Wikipedia:Reference desk/Archives/Science/2013 June 22

= June 22 =

Sound of Sound Waves Based off Certain Characterisitcs
Promise this is the last question I'll ask for a while. Given two curves f and g so f(t) >= g(t) for all t; a "template wave piece" w(t) with t ranging 0 to 1, w(0) = w(1) = 0, and there being 0 < a < b < c < 1 so w(a) > 0 is maximal, w(c) < 0 minimal, and w(b) = 0; and an increasing sequence of points t(0) = 0,...t(N) = k. Make a wave W by taking rescaled versions of w and placing them so that W on [t(3m), t(3m + 2] will have t(3m), t(3m+1), t(3m+2) correspond to 0, b, 1 on w (,but be mapped to the midpoints of f(3m+s) and g(3m+s) ) and the points x, y corresponding to a, c will have W(x) on f, W(y) on g.  Essentially, f and g are upper and lower envelopes, w defines the "shape" of one full cycle of the wave, and t(i) determines a frequency. My questions are: what effects do these three components (w; f and g; t(i)) have on the sound of W? In other words, what characteristics of sound are controlled by the various components, even if only approximately? And, how do various transformations affect sounds: for example, doubling the distance between t(i)'s; or stretching out f and g, but adding points to t(i) to keep it relatively the same spacing; etc.? I apologize for any poverty in my description, thank you for any help:-) [also, is there a family of w's so that any wave W can be "closely approximated", in the sense of sound perception, by using it in this fashion with the same f, g, and t as in the original?]Phoenixia1177 (talk) 04:40, 22 June 2013 (UTC)


 * Or, if it's simpler, if someone could point me towards some free/opensource tools I could use to implement my own version of the above and just test it out. I'm fairly confident I could code something from scratch, but I think it would end up very inelegant and unuseful for analysis, so if there are any libraries that could help with that, I'm willing to try that and look into it myself.Phoenixia1177 (talk) 04:44, 22 June 2013 (UTC)
 * Since my description above isn't the best; what if, this is essentially the same, we took a periodic wave with max = 1, min = 0, and rescaled it so the maxs lay on f, mins on g, and where it crossed the midpoint between f and g was at points t(i)?Phoenixia1177 (talk) 06:18, 22 June 2013 (UTC)


 * I think you are describing a form of amplitude modulation, but I can't help with the rest of your questions. Gandalf61 (talk) 13:07, 22 June 2013 (UTC)


 * If you don't have MATLAB and you want to mathematically manipulate raw waveforms, you can use the free software Python Multimedia library. It can import and export PCM and WAV files, allowing you write simple software to directly edit the samples.  You will probably also find Audacity helpful for visualizing the waveform, and for performing higher-level audio processing operations.  Nimur (talk) 16:19, 22 June 2013 (UTC)
 * Octave_(programming_language) is essentially a free MATLAB clone, which would also work. SemanticMantis (talk) 18:10, 22 June 2013 (UTC)


 * Thank you for all of the answers, Octave looks very useful ( :-) ); far more useful than my DIY approach. It seems very difficult to track down any books/papers that deal with how the various aspects of sound waves link to actual types of sounds (hence my questions), I've found a lot of material that talks about how sound waves work, which is useful, but little of it makes clear how that impacts what you will end up hearing (perceptually, not physically). Maybe it's just something you have to experiment with.Phoenixia1177 (talk) 06:45, 23 June 2013 (UTC)
 * The same book I always recommend - which is available for free online: Physical Audio Signal Processing - starts off in the very first chapter: how does it sound? The book is a must-read for anyone who wants to work on home-brew waveform synthesis.  Nimur (talk) 19:22, 24 June 2013 (UTC)

Why humans have more distinctive physical features than most other species?
Aside from sexual dimorphism, it's often hard to tell the difference between the representatives of one species (like between two mosquitos, two rats, two zebras, two chimps etc.) whereas humans feature various eye color, hair color, skin shade and so on. Even when considering a species that, like humans, lives almost everywhere (like mice), that species exhibits less physical features than humans. Is it because humans are a sort of evolution's apex or some other reason?--93.174.25.12 (talk) 08:51, 22 June 2013 (UTC)
 * First, who says your premise is true? Second, maybe you've heard western stereotyped comments such as "all Japanese look alike" (which they clearly don't, but that's the ironic point). Familiarity figures into it. Folks who study chimps or gorillas, for example, are easily able to tell one from another, as they spend so much more time with them than the average person does. Then again, maybe all penguins look alike to us. But maybe not to each other. To do this right, though, you would need to study the DNA variations in some particular species that you think "all look alike" and learn how genetic variations subtly affect appearance. ←Baseball Bugs What's up, Doc? carrots→ 10:30, 22 June 2013 (UTC)


