Testing audiophile claims and myths
71 dB
Headphoneus Supremus
There is no such thing as "passing a listening test." You do the test and then there's the result, the information. The result is valuable regardless of how "badly" you did in the test, because it tells you what matters and what doesn't matter. You know if the DAC matters or not. Being pressured and stressed by the test is silly.
No, not at all. When doing a test, an ABX for example, then you are listening far more intently, in a highly focused/concentrated manner. On the other hand, when listening to music in a relaxed state/casually, then you are listening far less intently and by definition, with less focus and concentration. However, this fact rather obviously contradicts your claim, unless you believe that more focus and concentration on listening for differences actually reduces our ability to identify/differentiate differences? But of course that would be a bizarre/ridiculous believe.
No! Clearly you don’t know what the point of an ABX test is or how it works! The point of an ABX test is to falsify a “null hypothesis”, which commonly (though not always) in audio would be: “There’s no audible difference”. So in a sense, there is a bias to succeed and falsify that null hypothesis, EG. To register an audible difference where in fact there wasn’t an audible difference. However, the fact you’re apparently missing is that while the goal, as you correctly stated, is to control bias, that does NOT necessarily mean it stops the test subject still suffering from bias, just that such a bias does NOT affect the results! In the case of a test subject believing they’ve heard an audible difference where there isn’t one, their answer must be either X = A or X = B and would on average be incorrect 50% of the time. This will have no effect on the results, the ONLY thing that affects the results is being far more (or far less) correct than 50%! Additionally, ABX testing does not and is not intended to reveal which “is actually better”, only to determine if a difference can be determined and of course there can be no bias towards the supposedly better DAC (or whatever else one is testing) because you do not know which is the supposedly better DAC. A bias towards the supposedly better DAC can only occur with a sighted test. Again, the opposite of what you are claiming!
No, yet again the opposite! In “an ideal world” you want the test subjects (“the person taking the test”) to know they are taking the test, you want them to focus/concentrate on identifying a difference and preferably with prior training so they know what difference to look for/concentrate on. Again, the point is to hopefully falsify the null hypothesis, by presenting the optimal conditions to discern a difference and thereby enable the results to be applicable to a wider range of people than just the test subjects. Also again, having “any stakes in the outcome of the test” is irrelevant because as explained above it will not affect the results, which is specifically why the ABX test was invented in the first place (as opposed to some other double blind test methodologies)! So “an ideal world” is achieved with or without “any stakes in the outcome”.
G
Last edited:
FunkyBassMan
500+ Head-Fier
@eq1849
You are right and gregorio is so very wrong.
There is reams of research that shows the unconscious perceives things that conscious awareness cannot. Among the many salient writings:
Marcel, A. J. (1983). Conscious and unconscious perception: An approach to the relations between phenomenal experience and perceptual processes. Cognitive Psychology, 15(2), 238-300. Proved that subliminally presented words can influence subsequent responses, indicating that the brain can process information without conscious awareness.
Weiskrantz, L., Warrington, E. K., Sanders, M. D., & Marshall,J. (1974). Visual capacity in the hemianopic field following a restricted occipital ablation. Brain, 97(1), 709-728. Blindsight patients detect and respond to visual stimuli without conscious visual awareness, demonstrating unconscious visual processing.
Dehaene, S., Naccache, L., Le Clec'H, G., Koechlin, E., Mueller,M., Dehaene-Lambertz, G., van de Moortele, P. F., & Le Bihan, D.(1998). Imaging unconscious semantic priming. Nature,395(6702), 597-600. Masked stimuli, invisible to conscious perception, can still influence reactions to subsequent stimuli, showing unconscious semantic processing.
Gazzaniga, M. S. (1967). The split brain in man. ScientificAmerican, 217(2), 24-29. Showed that each hemisphere in split-brain patients can process information independently, leading to actions without the conscious awareness of the other hemisphere.
