Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 26, 2016 19:30:48 GMT -5
This thread is starting to become ever so slightly fatiguing for me. Are you sure about that or are you just becoming burned out from burn in? Do you think you could reliably distinguish your current level of fatigue versus how you felt after being totally exasperated in a different situation? Yeah Monku, I just remembered how fatigued I was after that long repetitive night at Candy's house many years ago. In the morning I really had a hard time getting up. I finally got up in time to go to work. I guess long extended listening sessions are not as hard fatiguing as the Candy Sessions. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 26, 2016 19:47:26 GMT -5
Thanks Yves for the kind feedback. So if I have the test subject come into the room in a non-casual manner, like rush in or even run into the room would that help in your opinion and cause less or no bias? Boom said he would feel fatigue (maybe due to stress, I'm thinking) or listening too long. My test would be relatively short and I have tried to make it as easy and relaxing as possible. Can you please make any other comments or suggestions how I could make the experience valid in your opinion? Is playing thru same DAC for a number of track listenings a trick? I only have him leave the room and come back in so he can't see me when I change DAC's. Perhaps he could just stay in the room and I could cover the DAC's with a sheet so he can't tell which one is on. Would that be better? The idea if course is not to have him aware of which one is in use. Do you think that is wrong? More feedback on my test would be much appreciated. You've been on to me from the beginning? Can you expound on that more please, as long as you don't charge me for your services. Thanks  BTW: My ears and brain are in fact securely connected! I had that verified by a joint combined appointment with an otolaryngologist and a neurologist. Oh yeah, I almost forgot, Hugo and I used to get together sometimes for a quick beer. The test subject should be able to remain seated during the switch between test samples to compare. Also, the switch itself should be specifically designed to add only a short pause between test samples, with no distracting clicks and no noticeable fade-in after the pause or anything, and no cutoffs that differ in such way that they might give away the identity of test samples. Ideally the duration of the pause between test samples should be kept constant until moving on to the next test in the series. It gets far more difficult than that, though. If someone hears a certain subtle detail due to the DAC being more resolving than another DAC, for example, the simple recollection of this subtle detail can cause the person to hear it whilst listening to the lesser DAC next. (Even, if this lesser DAC truly lacks the capability to audibly reveal this detail, like, after you have in secret fact made completely sure of that by purposefully removing this detail from the test sample through editing software for only that particular DAC). This is because our brain tends to fill in the blanks, and can do so pretty successfully (to the point where it can be exploited by codecs like mp3 which loses tons of information). Especially where the brain has access to a memory of the sound, the effect is often stronger than you might assume. So this filling in missing information can be responsible for "both DACs sound the same" because the listener perceives it that way despite that audible information is genuinely obfuscated by the least resolving DAC, but not the more resolving one. Granted, this sort of thing doesn't happen all of the time. But nevertheless it happens systematically and frequently enough for it to add significant bias towards "hearing no difference", as unfortunately it doesn't take very many wrong answers for a listener to fail the test, and, assuming you don't know which is the most resolving DAC of the two (because... well, probably part of the entire reason why you are running a test was to find out if there is such thing as a more resolving DAC of the two, anyway in the first place) you get a 50 percent risk each time of the above explained bias getting added on top of the human stress factor. Although this is only one specific example of how bias easily creeps in though the back door, it helps demonstrate why knowledge of statistical analysis and psychoacoustics is compulsory. Correct interpretation of test results is not easy. Designing a proper test is not easy, and neither is conducting such a test. It is all extremely time consuming