1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Testing audiophile claims and myths

Discussion in 'Sound Science' started by prog rock man, May 3, 2010.
769 770 771 772 773 774 775 776 777 778
780 781 782 783 784 785 786 787 788 789
  1. old tech
    There you go again, making a claim based on your misuderstanding of statistics and your misinterpretation of results. This appears to be a continuation of your misinterpretation of the M&M study where you claim it proved some could hear a difference in normal listening environments, despite the flaws in your observations being clearly pointed out to you.

    I have yet to see one credible test where all other factors apart from 16/44 are controlled that supports your assertion of exceptions or even more absurd, that that they "fail so much of the time".

    Still flogging the dead horse? I like Emotiva products but peddling FUD on a sound science forum does you or your company you represent no favours.
    Last edited: Dec 19, 2018
  2. Killcomic
    Because no one EVER, in the entire history of the world, has overcharged for something. Cough.
    Also, it could be the same crap as magic juju oxygen-free kryptonite cables made of expensive materials. They make no difference but they cost more.
  3. analogsurviver
    Foobar2000 ABX is limited to comparing different PCM files - and can not test anything else, be it analog or DSD. That means I have next to none use for it.

    The only commercially available ABX box of sufficiently high quality that can do ABX testing properly - down to a piece of wire or a single electronic component, if required - is


    One could do with a less expensive home brew version - but no way yours or mine, even if built from the same schematics, could be thought of actually performing the same. The most troubles in audio are caused by switches and volume control in general - and with 0.2 dB or less volume matching - RELIABLY, REPEATEDLY - requirement, this IS tough in real life. You could realise in no time a potentiometer capable of required performance alone costs more than some receivers some claim to be audibly transparent.

    If I had the money, I would have gotten that box - despite being on the wrong continent regarding shipping/customs/voltage, all of which add in practice almost 50% on top of what buys one an unit in the US.
  4. Killcomic
    Oh for the love of... You'll first need to tell the difference between a high quality lossy vs lossless, only then do you even consider going higher.
    Everyone here thinks they got dog hearing.
    Last edited: Dec 19, 2018
  5. castleofargh Contributor
    something we'd be able to notice when it goes away, we'd notice with a switch. I get what you're suggesting, but what are the odds for something creating an impact to be at the same unnoticeable with rapid switching? it would need to have a really small variation compared to the other signal, and then I doubt the consequences would turn out to be meaningful anyway. or it would need to be something with fairly high energy but outside our hearing range(think the radiations of sound, strong but we just happen not to have sensors readily available for it). ultrasounds could probably be that if they happened to have some impact on our body in the long run while we're unaware of their presence. but then high energy inaudible stuff would most likely turn out to be bad for us.

    I don't think it's that easy to manufacture a situation where sound variations we don't notice with rapid switching turn out to have some significant impact anyway. at least I don't expect that to happen in the audible range.
  6. analogsurviver
    Sorry, do you know what PCM achronym REALLY stands for ?

    TBH, for my personal listening, I have stopped using it, thanks to a member banned from the head-fi ( and to my knowledge, not reinstated after last major change on head-fi - but I could be wrong on that one ), to whom bigshot played the role of the yes-man.

    The plot went around fooling the ear with (white) noise above 20 kHz added instead of the original ultrasonic content in the > 44.1k sampling files. I can go back to the exact files, if required.

    BOTTOM LINE : If a person is subject to PCM only diet for music listening, that DOES - unfortunately - de facto become the standard by which the sound quality gets to be judged.
    ANYTHING that stands out from this bandwidth limited corrupotion of the real thing gets to be considered - inferiour.

    I did fell prey to this devilish scheme. And from that time on, my CD players are gathering - dust. I dusted off some half a cm of dust off a player I will use for CD mat tests - to satisfy @bfreedma - sometime this summer. It was actually quite tedious to get it from storage - under piles of low interest gear . No prizes for guessing where it resides at the present ...

    It is live, analog record or DSD for me; even on the go. Even if it means strapping the brick of the external DSD capable DAC to the phone.