 * If you make measurements, human faces are remarkably alike, which is one of the reasons automated facial recognition is so difficult. We have a phenomenally good ability to analyze human faces, to the point of pareidolia, where we see human faces where there are none, but as Baseball Bugs says, it's very common for people of all ethnic groups to have difficulty distinguishing the faces of people from unfamiliar ethnic groups, even though they have no difficulty distiguishing themselves from one another, and may well have the same problem in the reverse direction. I think this phenomenon may well be related to our hyper-acuity in seeing faces: the viewer is so busy processing the unfamiliar difference in the person's face to be able to distinguish finer differences. This is rapidly fixed by becoming familiar with many members of the unfamiliar ethic group -- which can be done quite easily by living in one of the world's more cosmopolitan cities -- at which point, you stop seeing the macro difference in overall features, and start seeing the micro differences that make each person individual. There's a whole world of academic research to be done about this, and its relationship to both racism, and the experience of self and other: I wonder how much has been done? -- The Anome (talk) 14:34, 22 June 2013 (UTC)


 * As ever, Wikipedia (via Google and TV Tropes) has the answer to my question: it's called the cross-race effect, and lots of research exists. And it looks like research has been done that directly addresses your question, albeit just across human and non-human primates: -- The Anome (talk) 14:42, 22 June 2013 (UTC)


 * And also this, which also directly addresses the issue you've raised: -- The Anome (talk) 14:46, 22 June 2013 (UTC)


 * (ECx2) I think the premise is true. Looking at eye color, skin color, hair color, hair texture, height, and weight, humans do seem to have more variation in those characteristics than many other species.  We could throw in things like the chin dimples and oriental/occidental eye skin fold differences to complete the picture.  (Of course, people also sometimes choose to alter their appearance further, with differences in facial hair, tattoos, etc., but this isn't relevant to this discussion.)  Those species which do show as much variety in appearance as humans are often those bred by humans, like dogs, cats, and livestock.


 * Now, why would this be ? Humans have had a relatively large population, with many isolated populations, in different geographic regions, for thousands of years.  This is enough time to allow genetic drift to occur sufficient to change the appearance of a species, but not enough for the populations to diverge into separate species.  In some cases, like skin color, there are actual genetic pressures based on geographic areas (equatorial populations needed darker skin to protect from skin damage from UV light, while polar populations needed light skin to allow enough UV through to produce vitamin D).


 * Another factor might be that people are social animals which use visual means of identification, so it's helpful if each individual has a unique physical appearance. In animals which primarily use smell for identification, they might look identical but have widely varying odors.  And non-social animals just don't need the range of appearances as social animals, since they don't need to keep so many individuals straight.


 * Then some species have actual genetically identical offspring, like the armadillo with identical quadruplets. This virtually guarantees that they will look similar.  Identical twins, triplets, etc., in humans are relatively rare. StuRat (talk) 14:47, 22 June 2013 (UTC)

I'd agree that the premise is probably wrong. I recall reading about a study that confirmed human babies' ability to tell the difference between faces of different monkeys. Adult humans see no difference, however. Unfortunately, all I could find is this BBC article, but the proper scientific article is out there somewhere. This suggests that although we, adult humans, cannot spot the difference, others can and it certainly exists. Surtsicna (talk) 14:54, 22 June 2013 (UTC)


 * Try this:, http://www.cbcd.bbk.ac.uk/people/scientificstaff/mark/PDFs/Specialization -- The Anome (talk) 15:01, 22 June 2013 (UTC)


 * The only way really to find out if there are really differences in the amount of difference would be to take photographs of many individuals of each species (perhaps about 1000 might suffice?), identify corresponding points (corners of eyes, nose or beak, ears, etc.) and measure, and then perform a statistical analysis of inter-individual variations within each species. -- The Anome (talk) 14:54, 22 June 2013 (UTC)


 * You might also consider that some other animals identify each other in non-visual ways. Extremely well. A mother bat can find her baby (even an adopted baby) in a screaming mass of thousands of other seemingly (to us) identical babies, in the dark. I couldn't do that with my own newborn in a well-lit room of only twenty other white babies.