and resource heavy, and getting a large enough group of candidates to show up for statistics analysis to yield conclusive output is a rough job according to those who have tried. Detection rates go up if test samples are kept short. That's why usually they are kept much shorter than 10 seconds, but listening happens on a lot of different time scales. Some sounds you will notice immediately within seconds, maybe less, whereas others only arrive after several minutes, or more. When we hear sounds our brain externalizes them as objects that are creating these sounds. As we continue to listen the human cortex builds a map of these objects over time, and the period of time over which such integration would happen is called the conscious present. This natural mechanism gets broken if you do rapid switching, albeit you can do blind listening tests over long periods of time, which hardly anyone does because of other priorities. Like, important questions that you can't always answer without somehow being in control of the parameters (as opposed to listening to one thing and then another thing without knowing what it is you are listening to). For the most part, DBT are useful only to determine when we stop hearing a distortion. You run the risk of missing those qualities that are the very reason we listen to music in the first place. Not wholly dissimilar to using a fierce spotlight searching for your set of lost keys when you already know you've lost them in another continent, or maybe, if you want to study a river, do you take out a bucket of water and stare at it on the shore? : Professor Yves, PhD (Statistical Analysis and Psychoacoustics) has laid out here the quintessential considerations and format for the correct way to conduct your shootout. Please in the future use these very simple rules in conducting your shootout. Too bad he completely ignored my comments and questions. Maybe it was too simple for his mind. Some of my Mensa friends thought it was quite legit and refreshingly simple and non-fatiguing! |
|
|
|
Post by garbulky on Dec 26, 2016 20:22:05 GMT -5
Not doing that either.
|
|
|
|
Post by Jim on Dec 26, 2016 20:27:00 GMT -5
|
|
|
|
Post by garbulky on Dec 26, 2016 20:52:48 GMT -5
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 26, 2016 21:03:34 GMT -5
This thread is now starting to make some sense! If Garbulky and the Boom followed the Yves procedures it would have taken them several weeks! |
|
|
|
Post by Jim on Dec 26, 2016 21:28:22 GMT -5
This thread is now starting to make some sense! If Garbulky and the Boom followed the Yves procedures it would have taken them several weeks! Years |
|
|
|
Post by unsound on Dec 27, 2016 2:12:39 GMT -5
Just out of curiosity, and I haven't read this entire thread, has a DBT ever been done when minor imperfections were deliberately introduced in the DACs? So, 2 versions of DC-1, with one of them having a small (and variable) amount of conversion imperfection (not sure if this is easily possible). At what percentage of difference between the 2 do trained (and non-trained) listeners reliably identify difference (preference is immaterial).
|
|
|
|
Post by yves on Dec 27, 2016 3:06:32 GMT -5
Attention Boomzilla and Garbulky: Professor Yves, PhD (Statistical Analysis and Psychoacoustics) has laid out here the quintessential considerations and format for the correct way to conduct your shootout. Please in the future use these very simple rules in conducting your shootout. Too bad he completely ignored my comments and questions. Maybe it was too simple for his mind. Some of my Mensa friends thought it was quite legit and refreshingly simple and non-fatiguing!  |
|
|
|
Post by sahmen on Dec 27, 2016 3:34:43 GMT -5
Just out of curiosity, and I haven't read this entire thread, has a DBT ever been done when minor imperfections were deliberately introduced in the DACs? So, 2 versions of DC-1, with one of them having a small (and variable) amount of conversion imperfection (not sure if this is easily possible). At what percentage of difference between the 2 do trained (and non-trained) listeners reliably identify difference (preference is immaterial). Well, this is not about either DBT or the DC-1, but you may find it somewhat relevant. It is about my Audio GD Reference 5.32 DAC. Barely a week ago, I had Audio GD install a DSP firmware upgrade on it, and I also upgraded the USB from a USB32 to an Amanero384 combo module. The firmware upgrade was done via my laptop by the manufacturer from China with the help of the Teamviewer 12 remote access software, and the manufacturer also supplied the parts and instructions for the