    I break this rule only when having to use sealed bluetooth headphones to demonstrate my recordings in noisy environments - and when some music I like is not available on anything else but PCM.

    And for "enjoying" comparing issued CDs I recorded to the DSD master.
  7. Killcomic
    Well then, we are looking forward to your subjective test results demonstrating conclusively that you can pick a DSD file over a 256kbps AAC file. Then, and only then, will I put a poster of you in my bedroom and wish upon the first evening star that I could be just like you.
    *Wipes away a hopeful tear*
    Last edited: Dec 19, 2018
  8. bigshot
    It's absurd. All I want is one DAC that someone has good reason to believe sounds clearly different. That shouldn't be difficult. All this stuff about tests not being up to their standards and show me tests to show that the tests of the tests are being tested correctly stuff is pure and simple argumentativeness being used to prevent people from seeing the nakedness of the emperor. I think some people will never be satisfied. They will be self justifying their absurd biases until the day they die.

    Here is the last time I quote you before I mute your pointless, useless, semantic, repetitive, circular and rambling posts...


    I know someone who failed to identify lossy from lossless and claims that.
    Last edited: Dec 19, 2018
  9. analogsurviver
    Will you send a decent ABX box over here for testing ? Two EQUAL DACs capable of native DSD ?

    IIRC, using jRiver it should be possible to simoultaneously playDSD and from this DSD derived AAC file via two "zones" - each "zone" using each own own DAC ( has to be two equal native DSD capable type DACs in this case ) - with the output of those two DACs DBT ABXed over a decent ( < 0.2dB level matched ) ABX box.

    I only have an older jRiver version installed on my desktop - because it is so damn "predatory" (OK, it IS possible to set it up - EXACTLY how do you want it to handle various types of files during the initial installation ) and will usurp any device settings for other software also dealing with DSD, I did not want to see it on my laptop, etc.

    I do not have two exactly same DACs at my disposal - but could muster iFi Micro iDSD and iFi Micro iDSD Black Label ( same thing as seen by the computer ) to see if I can get jRiver to simoultaneously play both DSD and from this DSD derived AAC files.

    That "from DSD derived AAC" involves some kind of format conversion... and these ARE known to be considerably different from each other. And, they ARE constantly evolving - what held true a year ago does in no way mean it is correct at the time of writing .
    Particularly not so for F2K - which went from almost zero to almost hero regarding DSD over the years... - the more recent improvements are regarding the filtering options.

    DSD is, like it or not, more Windows than Mac thing ... - and I do not have any direct DSD to AAC conversion software. DSD to WAV, then WAV to AAC - yes, but direct - no. At least not yet - because I do not have any real use for AAC yet. Most people still require MP3s as proof of the recording being technically OK - and MP3 at 48kHz and 320kbps has been used by me so far for this purpose. I know AAC is better, but if no one requires it and I personally do not listen to lossy , I did not investigate further as much as I perhaps should have.
  10. Killcomic
    You don't need to convert DSD to AAC directly. That's even better actually. Because if you can't tell the difference of a dual encoded lossy file and the DSD iriginal, well, that's solid evidence right there. However, if you can tell the difference consistentlg, that's a pretty good indication that there may be something to higher resolution media.
    After that, it's just a matter of narrowing down where audible differences begin.
  11. analogsurviver
    For starters, I would like - LOVE is the more appropriate word - DSD to PCM conversion software would be better in the first place.

    The annoying phase differences between the two channels that some ( most ?) conversions from DSD to PCM produce are audible in the extreme ... - and there ARE combinations of software/hardware that would do that with PCM, too.
  12. Killcomic
    Goodness, who knew the burden of proof was so burdensome?
    Well then, so in light of a lack of evidence, claims if audible superiority must be duscarded till actual proof in obtained. Otherwise yuo are chasing ghosts. Do they exist?
    Maybe, but till proof is obtained, it's best to go on as if they don't.
  13. analogsurviver
    What would satisfy you guys ? F2k ABX is NOT - and I REPEAT - NOT usable for anything but PCM. Which is poor, regardless of degrees of poorness, lossy or non lossy.

