I think I hear it" from Tube, nobody else has provided any indication to me they hear what I hear. Was all my effort to learn how to use Audacity, select that segment, post the two songs' rendition of it, post my winning ABX log, and make a video on it hoping to point out to others what to listen for, all for nothing?

Last edited by m. zillch; 08-17-2014 at 11:42 AM.
m. zillch is offline  
post #194 of 222 Old 08-17-2014, 11:50 AM
Member
 
rongon's Avatar
 
Join Date: Jul 2013
Posts: 44
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 11
No symphonic music?

I admit I didn't give a close look or a search, but why no symphonic music? I would think that is where you'd really hear a difference. There's nothing as difficult to capture with a microphone as massed strings, and big Post-Romantic orchestral music (Rimsky-Korsakov comes to mind) certainly has wider dynamic swings than just about all pop music.

Just curious. I wish I had the time to devote to doing this test. Will try to figure out how to do it...

--

Hopelessly addicted to this stuff.
rongon is offline  
post #195 of 222 Old 08-17-2014, 12:12 PM
isa
Member
 
Join Date: Jun 2014
Posts: 37
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 33 Post(s)
Liked: 18
Quote:
Originally Posted by arnyk View Post
It is especially true that if the distortion or artifact can be completely eliminated or vastly reduced at a reasonable cost, this begs the question, why not just do it?
Quote:
Originally Posted by amirm View Post
Yes, yes, yes! Why not indeed? We can get the original masters prior to conversion to 16/44.1 with even less than "reasonable cost." That track already exists so it doesn't need to be produced. Indeed it takes work to create the 16/44.1. The bandwidth costs are negligible compared to the royalties of the track. And storage of said file locally costs next to nothing. And just about any playback hardware can play these files.

So let's use this opportunity to come together as Arny has and embrace the concept of getting original high resolution masters delivered to us before the CD starts to die off and we are stuck with MP3s.
I also heartily agree, if the common concept is to advocate delivery of 24/96k uncompressed lossless audio to avoid the foibles of downsampling as we've admired to death here and other quality-degrading processing. Disk space is cheap. 24/96k DACs are cheap. To more specifically describe what I mean by "uncompressed", I mean music that has not gone through the dynamic range-crushing process of the Loudness Wars that is ubiquitous in non-classical music. To state yet another way, if we can leverage 24/96k to defeat the Loudness Wars, that would be huge!

Additional suggestions to make it happen: Mark Waldrep has created a trademark/brand/label he calls "Ultra HD-Audio", and I regard Waldrep as truly devoted to high quality audio. What if, as a part of that brand, he defined a production certification standard to meet that definition that included aspects vital to high quality music, like pure digital floating point processing during production, no DSD/PCM translations during production (or define a way to allow that without quality loss: I'm not an expert in this area), no dynamic range compression (or no such compression beyond certain defined limits to ensure high quality and preserve artistic intent that are beyond my expertise to define). And then Mark and his supporters enlisted other high quality music producers to adopt and certify their production standards to this brand, and mark and market their music accordingly.

With my business experience, I see three major potential outcomes: 1) no one cares and this goes nowhere 2)everyone sees the benefit and opportunity to slap the label on all their music regardless of actual quality, making the entire effort useless at best 3)it actually works and helps raise the bar for consumer audio knowledge, expectations and high quality available music. I think it will take close to a miracle to avoid #1 or #2 and have #3 happen, but wouldn't it be more fun and fulfilling to redirect the energy of the participants of this forum into trying for #3 instead of "debating"?

Perhaps a key area of opportunity is timely: Pono has announced their music provider is Omnifone. Omnifone has apparently and magically declared they have millions of high rez tracks. As observed in many places, this can be best described as "baloney". What if Pono were convinced it was in their best interests to subject Omnifone to some kind of certification process to validate that their tracks labelled hi rez actually were? While I'm deeply cynical regarding the true versus claimed motives and aspirations of Pono, I could see how Pono would not want to be harmed by crappy hi rez songs and find a way to "turn the screws" on Omnifone to actually deliver high quality music. My issues in previous posts aside, I believe Amir and many others here are truly devoted to high quality music, and a combo of Amir, Mark, other individuals and audio websites with the highest web presence like avsforum, the parent company of stereophile, audioholics, hometheatershack, head-fi and others approaching Neil Young on promoting a brand that represents truly high quality music and certifying Omnifone's library against that brand seems very interesting to me...