Amanero384 combo upgrade, which I manually installed successfully. Both upgrades are supposed to bring substantial improvements to the table, and in a way they have, in that they've dramatically transformed the sound signature of the DAC by introducing slight, but notable, technical "alterations," such as a wider soundstage, and slightly better detail retrieval and resolution. They have also brought the highs and mids forward in a way that has caused the bass to sound slightly leaner and recessed than before, although it still goes as deep as it used to and sounds a little more controlled and textured than before... All these are supposed to be technical improvements, but my immediate reaction to the new sound has not been very favorable, because, notably, of a certain occasional treble glare that has crept into the sound field, and which was not there before. Also this DAC came with an R-2-R (as opposed delta sigma) topology, and is as a result supposed to sound more smooth at the top, organic, natural, and musical, without necessarily being less detailed or resolving than delta sigma DACs in general (if one is to believe certain "experts" on head-fi.org. However, the two upgrades seem to have robbed it of that R-2-R organic smoothness. In my own opinion, all these "improvements" have made some tracks I previously used to enjoy on the DAC less enjoyable, and sometimes, even intolerable, to the point where I am thinking I should re-run DIRAC on the computer to which the DAC is attached in order to EQ the sound to frequency levels that would be more pleasing to my ear than they are now. I have played some musical instruments before, and also sung in a couple of choirs, As a result, I am habitually inclined to separating out automatically and analyzing the sound of different individual instruments and voices in any pieces I listen to, although, beyond that, I have not had any kind of formal training in "listening." I can tell you that the differences between the sound signatures of this very same DAC, before and after the upgrades, are such that I might sell the upgraded one if rerunning DIRAC does not help to "correct" the new sound signature, and make it a lot more pleasing to my ears that it is now. I feel confident that if I had another (pre-upgrade) AGD Ref. 5.32 available, it would be relatively easy for me to tell it apart from the upgraded one, because of how tracks I have long enjoyed on the pre-upgrade version now sound on the upgraded version, especially when listening to the tracks with headphones (especially, with my Hifiman he-6)... I leave you to make of this little story what you will |
|
|
|
Post by yves on Dec 27, 2016 8:02:12 GMT -5
Just out of curiosity, and I haven't read this entire thread, has a DBT ever been done when minor imperfections were deliberately introduced in the DACs? So, 2 versions of DC-1, with one of them having a small (and variable) amount of conversion imperfection (not sure if this is easily possible). At what percentage of difference between the 2 do trained (and non-trained) listeners reliably identify difference (preference is immaterial). Well, this is not about either DBT or the DC-1, but you may find it somewhat relevant. It is about my Audio GD Reference 5.32 DAC. Barely a week ago, I had Audio GD install a DSP firmware upgrade on it, and I also upgraded the USB from a USB32 to an Amanero384 combo module. The firmware upgrade was done via my laptop by the manufacturer from China with the help of the Teamviewer 12 remote access software, and the manufacturer also supplied the parts and instructions for the Amanero384 combo upgrade, which I manually installed successfully. Both upgrades are supposed to bring substantial improvements to the table, and in a way they have, in that they've dramatically transformed the sound signature of the DAC by introducing slight, but notable, technical "alterations," such as a wider soundstage, and slightly better detail retrieval and resolution. They have also brought the highs and mids forward in a way that has caused the bass to sound slightly leaner and recessed than before, although it still goes as deep as it used to and sounds a little more controlled and textured than before... All these are supposed to be technical improvements, but my immediate reaction to the new sound has not been very favorable, because, notably, of a certain occasional treble glare that has crept into the sound field, and which was not there before. Also this DAC came with an R-2-R (as opposed delta sigma) topology, and is as a result supposed to sound more smooth at the top, organic, natural, and musical, without