tired hearing

General bodily fatigue often reduces one's ability to concentrate and thus increases the probability of missing details. What I want to know about is something a bit different. I'm sure it must have been researched in the laboratory and mapped in some detail, but so have a million other things. Not being versed in that literature, I don't know where to start.

I'm reasonably certain people have been tested to determine how well humans can discriminate various kinds of auditory details. For instance, it is no doubt know that, in general people can tell the difference between frequencies if they are yea-so many Hz apart (no doubt depending somewhat on where in the frequency spectrum we are measuring) or the difference between the same tones at differing loudness, etc. etc..

What I am wondering is whether or not there is any know special fatigue factor for hearing. If, for instance, it has been established that a subject can tell the difference between two equal loudness tones that are 10Hz apart (or 100Hz or 0.1Hz, whatever is real) and if the experiment keeps presenting trials for the person to judge, does the ability to make the discrimination decrease after some fairly short time (2 minutes, 20 minutes, ??) regardless of whether the overall state of the person is good, alert, energetic, etc.?

Any information is appreciated. What I'm mainly looking for is that such a factor is know to exist or not know to exist, and if know to exist, what are the time factors.

Yet again the kind of absolute quantification your question requires does not, and can not exist.  You could look at protocols for bioengineering threshold testing, and if you deconstruct that basic information you will see that 'fatigue' is a primary concern of any threshold testing. 

Those protocols are readily available online.  The MIT open course material might be a good place to start.

If you want more then, "Yes 'fatigue' is a significant auditory function and because it is system (the pressure wave(s), subject, environment) specific the kind of quantification for which you are asking is not possible.", then entry level anatomy & physiology of mammalian auditory function would be next step.  Even high school AP info would be sufficient.

You would then need to follow that with an entry level psychoacoustics reference/seminar, etc.

Then if you could find some of the OSHA benchmark testing . . . but you kind of need more then the raw data here and it's that, that's hard to find.  As with all the OSHA stuff there is creative and fluid compromise between health and economic concerns.  But some of the studies OSHA earmarked, I think in the 70's, is pretty interesting. 

this might be a place to start on some OSHA data

(here's the link (unparsed): http://www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=FEDERAL_REGISTER&p_id=14206 in case the 'this' doesn't translate effectively, which for reasons I've never quite figured out it seldom does, on this forum)

What do you mean by "absolute quantification"? I'm sure that if such a factor exist, it will be quantified by normal statistical analysis: e.g. 97.3% of the population fall in the range of X minutes to Y minutes with everyone else further out on the curve,  probably a bell curve.

I chose those two particular examples of basic auditory discrimination because I'm sure such measurements have been done a number of time, starting quite a few many years ago. What I don't know is if a "fatigue" or "desensitization" factor exists, something that would effect such "threshold" discrimination, but not be a significant factor under "normal " conditions (obviously, people can use their hearing all day under "normal" circumstances and it still continues to function just fine).

If it does exist, and it is shorter than say an hour, it would probably have been investigated and measured as a matter of course. If not already measured, it would probably be noticed from other testing data. In the latter case, someone would then do the experiments. Grad students are always looking for easily defined projects.

That OSHA reference, to some hearings held years ago, is probably at least six layers of bureaucratic paper pushing above any subject matter mention of what the hearings were actually about. Why do you think OSHA might be interested in such a thing (the subject matter of my inquiry)? It is highly unlikely to be a health or safety related factor (not that OSHA isn't quite often very intrusive and authoritarian).

If it does exist, and it is shorter than say an hour, it would probably have been investigated and measured as a matter of course. If not already measured, it would probably be noticed from other testing data. In the latter case, someone would then do the experiments. Grad students are always looking for easily defined projects.

... and that absolutely isn't one - which is why the data you claim to exist doesn't. You simply cannot control that experiment in the slightest, because you absolutely cannot control the baseline, and there is no realistic control experiment that you can do either.

"the data you claim to exist doesn't"

No one has run experiments to determine
(a) the minimum difference in Hz between two tones necessary for someone to tell that they are different
or
(b) the minimum difference in loudness between two tones of the same frequency necessary for someone to tell that they are different
or
(c) either a or b?

"you absolutely cannot control the baseline"

I presume the relevant definition is
‘a set of critical observations or data used for comparison or a control'
but data for what?

Obviously, if there is some reason the two experiments considered in the first part can not be done, there is no way to go to the next step I'm curious about: does fatigue, or desensitization, or whatever label is most relevant, lead to deterioration in discrimination in a relatively short time?

My questions here are not argument that you are wrong. I don't understand what you wrote. Since I'm interested in the question, if there is no way to get an answer for it, I would like to at least understand why.

I presume the relevant definition is
‘a set of critical observations or data used for comparison or a control'
but data for what?