The another big dog to approach is of course Apple, and also approaching them would seem important. But if rebuffed, there's always Google and Amazon trying to take a byte out of Apple and thus may be better positioned to jump on this.

Firms I've not considered to approach are the major labels. While I think their power in making this happen or defeating it greatly exceeds Pono, Apple or anyone else, I also think they could care less or would feel threatened by the effort, so they seem best viewed as reluctant followers or an obstruction rather than a facilitator. At best I expect major labels to create a competing brand that is comical and for which all their music qualifies to specifically defeat the intent of delivering high quality music to consumers. That may seem paranoid, but who 30 years ago predicted that the advent of digital audio would be used to ruin dynamics in popular music?

This topic is perhaps best served in a new thread, but I defer to others to decide that. In any case, I would rather spend my time finding a way to leverage 24/96k in a manner to defeat the Loudness Wars than continue to "debate".

P.S. There is also a much lower priority technical nit for people like me with audio systems that have significant distortion issues in reproducing audio above about 22k, so in addition to the above it strikes me that something should be done at the point of playback to make it easy for people to test their systems for 24/96k playback and have an easy option to lowpass such music to avoid problems, but I can only admire the problem for now and have no good suggestions to easily deal with it. Perhaps we conclude that 24/96k music with realistic content above 22k does not pose enough of a threat to equipment or ears to warrant anything on such matters, but I think it should be looked at closely by people more informed than me to see if anything is needed to avoid harming equipment or producing really bad sound in consumers' systems that would undermine the overall effort.
isa is offline  
post #196 of 222 Old 08-17-2014, 03:05 PM
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,530
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 851 Post(s)
Liked: 1214
Quote:
Originally Posted by m. zillch View Post
I'm curious Arny, under your definition of "gaming" the system [I literally had to look up that use of the word "game", as a verb], when I repeated the short section of the song Mosaic and could pick up on that lower pitched "CHA" sound vs. a higher pitched "CHIP" sound at the guitar riff's conclusion [proving the ~10 millisecond delay problem has audible consequences], is that "gaming" or cheating the ABX system in your opinion?

First off I'd like to thank especially you and also everybody else who contributed results, for the exceptional quality effort that you put into these two projects related to high resolution audio. It has definitely been a learning experience for me and your efforts have contributed greatly to it in a positive way.

IMO these tests were gamed both intentionally and unintentionally. What surprised me was how easy that happened.

The purpose of both tests related to the audibility of high resolution audio, and I'm taking the position that any positive results that were observed that were not due to differences in sample length or sample rate constituted gaming.

IMO what's described above was gaming because the positive results were not due to the object of the experiment. the gaming of the experiment was unintentional because the time shift that was eventually diagnosed was clearly an accident.
arnyk is offline  
post #197 of 222 Old 08-17-2014, 04:31 PM
AVS Special Member
 
RichB's Avatar
 
Join Date: Nov 1999
Location: Massachusetts
Posts: 9,079
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 442 Post(s)
Liked: 312
Quote:
Originally Posted by isa View Post
...
To state yet another way, if we can leverage 24/96k to defeat the Loudness Wars, that would be huge!
That is not enough. I have HDTracks 96/24 downloads with DR ratings of 5 and 6.

Quote:
Originally Posted by isa View Post
... but who 30 years ago predicted that the advent of digital audio would be used to ruin dynamics in popular music?

This topic is perhaps best served in a new thread, but I defer to others to decide that. In any case, I would rather spend my time finding a way to leverage 24/96k in a manner to defeat the Loudness Wars than continue to "debate".
Sad but true. I have recording from the late 80's that sound great with DR ratings from 13 to 17.
That is hard to find now. But what is the purpose of HD audio with the same crappy masters.
There is no good quality metric and the HD sites don't have buyer feedback.


I have HDTracks 96/24 downloads with DR ratings of 5 and 6.