Reber, A. S. (1967). Implicit learning of artificial grammars.Journal of Verbal Learning and Verbal Behavior, 6(6),855-863. Proved individuals can learn and retain complex information implicitly, without conscious awareness of the learning process.
Last edited:
I have no idea why you keep doing this to yourself, do you actually like repeatedly stating I’m very, ridiculously or laughably wrong, only for me to prove/demonstrate that you don’t have basic reading skills and/or a basic understanding of the facts?
In this case, you’ve presented a list of reliable evidence showing that we can subconsciously perceive things our conscious mind can’t but how does that demonstrate that my statement was wrong in any way whatsoever, let alone “very wrong”? Either quote where I stated that we can’t perceive anything subconsciously/subliminally or this is yet another in a long list of examples of you making a fool of yourself!
Actually your response is quite telling. If I’m apparently “very wrong” then you must obviously believe that if you concentrate/focus on something then you actually understand/identify/differentiate less than if you don’t. That would explain a lot, maybe if you did concentrate/focus on what I’m actually stating then you wouldn’t keep posting these ridiculous responses!
G
Last edited:
Couldn't you find any older research?
The Jester
1000+ Head-Fier
Plenty of that around here .. 
castleofargh
Sound Science Forum Moderator
- Joined
- Jul 2, 2011
- Posts
- 11,039
- Likes
- 6,927
Just because unconscious is written somewhere for something doesn't mean any belief related to unconscious hearing is validated by association. Unconscious isn't a word god to justify anything we want to be real without proof.
The upper treble is already failing to account for more than "air" in our experience of music. Without knowing anything else, how much could ultrasounds possibly do for us and music? I don't get that need to grasp at straws and blow the impact of ultrasounds out of proportion. Well, I don't for listeners. I fully understand why hires was marketed how it was. Who wouldn't want to invent more products out of the same product? It's kind of badass.
But then there is the part of the audiophile community that, let's remember, has since the first advertising campaign for a hires product, claimed unequivocally to clearly and very consciously hear the difference. And it still goes strong to this day. Some of them see those papers looking for brain activity with hires. The papers say there is no evidence of conscious impact, and mention some more failed listening tests to add to the pile. The clear conclusion should be that those audiophiles made and keep making false claims(they're not hearing it, or what they're hearing has a different cause). That is what research has kept pointing to.
And yet, we instead see those papers brought up from time to time by people acting like they just found total vindication of their claims. Amazing!
The upper treble is already failing to account for more than "air" in our experience of music. Without knowing anything else, how much could ultrasounds possibly do for us and music? I don't get that need to grasp at straws and blow the impact of ultrasounds out of proportion. Well, I don't for listeners. I fully understand why hires was marketed how it was. Who wouldn't want to invent more products out of the same product? It's kind of badass.
But then there is the part of the audiophile community that, let's remember, has since the first advertising campaign for a hires product, claimed unequivocally to clearly and very consciously hear the difference. And it still goes strong to this day. Some of them see those papers looking for brain activity with hires. The papers say there is no evidence of conscious impact, and mention some more failed listening tests to add to the pile. The clear conclusion should be that those audiophiles made and keep making false claims(they're not hearing it, or what they're hearing has a different cause). That is what research has kept pointing to.
And yet, we instead see those papers brought up from time to time by people acting like they just found total vindication of their claims. Amazing!
redrol
Headphoneus Supremus
My experience with A/B testing (not x) since im the one doing the tests is that it works best over a long period of time. After awhile I notice certain things in dacs and amps that I either like or dislike. For instance, I have a Motu MK5 interface with a Sabre run headphone out. Over time this output gives me the impression of a bit of graininess in the sound and also a harsh upper end. It also has poor stereo separation meaning it is more center focused than other devices. Finally I looked up the cross talk specs of the thing and it rates 40-50db which isn't super good.
One issue is that my impressions of what is better is only what I like more, not what is actually objectively better, so I am at peace with this kind of testing.
