|   DBT question | Thomas A
Apr 9, 2001 5:54 AM | | If the auditory "short-term" memory is so short (seconds or fractions of seconds), how can a DBT be regarded as a reliable test for differences? Isn't the mismatch negativity test a better measure for difference?
Just curious.
Thomas |
| once you... | gonefishin
Apr 9, 2001 8:18 AM | | ....when/if you establish a difference you must then STILL esablish a perference. Unless your goal is to find out what and if there is difference....instead of having a goal of finding your preferences.
enjoy the differences.....or not
;) |
| re: DBT question | Eyespy
Apr 9, 2001 10:13 AM | | When testing for the purposes of verifying if small potential sonic differences, as imparted via an altered audio output and conveyed as soundwaves transmitted through the atmosphere and coincident upon the peripheral detectors (ears), are indeed audible, than well constructed bias controlled double blind listening tests are the most reliable methodology available. |
| re: DBT question | Thomas A
Apr 10, 2001 2:12 AM | | The MMN should be an even more objective methodology in my mind. It correlates with perceptual accuracy and have the advantage that it is not dependent on various tasks and does not require attention by the tested persons (which is they can watch a silenced video while testing). It requires EEG equipment though.
Thomas |
| re: DBT question | mtrycrafts
Apr 9, 2001 11:50 AM | | >>>If the auditory "short-term" memory is so short (seconds or fractions of seconds), how can a DBT be regarded as a reliable test for differences?<<<<
It has been found by experimentation by experts that rapid switching of short duration listening, less than 30 seconds, and at times only single notes have been listened for, this is the only method and way of reliably identifying audible differences.
>>>> Isn't the mismatch negativity test a better measure for difference? <<<<
Please explain what you mean here. |
| re: DBT question | Thomas A
Apr 10, 2001 1:24 AM | | The mismatch negativity (MMN) is generated by the brain's automatic response to change in auditory stimulation. It's an objective measure and can be measured in the absence of attention and without any task requirement.
See
International Journal of Psychphysiology 37 (2000) 3-10
for a review.
Thomas |
| re: DBT question | mtrycrafts
Apr 10, 2001 9:50 PM | | Oh, yes, there was a Japanese author tried to show audibility of ultrasonic sound using this method.
It doesn't guarantee that you are hearing it, only that it registers an EEG pulse that you can measure. This can be generated by other sensory pickups besides your ears. Perhaps there have been tests showing that you can only have these EEG pulses if your ears are the only sensory detectors? Maybe plugging ones ears with earplugs and ear muffs?
I get a kick out of this working while not even trying to listen? Great. Now you can listen to music while watching your favorite movie and still register the music on EEG? What fun. I love it.
Has the Acoustics Society of America accepted this methodology for testing JNDs? Audible differences? Still in the experimental stage?
|
| re: DBT question | Thomas A
Apr 11, 2001 1:17 AM | | I think you have misunderstood the method. In the latest publication by the japanese group (Oohashi et al, 2000) the used EEG to measure alpha-EEG, a totally different method from MMN.
MMN is a much more reliable test which correlates with perceptual accuracy. The MMN only gives a signal when the sound suddenly changes in e.g. frequency, duration, or level. There is no resemblance between the japanese groups alpha-EEG method with the MMN method developed by Nääten et al. 1978.
Regards,
Thomas |
| re: DBT question | mtrycrafts
Apr 11, 2001 9:17 PM | | Then, if it is that accurate, which hardly seems to be if you can do a completely different task and still register, then perhaps this can be compared with the DBT results for comparison. How JND levels have they measured with this, in all three parameters?
By the way, That Japanese group is the one I was thinking of as I have read thei AES conference presentations from some time back. Not very impressive at all, nor does their lates work which you seem to think likewise? |
| re: DBT question | Thomas A
Apr 13, 2001 4:53 AM | | I have not been able to track down every reference, but as I understand this, MMN correlates with JND tests. Thus when the level is below JND there is no MMN response generated. The advantage of "non-attention" is that this can e.g. be done also in small children to register a hearing loss in early age. So the research points to that a difference in sound is automatically generated by the brain.
Only exception is that musicians appear to elicit a "better" MMN response than non-musicians, so a long term pre-auditory stimulation/learning may be important also for the MMN. This field appear to be contradictory though.
I've read the latest report from the Japanese group. I don't really believe this, and even if it was true, it would not matter since most speakers for home HIFI cannot reproduce "super-sonic" freqeuncies.
Thomas |
| re: DBT question | mtrycrafts
Apr 16, 2001 8:57 PM | | Thanks for your added input. I understand the need in children, less so in adults, especially when the attension aspect can negate anything that you will consciously register as sound, even though your brain may have picked up the impulses, somehow. This is also important why perceptual coders work so wel is that much of the information is masked and lost in reality that matters, although it may be a neuron impulse someplace that can be measured.
I gather that this type of testing may be up your alley:-) |
| I'm not so sure it is that short,but it is tricked very easily.. | Joe B
Apr 10, 2001 5:53 PM | | That is to say as soon as a new sound comes along , the old sound soon evaporates. With DBT 's You have a good starting point because too many people listen with thier eyes. In a perfect test, switching would be instant, and often. |
| I'm not so sure it is that short,but it is tricked very easily.. | Steve H.
Apr 10, 2001 6:57 PM | | And so you propose that everyone goes out and buys testing equipment to get that perfect test?