Quote:
Originally Posted by isa View Post
P.S. There is also a much lower priority technical nit for people like me with audio systems that have significant distortion issues in reproducing audio above about 22k, so in addition to the above it strikes me that something should be done at the point of playback to make it easy for people to test their systems for 24/96k playback and have an easy option to lowpass such music to avoid problems, but I can only admire the problem for now and have no good suggestions to easily deal with it. Perhaps we conclude that 24/96k music with realistic content above 22k does not pose enough of a threat to equipment or ears to warrant anything on such matters, but I think it should be looked at closely by people more informed than me to see if anything is needed to avoid harming equipment or producing really bad sound in consumers' systems that would undermine the overall effort.

I cannot see the advantage in delivering context above 22K, it might not do harm but it could.


- Rich

Oppo Beta Group

Oppo BDP-105D | Oppo HA-1 | Oppo PM-1 | Parasound A21 + A21 + A51 | Revel Salon, Voice, Studio | Velodyne HGS-15
For Sale: Revel Salons and Revel Studios (together or separately)
RichB is online now  
post #198 of 222 Old 08-17-2014, 07:09 PM
AVS Special Member
 
m. zillch's Avatar
 
Join Date: Oct 2006
Posts: 4,795
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 531 Post(s)
Liked: 358
Quote:
Originally Posted by arnyk View Post
First off I'd like to thank especially you and also everybody else who contributed results, for the exceptional quality effort that you put into these two projects related to high resolution audio. It has definitely been a learning experience for me and your efforts have contributed greatly to it in a positive way..
Thanks.


Quote:
Originally Posted by arnyk View Post
IMO these tests were gamed both intentionally and unintentionally. What surprised me was how easy that happened.

The purpose of both tests related to the audibility of high resolution audio, and I'm taking the position that any positive results that were observed that were not due to differences in sample length or sample rate constituted gaming.

IMO what's described above was gaming because the positive results were not due to the object of the experiment. the gaming of the experiment was unintentional because the time shift that was eventually diagnosed was clearly an accident...

Like I said, I'm new to the use of the word game as a verb, but I assume you mean to cheat, or take advantage of dishonestly:


GAME
intransitive verb
: to play for a stake

transitive verb
1
archaic : to lose or squander by gambling

2
: to take dishonest advantage of : cheat <game the tax system


http://www.merriam-webster.com/dictionary/game


Since I was completely upfront from the get go that the audible difference I heard was due to problems with the test material's presentation [a ~10 mS delay on one version of Mosaic I was the first one to hear, consciously at least, post ABX scores documenting my accuracy, and then later taught myself how to use Audacity to measure it], I wouldn't use the word "game", exactly, but since I am new to its exact usage, I won't press the matter.
m. zillch is offline  
post #199 of 222 Old 08-17-2014, 08:36 PM
Member
 
Kele's Avatar
 
Join Date: Jun 2011
Location: Aloha
Posts: 28
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
Yes arnyk, on my part it's not a blind-test. That was why I added about wife... it is blind when she has no idea that SACD even exists let alone what's on play. Unprompted, she throws out a rare comment "that sounds real good" while cutting through to the laundry room. I'm surprised how this thread is going - I skimmed through some. I cannot tell CD from SACD files on my PC mini-speakers/base-module, but with big stereo, it's just so easy to hear the tighter base, etc., it doesn't seem necessary to perform serious AB comparisons. This is why I threw in the part about Stereophile class C. Not killer resolving, but enough to appreciate the higher bit-rate. I'm not sure I could always tell 24bit/ 88khz from 192khz... with this I'd say it affects the cleaner silence with higher freq sampling. Wow, the source makes heap big importance to better realism. PS: I can hear differences between some cables too ; )
Kele is offline  
post #200 of 222 Old 08-17-2014, 08:56 PM
AVS Special Member
 
m. zillch's Avatar
 
Join Date: Oct 2006
Posts: 4,795
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 531 Post(s)
Liked: 358
Quote:
Originally Posted by Kele View Post
Yes arnyk, on my part it's not a blind-test. That was why I added about wife... it is blind when she has no idea that SACD even exists let alone what's on play. Unprompted, she throws out a rare comment "that sounds real good" while cutting through to the laundry room.
And then all of the sudden, one day...http://www.iabmed.com/images/WifeWalkOut.jpg
m. zillch is offline  
post #201 of 222 Old 08-18-2014, 07:06 AM
Member
 