You are talking about the factors you mention in terms of fatigue - and it's fatigue that there is no possible baseline for. You are assuming that the factors you mention would change with fatigue levels, and there is no evidence that I'm aware of that either the ear's critical bands or ability to detect loudness changes alter in the slightest under these conditions. Fatigue doesn't affect physical discrimination per se; it affects your attitude to what you hear, and your ability to make critical decisions about it, which is not the same thing at all. For instance, I can be completely exhausted and still be able to discriminate pitch changes that some other people can't even detect, quite accurately. It's an inherent ability (somewhat trained-up as well, a long time ago, admittedly) but fatigue isn't a factor in this in the slightest.

And it is universally acknowledged that there are massive differences in peoples' abilities in this area anyway. Not only that, but there are a large number of factors that can influence your attitude to what you are doing, even in the short term, and even if you are tired, and these can make a difference too. But there is no evidence that what amounts to a largely autonomous pair of functions is altered in the slightest in any predictable way  by anything like fatigue.

Hallucinogenic substances might do it perhaps... and you'd probably get more students prepared to sign up for the experiments!

Hallucinogenic substances might do it perhaps... and you'd probably get more students prepared to sign up for the experiments

gawd SteveG, you certainly have no deficiency of fresh perspectives

The first time I saw this post, I passed it by since there is no real way to address it, much less answer. Sometimes I wonder if there is any other perspective to view things than Bacon or Decartes. Even if what you seek could be measured, would there be any meaning to ihe data?

Maybe it might be more profitable to inquire into the metaphysics of measurement (what is it, what is meant by it, what is its essence) rather than to attempt to force things that don't belong together into a makeshift relationship, AndyH. I say that with the greatest of respect, just as I follow up with the observation that most of the more abstract questions you have posed here could only be characterised as "trying too hard." The impasses you are experiencing are inherent in the frame you inhabit, not what exists outside, and thats where the insights lie.

Best wishes.

gawd SteveG, you certainly have no deficiency of fresh perspectives

You have to bear in mind that I come at this from the perspective of being quite an experienced and qualified acoustics researcher, who quite regularly gets all sorts of daft research ideas put in front of him for comment. So please excuse my ever-so-slightly  tongue-in-cheek cynicism... One way or another, I know what students are like!

And I have to say that having not considered this for a couple of weeks, so effectively coming to it freshly, I still don't think that 2+2= 54.7...

I believe the difficulty is more one of communication. I think I can describe how to collect reasonable data on the topic. Perhaps that will paint a different picture than the one I have so far managed to get across. Note that I have not maintained that the topic is "fatigue." If this characteristic exists in human hearing (and I have only asked, not declared), it may be a localized result within the auditory system that has little or no effect in the rest of the body.

While I don't know the data, there is significant knowledge about people's ability to discriminate frequencies. Certainly this differs significantly from one person to another, probably producing the good old bell curve if plotted, and it differs depending upon what part of the audio spectrum we focus.

We start by picking a base frequency where the most people have the best discrimination. We use available data to simplify the task so we don't need to repeat work already done to determine that frequency nor to find the difference generally necessary for discrimination (two tones 10Hz apart or 100Hz apart or ... . Again, I don't know the data, but I sure this has been investigated).

Since people differ significantly, we have to measure for each subject, to determine that person's discrimination ability. The test seems simple enough but other people may be able to think up a more elegant design:
Here is tone A. Is tone B the same or different?
We thus find the minimum difference that each subject can reliably identify. "Reliable" is necessarily determined by the math model that will be used to analyze the data. I would guess that it would be somewhere between 50% and 100% of the time.

Once the baseline for the subject is established, the actual trials can began. I think that would reasonably have to be after some rest period, perhaps even a day. Let us just say, so I can be less vague in how I describe the experiment, that this subject has the ability to discriminate tones that are 10Hz apart (in this part of the audio spectrum) but no closer.

Many pairs of two tones 10Hz apart can be prepared, preferably by a computer program. In some number of the pairs, probably 50%, both tones will be identical. The experiment is for the subject to continue listening to pairs, and declaring if B is different than, or the same as, A. The process must continue for some extended, but reasonable time. Perhaps an hour would be good (and no doubt boring).

The goal of analyzing the data is to find it there is some fall off of accuracy over the time of the subject's engagement in the experiment. There will have to be enough subjects to ensure statistical accuracy. Results will have to be combined for all subjects in order to discover the general trend as some subjects will be effected individually in ways not directly related to this hearing characteristic, if it exists (e.g. some may give up actually trying to discriminate after awhile, others may develop some symptom, physiological or psychological, that actually interferes with discrimination, etc.). Therefore, a fairly large pool may be necessary to counter objections to the data.

I am not saying anyone here should be excited about such an inquiry. I merely asked if anyone knew whether investigation of this subject had been done and, if so, the results.

The process must continue for some extended, but reasonable time. Perhaps an hour would be good (and no doubt boring).

I'm already  bored. It's pointless.

Any particular thing can be boring to any particular person but it is not pointless if one need to know whether such a factor exists. And, speaking of points, while my last post is only a sketchy outline, I don't claim to know enough about experimental design to fill in all the necessary details, this experiment does overcome your earlier objections about no stable baseline and factors that are too indistinct, no?