 * Animals who differentiate with another sense (or eyes that can see beyond our colour spectrum and in ultra-high resolution), don't need to look different (or different in our eyes), for the same reason we don't need to smell, sound or feel different (though we do, to an extent). It'd be rather pointless.


 * And yeah, what they said. InedibleHulk (talk) 15:09, 22 June 2013 (UTC)


 * It's an interesting question: is increased variation in facial appearance something that is selected for, perhaps for improved ability to be the recipient of favourable social behavior from others, or merely a consequence of other factors like genetic drift, that happens to be latched onto by other mechanisms for those purposes? -- The Anome (talk) 15:14, 22 June 2013 (UTC)


 * Dogs (for example) exhibit hugely wider differences in size, shape and coloration than humans. It's true that the species has been selectively bred for these differences, but I think it throws your initial assumption into question. --ColinFine (talk) 15:41, 22 June 2013 (UTC)


 * Also, we have specialized brain parts devoted to recognize human faces. You can have a set of pictures of human faces that look totally different, rotate them 180 degrees and then they all look almost identical. Count Iblis (talk) 15:42, 22 June 2013 (UTC)
 * Number 3 on this list touches on that. It's also a great picture of Obama. (That first illusion is trippy as hell, too.) Our vision definitely isn't foolproof, but great with human faces. InedibleHulk (talk) 16:19, 22 June 2013 (UTC)


 * Related: sheep are very good at recognizing sheep faces, and it is hypothesized that better abilities to recognize individuals are tied to more social species. The idea is that better ability to recognize individuals is more beneficial in social animals than in non-social animals (e.g. humans and dogs, as compared to mice or rabbits). This allows for a synergistic selective pressure between social behavior and recognition. This sort of thing is very hard to demonstrate conclusively, but it is an explanation that fits most of our theories of behavior and genetics. For example, all the social insects are rather good at kin recognition.
 * Thus, I think it is fair to say that humans are rather good at recognizing and displaying individual identity. But, we are not alone in that aspect, and we are certainly no "evolution's apex"! (Actually, we are evolution's apex, in a sense, but then so is every other species alive today-- But that's a discussion for another day :) SemanticMantis (talk) 18:05, 22 June 2013 (UTC)


 * The showiness of human faces, eye, eye color, breasts and buttocks, the fact that we have the largest penises among the apes is a result of sexual selection. We are like birds with showy feathers trying to gain mates.  Were we not such a food-chain dominant animal we might visually plain as most rodents. μηδείς (talk) 18:49, 22 June 2013 (UTC)
 * And as furry. ←Baseball Bugs What's up, Doc? carrots→ 05:03, 23 June 2013 (UTC)

Headphone frequency range
Do all headphones have weaknesses somewhere in the frequency range? Is there such thing as a headphone which is strong and consistent throughout the frequency range? Why/why not? — Preceding unsigned comment added by Clover345 (talk • contribs) 09:34, 22 June 2013 (UTC)


 * Our article Electrical characteristics of dynamic loudspeakers is applicable also to headphones.--Aspro (talk) 13:40, 22 June 2013 (UTC)


 * Traditionally, deep base frequencies were difficult to achieve with tiny speakers. There are new technologies which do a better job there.  I believe Bose Corporation is able to get deep base out of small speakers, for example, using the Bose Wave System, although I haven't heard of them mating this to headphones yet (it might still be a little too big for that).  Bone conducting audio is another way to get bass you can feel "in your bones".  Another concern is that bass takes more energy to produce, so will drain batteries more quickly. StuRat (talk) 14:25, 22 June 2013 (UTC)


 * Yes, the bass range is where headphones generally suffer. Part of that is because we pick up deep bass frequencies largely with our bodies rather than our ears, but part is also because it is hard to produce low frequencies with small vibrators. Looie496 (talk) 15:19, 22 June 2013 (UTC)


 * Just make enormous headphones. Problem solved.--Jerk of Thrones (talk) 06:12, 23 June 2013 (UTC)
 * is that why many people judge the quality of headphones on the bass? Clover345 (talk) 09:04, 23 June 2013 (UTC)