Kees de Visser's Avatar
 
Join Date: Jan 2009
Posts: 42
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 19
Quote:
Originally Posted by arnyk View Post
the gaming of the experiment was unintentional because the time shift that was eventually diagnosed was clearly an accident.
I did warn Scott though within 24h after publishing the first test that there were level and sync differences between the files. The level difference was corrected, but not the sync problem. Perhaps I didn't explain the issue clearly enough to him. It could be a lack of experience with fast-switching ABX tests, or (very unlikely) an intentional attempt to provoke false positives. That's a pity since the positive ABX results seem to indicate that the Sonic Process which was used for SRC and 16 bit dithering is not audibly transparent.
Are there any plans for a follow-up test with less (or at least controlled) variables, e.g. only 24 to 16 bit or only SRC ?
Kees de Visser is offline  
post #202 of 222 Old 08-18-2014, 12:26 PM
AVS Special Member
 
m. zillch's Avatar
 
Join Date: Oct 2006
Posts: 4,795
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 531 Post(s)
Liked: 358
Quote:
Originally Posted by Kees de Visser View Post
I did warn Scott though within 24h after publishing the first test that there were level and sync differences between the files.
Maybe he read that:
Quote:
Originally Posted by Kees de Visser View Post
Scott, I noticed two problems with the testfiles:

1) The pairs are not sample accurately aligned. This is essential for seamless ABX switching...

...as meaning just "a sample or two", and felt that you were overreacting to what he assumed was an inconsequential, almost certainly inaudible difference, whereas it turns out to be more like a ~ 1000 samples misalignment, at least in the case of Mosaic A2 vs B2, which I've documented I can successfully ABX due to that very misalignment, so it most certainly is audible.

Last edited by m. zillch; 08-18-2014 at 12:32 PM.
m. zillch is offline  
post #203 of 222 Old 08-18-2014, 01:01 PM
isa
Member
 
Join Date: Jun 2014
Posts: 37
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 33 Post(s)
Liked: 18
Quote:
Originally Posted by Kees de Visser View Post
I did warn Scott though within 24h after publishing the first test that there were level and sync differences between the files. The level difference was corrected, but not the sync problem. Perhaps I didn't explain the issue clearly enough to him. It could be a lack of experience with fast-switching ABX tests, or (very unlikely) an intentional attempt to provoke false positives. That's a pity since the positive ABX results seem to indicate that the Sonic Process which was used for SRC and 16 bit dithering is not audibly transparent.
Are there any plans for a follow-up test with less (or at least controlled) variables, e.g. only 24 to 16 bit or only SRC ?
Don't be too hard on yourself. After your rather clear post, I repeatedly mentioned that the songs had an average of 90 sample differences in addition to level issues, Andyc56 quickly produced resampling results that had neither the level nor sample alignment issue using SoX, and yet Scott still anchored on declaring that the Sonic tool and settings were SOTA and would not budge in using a better tool or settings despite obvious evidence otherwise.

As incompetent as Mark and Scott showed themselves in basic resampling in this test, I think they are well above the average competency level of those typically engaged in mixing and mastering, so avoiding the need for resampling expertise and just distribute native 24/96k would eliminate almost all need to understand resampling. That may sound cynical, but the refusal to go with SoX or a similarly competent resampler after errors were clearly noted speaks for itself.
isa is offline  
post #204 of 222 Old 08-18-2014, 01:35 PM
AVS Special Member
 
RichB's Avatar
 
Join Date: Nov 1999
Location: Massachusetts
Posts: 9,079
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 442 Post(s)
Liked: 312
^^^
If the software has errors, the operators should not be called "incompetent".


- Rich

Oppo Beta Group

Oppo BDP-105D | Oppo HA-1 | Oppo PM-1 | Parasound A21 + A21 + A51 | Revel Salon, Voice, Studio | Velodyne HGS-15
For Sale: Revel Salons and Revel Studios (together or separately)
RichB is online now  
post #205 of 222 Old 08-18-2014, 02:26 PM
Member
 
Kees de Visser's Avatar
 
Join Date: Jan 2009
Posts: 42
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 19
Quote:
Originally Posted by RichB View Post
If the software has errors, the operators should not be called "incompetent".
I partly agree. It is however the responsibility (competence if you like) of the operators to verify that the software they use doesn't produce errors. If it does, the errors should be fixed or they should use alternative software that performs better.
NB: it has not been established that the software is to blame. Amir has tried another SRC (Audition) which was not transparent either. It is not clear if the 96 to 44.1 kHz conversion, the 24 to 16 bit reduction or the playback system (distortion) is responsible for the reported audible difference.
Kees de Visser is offline  
post #206 of 222 Old 08-18-2014, 07:43 PM
Senior Member
 