I don't claim to know enough about experimental design to fill in all the necessary details, this experiment does overcome your earlier objections about no stable baseline and factors that are too indistinct, no?

No, not at all. For a start, you said "There will have to be enough subjects to ensure statistical accuracy". Well, that's a contradiction in terms - there's no such thing as statistical accuracy, because statistics by their very nature do not produce results that are individually accurate. Therefore there is no individual direct benefit to be gained from this research - we already know that the effects exist to a degree, and that the problems that they cause have individually varying thresholds that vary widely - so what would any of this prove, and what earthly use would it be? It is not going to lead to any tangible outcome at all, and nobody's going to be able to make money or a useful product from it. But as I said, and you conveniently ignored earlier, it's more flawed than that...

Okay, the first specific that this falls down on:

We thus find the minimum difference that each subject can reliably identify.

How tired are they when you test them? How much training/experience have they had? How do you quantify this? How do you test for aptitude? What levels are you going to test at? How are you going to account for individual differences in masking detection thresholds? And I could continue. Basically, you have no means of determining the physiological state of each individual without repeated testing over several sessions, and even then you'd only end up with an approximation.

And anyway, it's already been demonstrated to my satisfaction that the ability to discriminate in this way isn't directly related to any form of tiredness, except in an extreme case*, so I'd say that in doing the 'test' (which isn't) at all, you are prejudging the outcome. That's bad science. You clearly haven't even made even the slightest  attempt to understand what the mechanism involved with frequency discrimination is, or how it works - you are just flailing around in complete darkness as far as I'm concerned. First rule of research proposals - do a comprehensive background study before proposing anything, or be prepared to be shot down in flames by somebody who has. In this case, if you want to find out just how much you don't know, go and look at the background research into how ATRAC and MP3 works - then perhaps you will realise why your tests are effectively meaningless.

So as it stands, this would demonstrably a waste of a very considerable amount of time and money, and that isn't good - especially when there is almost certainly research that could be undertaken into areas where real benefit could be obtained. To do something like this, which would reduce the possibilities of something meaningful being studied, could be construed as a research crime.

*subject is asleep, and therefore cannot determine any differences at all.

Thanks for the more explicit rundown on objections. I dislike having to assume I know what someone else means, especially when I don't have details. That can be too much like the woman wailing ‘if you really loved me, you would know what I want.'

I don't intend to pursue this further, I don't think there is anything there, I did not think it very likely to begin with. We do know, however, that what we believe to be true doesn't always stand up to objective evidence. I just wanted to know if there was any real evidence. I do intend to register a few objections, however.

While my wording may not have been specific enough to satisfy you, it should have been clear that ‘statistical accuracy' can only mean determining how high a confidence level is reasonable for such a question and then gathering enough data to compute it. What else could it be? Yes, this is somewhat a turn about from my first paragraph, but not in its full meaning.

I attempted, several times, to separate "fatigue" from this topic. My title choice was unfortunate. I first attempted to better define it in my second sentence. It seems first impressions cannot be overcome. "Fatigue" might have some effects, but that wasn't really the question of interest. By analogy, is it reasonable to tie bodily or mental fatigue to the decreased sensitivity to strong odors that comes with (fairly short term) continued exposure?
 
If everyone were expected to fully research all possible background to a subject, when all they really want to know about is some immediate aspect, there would be darn few discussions on any forum. Yes, it means that people are continually operating out of ignorance, but only so much is possible for one person. One does not need to become a master plumber before replacing a leaking faucet washer. If everyone first did enough research to fully understand a subject, how many clients would you have in your professional life? When I opened this thread I did not intend to propose some research. That outline in my fourth post was just an attempt to illustrate my question and never presented as anything different.

If everyone first did enough research to fully understand a subject, how many clients would you have in your professional life?

Plenty, and ones I'd be much  happier with. I like clients who know what they want, and know enough of the background to a problem to understand whether their requirements are achievable. And generally they like people who appreciate that too.

If everyone were expected to fully research all possible background to a subject, when all they really want to know about is some immediate aspect, there would be darn few discussions on any forum.

Of course there would be - don't be silly. But they would generally be rather higher-quality ones, that's for sure. Also, your reaction doesn't take account of people actually asking informed questions simply and intelligently rather than just postulating half-baked ideas - which is what you did. Three people  have said that this entire idea was unreasonable - and yet you've continued on your merry way, disregarding everything that didn't suit you, and changing what you said around to give you more wriggle room. If you start out by assuming things that are unreasonable, you should expect the reaction you got. You keep saying 'I'm sure - I'm reasonably certain', etc when in fact you are really nothing of the sort. If you say things like that it makes it seem as though you have actually researched something - and yet on the other hand, it's pretty clear that you haven't even begun to. And you wonder at the reaction? Well you shouldn't.

Below are just two of the Google search results for "Human Hearing".  The second one especially seems to address some of the issues posed in the original question.  If not, my bad.

http://hyperphysics.phy-astr.gsu.edu/hbase/sound/earsens.html

http://www.silcom.com/~aludwig/EARS.htm