Might be worth checking out Stax earspeakers. I'm not too sure how flat the response is, but were certainly usable for 30 Hz signals such as idle booms in cars. I used to use them with a HEAD Acoustics binaural recording and analysing system. Cost about 100000 bucks all up. Greglocock (talk) 00:33, 24 June 2013 (UTC)


 * Here ya go, flat down to 20 hz, usual fun and games above 1 kHz (the interaction between the spekaer and the ear is complex at high frequency). http://wiki.faust3d.com/wiki/images/f/fd/StaxLa-p4_detail_sm.jpg. As somebody else pointed out, at low frequency a flat response for a headphone sounds as if there isn't enough bass, for example pipe organ music sounds good, but doesn't sound like you are standing in the cathedral. Greglocock (talk) 00:49, 24 June 2013 (UTC)


 * Yep - the problem with bass is that we don't hear it with our ears so much as feel it on our skin and in our gut. It doesn't matter whether your headphones can feed those very low frequencies into your ears - because that's not where you're sensing the experience.  There doesn't seem to be any way to achieve that effect without literally bombarding the body with sound waves at the same frequency and amplitude as the live event.  Since some of the things we like to listen to (like rock concerts and pipe organs) are pretty loud, the only way to get a truly realistic feel is to play the music (or at least the lower frequencies) incredibly loud - which tends to defeat the object of headphones!  You could possibly imagine a skin-tight suit - (or maybe just a waist-band!) that would vibrate at the right frequencies to produce the feel - but that kind of vibration would likely leak bass sound into the air at the same kind of volume that the original live event would have done.  So you're essentially doomed.  There isn't a solution - even with hypothetically perfect headphones. SteveBaker (talk) 01:47, 24 June 2013 (UTC)

Eigengrau for the ears?
The article Eigengrau gives a name to the universal experience of greyness that we see when there's no light to see: is there a corresponding article, or indeed a name, for the hiss that we hear when there's nothing but silence to hear? Do deaf people hear this hiss? -- The Anome (talk) 14:22, 22 June 2013 (UTC)
 * The word I know of is tinnitus, although that's usually used to describe cases where the background sound perception becomes so loud as to be obnoxious. People with certain forms of acquired deafness can experience it, although it's most common with partial hearing loss. Looie496 (talk) 15:41, 22 June 2013 (UTC)
 * And by the way, the material in that Eigengrau article should be taken with a grain of salt. The sources are antiquated -- one is from the 1890s! Looie496 (talk) 16:37, 22 June 2013 (UTC)
 * On further investigation, the term "Eigengrau" has basically fallen out of use -- the last scientific papers to use it were published in 2000. Nowadays, I believe, this effect is generally thought of as a type of hallucination, but its mechanism is poorly understood. Looie496 (talk) 16:56, 22 June 2013 (UTC)
 * From the same source as phosphenes, perhaps? I get a lot of both. -- The Anome (talk) 19:28, 22 June 2013 (UTC)
 * As a sidenote, it is useful to realize that what we see when there's no light to see are truly random, thermodynamic noise driven, processes (mostly the spontaneous isomerization events of the opsin molecules AFAIK). By contrast, what we hear when there's no external sound is not only the random events in the auditory hair cells and ribbon synapses, but also all the audible-range vibrations generated by our body: blood circulation, heartbeat, breathing, oculomotor activity, peristalsis, salivation, and so on. Furthermore, our outer auditory hair cells are capable of generating active motion in response to the stimulus (although I do not know whether they exhibit any spontaneous active motion or not). --Dr Dima (talk) 20:44, 22 June 2013 (UTC)

Supersonic/Hypersonic Movement at Low Altitude
How fast could a vehicle/anything else move at low altitude (below 300 meters) without being destroyed or destroying the terrain? Is very low altitude hypersonic/supersonic flight conceivable (not counting fuel efficiency, of course!)? Longraptor (talk) 18:38, 22 June 2013 (UTC)