WayneJoy's Avatar
 
Join Date: Feb 2013
Posts: 230
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 94 Post(s)
Liked: 63
maybe a comparison between 96khz and 48khz would have been better.
WayneJoy is online now  
post #207 of 222 Old 08-23-2014, 07:47 AM
Newbie
 
snake2002's Avatar
 
Join Date: May 2013
Posts: 1
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 0
Im quite happy because I was able to "solve" the test straight away on the first run which means:
i've good ears
and a good system altough a very modest one (but I live happy with it)
For those who say there' s no noticeable difference if I may a couple of suggestions:
- volume pretty high (not too much)
- first 30 seconds are enough, if you can't hear the difference at the start then you wont be able either later
- dont concentrate on the notes, instead close your eyes and let the music come to you
once you reach this state of mind the difference become more obvious
Of course if you're not carefully listening (say you're browsing the net) then it is almost impossible
because the overall picture impact is really close but..
the hd version have
on the street more energy
mosaic shakers more natural and "alive"
just my imagination is just more musical, a vinil kind of way
the cd version at first seems the be the same but its a colder presentation
like looking at 2 same pictures and one is slightly out of focus
sorry for the english (italy..) take care!
Andrea

my system
laptop with tons of tweaks applied for maximum music enjoyment
HRT II+ dac
old pioneer SA-710 with some tweaks applied
twisted OCF cables and no brand signal ones
blacknoise filters and a dedicated power line
snake2002 is offline  
post #208 of 222 Old 08-28-2014, 12:04 PM
Newbie
 
Join Date: Aug 2014
Posts: 3
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 0
A confession

I "gamed" these tests. I started with the level-corrected versions of the first musical selection. I used PC ABX. Using the start / end user interface, I created random short segments and listened to their start. After about ten tries creating segments, I came up with a segment which sounded audibly different. After memorizing the sound difference between A and B, I went straight through each X and the corresponding Y, picking the best match from my memory. It was clear that I could go on all day this way, until I made a stupid mistake and clicked the wrong button by mistake. (Pilots sometimes land their airplanes after forgetting to put the landing gear down.) There would be no point to posting results. IMO, it is likely that any "time alignment" of the sample files would not eliminate this gaming. So long as the waveform shapes are different (as they will be with sampling rate conversions of files with ultrasonic material) there will be startup transients that will make the ultrasonic material audible, e.g. by down modulating higher frequency energy into the audible band.

If you can "game" the casino you win. In gambling that's the way it is. If we put airs on and imagine our forum games are "science" then a different standard applies: we can blame the experimenters, in this case the people responsible for creating test files and suggesting test setups.

I will not try any more tests like these until suitable tools come along that enforce a real difference between "gaming" and "cheating". PC ABX fails in this regard. I used the user interface as instructed and got results. Being familiar with digital editing and its potential audible artifacts the situation I was already well trained to hear the artifacts created by the test equipment and recognize them for what they were. Others with less hands on experience may have fooled themselves into believing what they heard were differences in the actual files and not deficiencies in the test equipment.

As to listening to seven minute tracks to music that I don't particularly care for, suffice it to say that my attention span, concentration and aural memory are not up to snuff in the absence of substantial motivation. If I am going to critically listen to music, it's going to be music that I am remastering or restoring that others can enjoy. Incidentally, I no longer obsess with selecting the "best" hi-res to 44/16 conversion parameter choices, although I believe these are sometimes fairly easy to hear. My solution, if asked, would simply be to give the hi-res masters to any listeners who wanted higher resolution material. Suffice it to say, when I listen for my own pleasure I listen to the original hi-res. No eating food off the floor, no matter how clean...
Tony Lauck is offline  
post #209 of 222 Old 08-28-2014, 01:27 PM
AVS Special Member
 