necessarily being less detailed or resolving than delta sigma DACs in general (if one is to believe certain "experts" on head-fi.org. However, the two upgrades seem to have robbed it of that R-2-R organic smoothness. In my own opinion, all these "improvements" have made some tracks I previously used to enjoy on the DAC less enjoyable, and sometimes, even intolerable, to the point where I am thinking I should re-run DIRAC on the computer to which the DAC is attached in order to EQ the sound to frequency levels that would be more pleasing to my ear than they are now. I have played some musical instruments before, and also sung in a couple of choirs, As a result, I am habitually inclined to separating out automatically and analyzing the sound of different individual instruments and voices in any pieces I listen to, although, beyond that, I have not had any kind of formal training in "listening." I can tell you that the differences between the sound signatures of this very same DAC, before and after the upgrades, are such that I might sell the upgraded one if rerunning DIRAC does not help to "correct" the new sound signature, and make it a lot more pleasing to my ears that it is now. I feel confident that if I had another (pre-upgrade) AGD Ref. 5.32 available, it would be relatively easy for me to tell it apart from the upgraded one, because of how tracks I have long enjoyed on the pre-upgrade version now sound on the upgraded version, especially when listening to the tracks with headphones (especially, with my Hifiman he-6)... I leave you to make of this little story what you will The R-2R organic smoothness just pales in comparison to Sigma Delta done right TBH. Real unbridled transparency from top to bottom is what I look for in a DAC, and, only sometimes, I warm up the 12AU7 to give it that little bit of lounge-ish atmosphere because I like to listen to Frank Sinatra that way. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 27, 2016 8:16:17 GMT -5
Sahmen, I did enjoy your little story. However, you were getting a little too technical for my fuzzy brain. I would though ask one question as you must know by now I'm usually somewhat skeptical especially of the many descriptive words in reviews to describe the sound one hears. Can you explain what you mean by "organic" sound as you use in your post. Thanks |
|
|
|
Post by Boomzilla on Dec 27, 2016 9:42:40 GMT -5
...What the hell is LMDB? LMDB = Level-Matched, Double-Blind ...But if you " could NOT differentiate between these DAC's";then why did you make such a long thread in your " shootout" in which you listed many differences? Maybe you should go back and re-read your own first post in the your own thread? The FACT that I heard differences (and described them, to the best of my ability) doesn't mean that those differences were big enough to identify in a LMDB test. I agree that, on the face of it, it's fair to argue that the differences were so small as to be negligible. I'll agree with that. But it DOES NOT mean that the differences did not exist. I heard them. And don't blame expectation bias either, because after the selection was played (and I expressed my opinion), I usually had to ASK garbulky which DAC we'd been listening to. Does this mean that I have "golden ears?" No it doesn't. Had anyone else been there who had normal hearing, they'd have heard the same things. The proper question is not why I heard differences, but rather why wouldn't such differences be identified in a LMDB test? Thus the speculation that started this thread. LMDB testing has a horrid history, starting (?) when J. Gordon Holt ran one and found that there was no audible difference between a mass market stereo receiver and some high end (Mark Levinson?) amplifier. His conclusion was that consumers should buy by the best watts-per-dollar ratio and nothing else. This doesn't condemn Mark Levinson products, but rather the test method. There were (and are) audible differences between a Best-Buy Sony AVR and a Mark Levinson. Most can hear them easily. So why did Holt's test find the opposite? The method was (and is) somehow flawed. |
|
|
|
Post by Jim on Dec 27, 2016 9:46:26 GMT -5
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 27, 2016 9:48:42 GMT -5
because you KNEW it was a different DAC...therefore you purposely LOOKED for differences...which would probably not be noticed if you had no idea it was a different DAC. ever notice when anyone gets a new piece of equipment they all of a sudden can hear things they never knew was there before? its because they are looking for it to sound different therefore listen more intently....