 * Are you asking for the speed at which terrain avoidance becomes the limiting factor in low-altitude supersonic flight? Or are you asking what other physical effect is the current limiting factor that caps the maximum airspeed in low-altitude flight?
 * The answers to either of these questions will be difficult to find, because state-of-the-art aeronautical capabilities tend to be unavailable in published resources. But you can take a look at terrain following RADAR to get a ballpark idea of current capabilities.  And you can look at the specifications for modern air superiority aircraft to get an idea of the speeds of which they are (reputedly) capable.  Most modern aircraft are designed for highest speed at high altitude, because that is the mission they are engineered for.  Nimur (talk) 19:20, 22 June 2013 (UTC)
 * I think he means the speed at which terrain avoidance becomes the limiting factor. Whoop whoop pull up Bitching Betty 22:58, 22 June 2013 (UTC)
 * Our F-111 could fly at about Mach 1.2 (900 KEAS) at treetop height, the limiting factor being aeroelastic stress from dense, turbulent air at low altitude; our B-1 bomber is only slightly slower at low altitude. 24.23.196.85 (talk) 19:37, 22 June 2013 (UTC)


 * The 1999 7ECA Citabria that I fly starts losing engine power at "high altitude,", which is my speed-limiting factor above 5000 feet no matter how I lean the mixture; by my measurement calculated with my thin and light aviation-grade aluminum flight computer, I make Mach 0.16 at sea level and made just about Mach 0.145 at 5500 feet over California's central valley on a hot day. It may be about time to rate on a more complex aircraft.  On the other hand, because I fly a taildragger, I am always "in a landing configuration" which means that I may legally fly at treetop altitude, per 14 CFR 91.119 - unlike our supersonic counterparts!  Nimur (talk) 19:41, 22 June 2013 (UTC)

The US did some testing in 1968 on Sonic Booms Resulting from Extremely Low-Altitude Supersonic Flight: Measurements and Observations on Houses, Livestock and People with the McDonnell Douglas F-4 Phantom II. The test range was 5000 to 8000 ft ASL, however the test flights were done at 85 - 125 ft AGL and at speeds between 1.1 Mach and 3 Mach 1.3 Mach. CambridgeBayWeather (talk) 01:27, 23 June 2013 (UTC)


 * What are you talking about? The Phantom is not even capable of Mach 3 -- its limiting Mach number is "only" 2.4! 24.23.196.85 (talk) 00:31, 24 June 2013 (UTC)


 * Typo. It was supposed to be 1.3. I've corrected it. CambridgeBayWeather (talk) 00:58, 24 June 2013 (UTC)


 * This comes out to about 829 to 874 KEAS, which is pretty impressive! The Phantom must have been one tough bird in the structural sense!  24.23.196.85 (talk) 01:32, 24 June 2013 (UTC)


 * Sorry if I wasn't clear; I wasn't talking about terrain avoidance but factors such as atmospheric drag, air compression and so on. For instance, could you fly a fighter at treetop height at supersonic speeds without the airplane destroying the ground below or the air compression/friction tearing the jet apart. And what happens when you fly a fighter at maximum speed at low altitude? What are the air compression and friction forces involved? Longraptor (talk) 07:58, 27 June 2013 (UTC)

Chance of pregnancy per year vs age of parous and nulliparous premenopausal women
Where can I find a table giving the chance of pregnancy per year vs. age for parous and nulliparous, sexually active, premenopausal women? The tables at Contraception and Comparison of birth control methods give an 85% "failure rate" for the "No birth control" method, which I assume is an average across all women of child bearing age. I would like to be able to look up the comparable value for, for example, 50 year old, premenopausal, nulliparous, sexually active women. -- 179.202.36.245 (talk) 21:19, 22 June 2013 (UTC)
 * perhaps a graph like this would be helpful: http://www.babycenter.com/0_chart-the-effect-of-age-on-fertility_6155.bc  --Digrpat (talk) 22:03, 22 June 2013 (UTC)
 * Thanks. I'd like to see how the right side of the graph would change if restricted to premenopausal women.  (Separating parous &  nulliparous would also be interesting, but not as important.)  Your example gives me some better terms to search on, and I'll add a note if I find anything better. -- 179.202.51.55 (talk) 23:02, 22 June 2013 (UTC)
 * ... and all I seem to find is that chart or something similar in a lot of places. It is attributed to "Management of the Infertile Woman by Helen A. Carcio and The Fertility Sourcebook by M. Sara Rosenthal" and the two data points on the right of the graph give a 5% likelihood of pregnancy over the course of one year for 45-49 year olds and 0% (presumably +/ 0.5%) for 50+.  It would be interesting to learn the number of older participants in the study. -- 179.202.253.161 (talk) 20:54, 24 June 2013 (UTC)