    What would be accepted that does not automatically record on a computer ? A video of switching the switch by the person sitting in another room , where listeners can not possibly see or otherwise communicate with the switching person? Another video recording absolutely no cable switching etc has been performed ? Another video taken of the listeners, troughout the whole test? All videos of course simoultaneous, with exact time/date info displayed ?
    Another video of the people who would after the test take the questionaries and arrange them into a table?

    Would THAT be enough ?
  14. Killcomic
    Well, repeatable results are a requirement. So far no evidence has been presented.
    What are you saying here though? We should believe there are audible differences because it's inconvenient and difficult to prove?
    There is no correlation between the two statements.
    Mate, if you believe you can hear a difference, that's fine. But when it comes to convincing others, proof is an absolute requirement.
  15. gregorio
    1. No I didn't, that's a deliberate misquote. "There's no number 8: Repeatedly misquote and/or misrepresent the actual facts/science to serve your own agenda." Why are you not able to stick to the list YOU posted?
    2. What piece of equipment doesn't a world class recording studio have? You say you're "not a studio engineer" yet you repeatedly make assertions about studios and studio engineers when you've actually no idea what you're talking about. I've already stated that every commercial studio I've ever been in has at least one measurement mic but you simply ignore the facts and make-up nonsense anyway. Where's that on your list?
    3. Or maybe they'll just tell the customer that dead, frozen pigs can fly .... but how would you know? Where on your list does it say that if you don't know anything, you're free to make-up any ridiculous suggestion about it? That's your "pure science" is it?
    4. Oh the irony! You spend pages and pages arguing about the importance of capturing (ultrasonic) room acoustics and then argue for pages for the use of a mic specifically designed to ONLY be used in an anechoic chamber (where there are NO room acoustics). The self-contradiction is staggering!! Where on your list was that?

    1. You would have zero evidence that "one subject was right 18/20", in fact they were probably NOT "right", it was probably just a statistical anomaly (as you admitted). So why make the assertion that the subject "was right", when even by your own admission you would NOT know if they were "right"? That error does qualify as a fallacy, is "misusing statistics" and is "confusing statistics with specific facts", so why do you keep doing it? For example:
    1b. No, those results were NOT "a sort of null result"! You seem to be equating a "very high probability" with a "possibility". Those results were not just "quite suggestive" but very "suggestive". The results do not absolutely rule out the possibility that someone could tell the difference but as the expected/predicted distribution curve of pure chance was in fact achieved, the results are therefore very "suggestive" that no one could tell a difference. It was NOT a null result but neither was it conclusive/absolute proof, just evidence with a good degree of confidence.
    1c. That's a change, an assertion that you've actually just admirably demonstrated yourself!
    2. But that is pretty much never the goal! A lab controlled DBT results in a statistical probability, therefore the "goal" is NOT to determine "absolute fact" but simply to provide reliable evidence!
    2a. No it doesn't! You claim to have studied testing at college but don't seem to know the purpose of a DB listening test or what an ABX test asks for?
    3. Firstly, there's a very small chance of that and Secondly, it's a risk you seem to be particularly good at avoiding anyway! Why don't you follow YOUR OWN list and actually try your "OVERLAPPING samples" analogy, test it with music and find out for yourself what actually happens?

    1. Nope, that's NOT what the tests/evidence shows.
    2. What exceptions? The exceptions of people comparing different masters, the exceptions of people doing tests at effectively dangerous levels or the exceptions which aren't exceptions because they fall entirely within the expected distribution curve of random guessing? Despite decades of trying, no one has so far found any exceptions, so how can we concentrate on them?
    3. Another fallacy! You are conflating the container with what it contains and worse still, you are already aware of this.

    Round and round and round we go!

769 770 771 772 773 774 775 776 777 778
780 781 782 783 784 785 786 787 788 789

Share This Page