And since I've never spoke to you on these coveted DBTs that so many reference around here. Do you feel that testing someones hearing for hearing loss prior to any DBT would be essential? Because that is the major thing that was not done in these tests that they always reference. I could not back up the results no matter what the outcome was. Yet they speak so highly of these people. Maybe not directly, but they do quote word for word often. |
| I'm not so sure it is that short,but it is tricked very easily.. | Eyespy
Apr 10, 2001 8:28 PM | | "And so you propose that everyone goes out and buys testing equipment to get that perfect test?"
I'm sorry. Where was that proposal?
"Because that is the major thing that was not done in these tests that they always reference"
You don't seem equally concerned about the hearing acuity of the self-proclaimed golden ears, when they pronounce that they can hear the difference between wire insulations, or one additional foot of Monster cable, or amongst similarly gauged speaker wires of disparate price, or any other unsubstantiated claim of an audible difference. Yet when they claim they hear these things in a sighted listening, but their ability to do so vanishes moments later when they have to rely only on their ears, you find the fault is with the blinded listening, not with the original unfounded claim. How fascinating.
Now, nowhere is it said that someone has to perform DBTs in a home audio situation, for stating a personal preference. But when unsubstantiated and implausible, even silly, claims are asserted by these individuals based on unreliable sighted listening, and some faith-based false belief system is the basis of the assertions passed on as if they were factual, don't be surprised when such notions are justifiably challanged. It's really very simple: make a testable claim, generate a challange. State a personal preference, no challange is generated, or warranted. |
| I'm not so sure it is that short,but it is tricked very easily.. | dank
Apr 11, 2001 5:32 PM | | <<State a personal preference, no challange is generated, or warranted.>>
So how would you respond to this:
"I personally prefer my Sonic Granite speaker cable because it makes my system sound more open/neutral/dynamic/airy/pickyerword."
Personal preference or testable claim? |
| I'm not so sure it is that short,but it is tricked very easily.. | Eyespy
Apr 11, 2001 7:13 PM | | Testable claim. |
| I'm not so sure it is that short,but it is tricked very easily.. | dank
Apr 12, 2001 10:19 AM | | Ok. So if I understand correctly, sonic performance doesn't qualify as a preference. Preferences would be color (yours are lavendar, aren't they?), thickness, sexy connectors, and the like. Correct? |
| I'm not so sure it is that short,but it is tricked very easily.. | Eyespy
Apr 12, 2001 9:30 PM | | A preference which is not predicated on some testable condition is immune from any challanges for substantiation, as it should be. As soon as an assertion of some testable state is introduced, and especially when that assertion is no more than an unsupported claim for an implausible, impossible, and or silly notion, failing to question that assertion would be remiss. The more extraordinary the claim, the more extraordinary the required proof. If the proof is there, so be it, the body of knowledge advances. If not, that relegates the claim into a different status. |
| I'm not so sure it is that short,but it is tricked very easily.. | mtrycrafts
Apr 10, 2001 9:57 PM | | Souds like you are getting confused by the day around here. Don't try so hard. Don't need to justify a null result. But, you do need to justify the audible difference, right? Whom have you taken to task lately on this? I must have missed your challenge. Please enlighten all of us. Or, you may not like the answer? What, no audible differences? How come? How could all those thousand be wrong? Oh, that happens all the. |
| Besides | mtrycrafts
Apr 10, 2001 10:01 PM | | Those poor dissappointed 'golden ears' would have failed the test regardless. Or, they really do need a crutch to hear with their eyes, right?
But, I suppose, with a test you could have disqualified them beforehand and tell them they cannot hear well enough, right? Naw, you have nothing to stand on except some credible positive outcomes which you don't seem to have, right? Why? |
| I'm not so sure it is that short,but it is tricked very easily.. | Thomas A
Apr 11, 2001 1:27 AM | | I think the auditory memory trace has been estimated to be around 10 seconds (Cowan, N., 1984). This duration of the memory trace found by the MMN method also appears to be around 10 s (Böttcher-Gandor and Ullsperger, 1992; Sams et al 1993).
This means that switching of sound must occur within 10 s to generate an audible difference when small changes are to be evaluated. All "home-made" DBTs which include testing of cables + manual switching of cables etc are therefore useless.
Thomas |
| I'm not so sure it is that short,but it is tricked very easily.. | Eyespy
Apr 11, 2001 8:15 AM | | "All "home-made" DBTs which include testing of cables + manual switching of cables etc are therefore useless."
But not "as useless" as all home-made sighted listening comparisons, which are even "more useless". SOme of my personal favs for uselessness are the testimonials offered by golden ears who change some wire or make some other sonically inconsequential change, and the wife comes home later and can tell there is a difference in sound from what she heard the day before or last week, or an hour ago before going to the A&P. There are limitless examples of useless claims for audible differences, and you have pointed out one of the significant reasons that it is so. However, the golden ears have elevated the importance of such stories to "I heard it with my own ears"-like status. |
| I'm not so sure it is that short,but it is tricked very easily.. | Thomas A
Apr 11, 2001 9:26 AM | | Hi Eyespy,
I agree with that sighted listening is useless. I just want to point out to all people that want to try such things at home that we probably need some high quality switch box to do "home made" cable/equipment comparisons, and do the switches rapidly.
And also, our long term audio memory appear to be different from this short term one. It may play a role in how we percieve sound and detect differences. Thus a DBT performed at home may have different outcome than one performed at a test lab.
Regards,
Thomas |
| |