One issue is that my impressions of what is better is only what I like more, not what is actually objectively better, so I am at peace with this kind of testing.
old tech
500+ Head-Fier
Of course you can do both of them. Many of the subjects in the Meyer/Moran study engaged in a long term listening test comparing CD quality to hi res over some six months in their own time, in their own homes on their own stereos. The key difference with what you say is that it was always blind, i.e. the listener never knew if the source was A or B. As bigshot says, a short test is all you need to determine if their is a difference but a properly executed long term blind test may help to determine a real preference (if there is a difference).
bigshot
Headphoneus Supremus
I would tend to think that someone who is worried about sound they can't even prove they can hear has already perfected the sound they can clearly hear, but I doubt that's the case. A lot of audiophiles enjoy mental monkey slapping. Not everyone is looking for practical solutions to real world problems. I guess tilting at windmills is fun for some people, but I'm a busy 21st century person. I have more important things to address.
I’m not sure what “graininess” actually means in terms of audio properties. “Harshness” typically means too much mid or high-mid freqs. So along with the “poor stereo separation” that’s 3 “impressions”, one of which appears to correlate with the crosstalk specs (although 40dB or so is typically about the limit of detection for crosstalk) but from that single possible correlation you’ve concluded that your “impressions” are actually real observations, even though your impression of “harshness” does not correlate with the facts, as there are no boosted mid or high-mid freqs.
Unfortunately, this “jump to conclusions” approach is entirely typical of the audiophile community and is what so often puts it at odds with the scientific/engineering communities, who verify impressions and actually make real observations. It’s been proven/demonstrated to the point of “well established fact” that long term listening testing is far more fallible, prone to biases and unreliable than fast switching between short excepts, due to echoic memory.
Unfortunately though there’s more than that “one issue”, there’s the prior and far larger issue as detailed above. Namely, that “impressions” are not real observations until actually verified as such. Therefore, your impressions may not be either “what is actually objectively better” NOR “what you like more”! In other words, regardless of what is objectively better, your assertion apparently just accepts that your impressions are real observations of “what you like more”, which is fallacious. For example, I tend to like a bit more bass (which typically is not “actually objectively better”) and I may have the impression over a period of time that a particular DAC, headphone or whatever provides that little extra bass, therefore I like it more. However, that impression over time is entirely likely to be the result of expectation, confirmation or some other bias or simply fallible long term memory. Many audiophiles would respond with something along the lines of; “I don’t care, if my impression is that it has more bass and I like it more, that’s good enough for me”. The problem of course is that it’s based on a fallacy, it is very likely not true and sooner or later that impression will change, due to a new expectation or a different listening experience that raises doubts. This, IMHO, is the basis of “upgraditis” and the plague of the audiophile community, as audiophiles lurch from one unsubstantiated impression to the next and in the process become more and more detached from reality/the facts.
Of course it’s entirely up to you what kind of testing you are “at peace with” but personally I can’t be at peace with a kind of testing that is inherently fallacious.
G
Last edited:
Graininess is often associated with peaky treble.
I always called distorted treble, grainy. At least for me, treble start to sounds grainy if it distorts too much
71 dB
Headphoneus Supremus
I believe graininess of sound is mostly about the high frequencies being modulated by low(er) frequencies. Quantization error (no dither used) sounds grainy, because the quantization error is modulated by the signal and the high frequencies in the quantization error modulated by lower frequencies in the signal generate the "graininess." Steady 5 kHz sinusoidal harmonically distorted in a steady manner generates steady harmonics (10 kHz, 15 kHz, etc. depending on the distortion type). This makes the sound brighter rather than grainy. However, if the amplitude of the 5 kHz sinusoidal is modulated say by a 110 Hz sinusoidal, the harmonic components are also modulated by 110 Hz and this starts to sound grainy. Instead of 110 Hz sinusoidal, the modulating signal can be for example the bass frequencies in music including infrasonics!
Users who are viewing this thread