m. zillch's Avatar
 
Join Date: Oct 2006
Posts: 4,795
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 531 Post(s)
Liked: 358
Quote:
Originally Posted by Tony Lauck View Post
IMO, it is likely that any "time alignment" of the sample files would not eliminate this gaming. So long as the waveform shapes are different (as they will be with sampling rate conversions of files with ultrasonic material) there will be startup transients that will make the ultrasonic material audible, e.g. by down modulating higher frequency energy into the audible band..
Have you looked at the waveform closely in the Mosaic A2 vs B2 guitar riff, right at the very end of the segment I posted (and that I can positively ABX)? That isn't a "transient", a pop, nor a click due to buttons being hit or deficiencies with the ABX software itself, as some have been lead to believe, it's part of the music. The extra 10 milliseconds of music tacked onto B2, because the files aren't properly time aligned, makes both the sound and the visible waveform of the fast, aggressive guitar strum, occurring right at the end of that segement, slightly different.

Listen to the full length version of Mosaic A2 (or B2), at that 1m52s mark, and you will hear the exact, aggressive guitar strum I isolated. The difference isn't an ABX software pop, click, thump, transient, hiccup, deficiency, etc., it is the correctly reproduced, guitar strum sound itself which is slightly different due to the added 10 milliseconds, or so, of waveform motion. There's at least one extra wave cycle (a crest and a trough) added to B2 that A2 lacks, and you can hear it. At least I can.


There's no problem with the ABX software here, folks. It is showing us the truth, without any artifacts.

Click on "full screen" in the lower right of my Youtube video below to see a magnified image so the extra wave cycle tacked onto the end of B2 (that A2 lacks due to time axis misalignment) is more easily seen:

Last edited by m. zillch; 08-28-2014 at 09:55 PM.
m. zillch is offline  
post #210 of 222 Old 08-29-2014, 01:06 AM
AVS Addicted Member
 
NorthSky's Avatar
 
Join Date: Jan 2013
Location: Canada - West Island: Vancouver, South Direction: Go East, dts:x Films: Under the Skin, 3D, Birdman, X_Machina, The Zero Theorem, Interstellar
Posts: 11,656
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 4902 Post(s)
Liked: 2026
Quote:
Originally Posted by Tony Lauck View Post
I "gamed" these tests. I started with the level-corrected versions of the first musical selection. I used PC ABX. Using the start / end user interface, I created random short segments and listened to their start. After about ten tries creating segments, I came up with a segment which sounded audibly different. After memorizing the sound difference between A and B, I went straight through each X and the corresponding Y, picking the best match from my memory. It was clear that I could go on all day this way, until I made a stupid mistake and clicked the wrong button by mistake. (Pilots sometimes land their airplanes after forgetting to put the landing gear down.) There would be no point to posting results. IMO, it is likely that any "time alignment" of the sample files would not eliminate this gaming. So long as the waveform shapes are different (as they will be with sampling rate conversions of files with ultrasonic material) there will be startup transients that will make the ultrasonic material audible, e.g. by down modulating higher frequency energy into the audible band.

If you can "game" the casino you win. In gambling that's the way it is. If we put airs on and imagine our forum games are "science" then a different standard applies: we can blame the experimenters, in this case the people responsible for creating test files and suggesting test setups.

I will not try any more tests like these until suitable tools come along that enforce a real difference between "gaming" and "cheating". PC ABX fails in this regard. I used the user interface as instructed and got results. Being familiar with digital editing and its potential audible artifacts the situation I was already well trained to hear the artifacts created by the test equipment and recognize them for what they were. Others with less hands on experience may have fooled themselves into believing what they heard were differences in the actual files and not deficiencies in the test equipment.

As to listening to seven minute tracks to music that I don't particularly care for, suffice it to say that my attention span, concentration and aural memory are not up to snuff in the absence of substantial motivation. If I am going to critically listen to music, it's going to be music that I am remastering or restoring that others can enjoy. Incidentally, I no longer obsess with selecting the "best" hi-res to 44/16 conversion parameter choices, although I believe these are sometimes fairly easy to hear. My solution, if asked, would simply be to give the hi-res masters to any listeners who wanted higher resolution material. Suffice it to say, when I listen for my own pleasure I listen to the original hi-res. No eating food off the floor, no matter how clean...
Fair enough Tony.
NorthSky is offline  
Reply Audio Theory, Setup, and Chat

Tags
frontpage

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off