|
|
|
|
Post by Boomzilla on Dec 27, 2016 9:57:39 GMT -5
So the criticality with which one listens results in whether or not differences are heard? I can buy that. jim - Why do you irritate yourself by continuing to read this thread? And, come to think of it, J. Gordon Holt's amplifier test was NOT the first instance of double blind testing. Bell Telephone Labs ran tests in the 1930's or 1940's. They placed an orchestra behind a curtain at one city with microphone feeds directly to (at least two) other cities. In the other cities, speakers were placed behind the closed auditorium curtains, and the audience had to vote whether they were hearing a real orchestra or a reproduced one. The result of the test was that a three-speaker array (R, Center, L) was the "minimum cut set" that was sufficient to fool audiences. In other words, with a three speaker array behind the curtain, the majority of the audience was fooled into believing that there was a real orchestra playing. Kind of ruins the theory of stereo reproduction, doesn't it? |
|
|
|
Post by Jim on Dec 27, 2016 10:18:10 GMT -5
jim - Why do you irritate yourself by continuing to read this thread? I've been reading it just for comedic value, pretty much from the start. Sorry, did I hit a soft spot?  |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 27, 2016 10:26:57 GMT -5
...What the hell is LMDB? LMDB = Level-Matched, Double-Blind ...But if you " could NOT differentiate between these DAC's";then why did you make such a long thread in your " shootout" in which you listed many differences? Maybe you should go back and re-read your own first post in the your own thread? I agree that, on the face of it, it's fair to argue that the differences were so small as to be negligible. I'll agree with that. But it DOES NOT mean that the differences did not exist. I heard them. And don't blame expectation bias either, because after the selection was played (and I expressed my opinion), I usually had to ASK garbulky which DAC we'd been listening to. Does this mean that I have "golden ears?" No it doesn't. Had anyone else been there who had normal hearing, they'd have heard the same things. The proper question is not why I heard differences, but rather why wouldn't such differences be identified in a LMDB test? Thus the speculation that started this thread. LMDB testing has a horrid history, starting (?) when J. Gordon Holt ran one and found that there was no audible difference between a mass market stereo receiver and some high end (Mark Levinson?) amplifier. His conclusion was that consumers should buy by the best watts-per-dollar ratio and nothing else. This doesn't condemn Mark Levinson products, but rather the test method. There were (and are) audible differences between a Best-Buy Sony AVR and a Mark Levinson. Most can hear them easily. So why did Holt's test find the opposite? The method was (and is) somehow flawed. You have that bass ackwards. Your logic is what's faulty. What you should have said was: " The fact that I think I heard differences"..... I don't really care at this point about the LM/BD tests. I want you to try my Easy Nut test, but you keep avoiding it and you won't correctly try it because you fear/know it will invalidate your findings/hearing in your very sloppy shootout conclusions. Bmoney hit it right on the nose in his short succinct post above. "The proper question is not why I heard differences, but rather why wouldn't such differences be identified in a LMDB test?"
The answer of course is they wouldn't be identified because you imagined you heard the differences knowing which DAC's were being used . Using my test will prove you can't really hear any such differences and your circle logic won't let you BS your way out of the results this time. Poor horse!  |
|
|
|
Post by Boomzilla on Dec 27, 2016 10:36:48 GMT -5
OK Chuckie -
We'll agree to disagree. You contend that UNLESS a LMDB test identifies the differences, that they don't exist (except in the imagination of the listener). I contend that they do, and that it's not merely imagination.
The Chuckienut test is but a variation of LMDB, so nothing different there, thanks. I really don't care enough to expend the effort.
And with that, my participation in this thread is (finally) completed. I don't ever expect to convince you, and (it goes without saying) you haven't convinced me. Such is life.
That said, I still enjoy your posts greatly, and will continue to look forward to your (wonderful & twisted) sense of humor!
Happy New Year to you & yours!
Cordially - Boomzilla
|
|
|
|
Post by garbulky on Dec 27, 2016 10:45:03 GMT -5
So what he's trying to say B'zilla he knows that you are wrong. It's a fact! So you should spend more time validating his knowledge and preferrably involve somebody to help you with it. I was going to help you but I found the blindfolds a bit too kinky. So I decided to use a towel to cover it but then my wife told me to put on some pants. So finally I just used these in both of my ears.  After several hours (and a divorce) I found out he was right. I heard no difference between the two. I have now changed my audio reccomendations to ear plugs from now on. |
|
