Wikipedia:Reference desk/Archives/Science/2011 April 15

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

April 15

Why is DC voltage used in public transportation

Help my boyfriend sleep better at night. He's currently tormented by the idea that today's worldwide subway, tramway, and electric train system use brutally inefficient DC Voltage rails to power motors. Since voltage must be supplied over a long distance (a whole rail, or at least have a long cable going up to the rail in question) there would be a lot of energy lost even when there is no train going on the rail (leaking capacitor).

So why does subway, maglev, tramways, and train use DC third rail exclusively? Esurnir (talk) 02:40, 15 April 2011 (UTC)

Electric motors operating on direct current (DC) are readily controlled to vary the speed. This allows the vehicle driver to choose the speed at which the vehicle is traveling. Electric motors operating on alternating current (AC) rotate at synchronous speed, or some fraction of synchronous speed, such as 3600 revolutions per minute (or 3000 or 1800 or 1500 rpm etc.) It would be feasible to have the distribution system supplying alternating current, and for each vehicle to have an AC motor coupled to a DC generator, and then have one or more DC motors driving the traction wheels, but that would double the cost, double the weight, and double the electrical losses.

Variable-speed AC motors are now available, but in the past they were not, and they are more temperamental than DC motors. Very-long-range power distribution is often done with direct current because of lower losses. Dolphin (t) 02:51, 15 April 2011 (UTC)

(ec) DC power lines are "leaky capacitors"; AC power lines are "giant antennas"; both suffer losses due to physical limitations. But, it's rarely a good idea to try to apply "first principles" physics to complicated, sophisticated engineering projects; estimating the losses requires detailed analysis of specific technologies and parameters for any particular project. While AC power has certain theoretical advantages, it also has certain practical disadvantages. In today's technology, high-voltage DC is probably more efficient than AC for long-distance power transmission, though in any particular instance, specific engineering details may sway the balance one way or the other. DC systems are usually supplied by AC from a power plant, so there is a conversion loss to worry about; but you never have to worry about phase matching, nor radiative losses. We have numerous articles on electric train topics; the most helpful will be Railway electrification system, which discusses AC and DC systems. Nimur (talk) 02:53, 15 April 2011 (UTC)

I don't get it - if HVDC is so efficient, why has everyone been drilled into thinking that long-range transmission requires AC? Wnt (talk) 05:09, 15 April 2011 (UTC)

High-voltage AC is easy to convert to low-voltage AC. It's harder with DC. --Trovatore (talk) 05:18, 15 April 2011 (UTC)

Yes, in a word transformers. These are only available for AC (except in network theory books). High voltage (sometimes as high as 400 kV) gives less loss over long distances but is unsuitably high for most power stations to generate directly (around 25kV is normal for anything with a turbine) and is way too high to be safe in a factory or a home. SpinningSpark 11:30, 15 April 2011 (UTC)

The AC vs. DC dispute is old, see War of Currents. Cuddlyable3 (talk) 12:18, 15 April 2011 (UTC)

The article link that Nimur provided in his comment provides a good discussion, but briefly, HVDC transmission makes economic sense only for long runs with few 'taps' off them. First, there is a small loss incurred each time current is converted from AC to DC or back again, so the line has to be long enough that the increased efficiency on the line more than makes up for the bigger losses in AC/DC conversions at the ends. Our article puts the break-even distance as about 50 km for undersea cables and 600-800 km for overhead cables, but it doesn't have a supporting reference. Second, each tap off the HVDC line needs to have an installed high-voltage inverter (to convert DC to AC) before the transformer, this increases the cost of each tap. TenOfAllTrades(talk) 16:33, 15 April 2011 (UTC)

Diagnosing an electrical problem

In my house all the lights on the even circuits are blinking, computers are rebooting on their own, etc., while the odd circuit numbers are fine. There are two "phases" coming into the house, one which feeds the even circuits, and one which feeds the odd circuits. So:

A) This indicates to me that the source of the problem is external to the house. Is this correct ? In this case, how can I convince the power company it's their problem ? (The power company is Detroit Edison.)

B) Is there any danger to continuing to use the even circuits until it's fixed ? (I installed an uninterruptable power supply for the computer, to stop it from rebooting when the voltage drops.) The other approach is to use extension cords from the odd circuits, but I doubt if the whole house can be run on half power like that, especially if this isn't fixed by the time we hit A/C weather (we use window air conditioner units). StuRat (talk) 03:59, 15 April 2011 (UTC)

No, the problem can be at the transformer, or it can be in the house. If you have a high-resistance arcing connection in the panel, disconnect or meter socket, you could have a fire. This should be checked out immediately. If you run any 240V equipment, it won't work efficiently, and if it's a 240V motor, like a heat pump or high-output air conditioner, you'll burn out the unit. Intermittent loss of one leg (it's technically not a phase) is often a sign of a failing circuit breaker - in this case, the main breaker. Have it checked as soon as possible by an electrician. Acroterion (talk) 04:17, 15 April 2011 (UTC)

There's no sound or smell of arcing, so it doesn't seem likely that's happening within the house. Nothing is running 240 in the house (we have gas heat, water heater, and dryer, and small window A/C units). StuRat (talk) 05:59, 15 April 2011 (UTC)

If the problem is at a transformer, neighboring homes will also be affected (unless yours is the only one on that transformer.) We had a situation years ago where some circuits were fine; others had no power. They had to replace the transformer to fix that.

I get the impression this situation has gone on for some time. If the power company won't check the situation, perhaps the only way to get their attention is to have an electrician tell them the fault is in their equipment. I'm quite sure our power company would be out quickly to check the transformer if I reported such a problem.

The above is just speculation. As Acroterion says, Have it checked as soon as possible by an electrician. Wanderer57 (talk) 05:28, 15 April 2011 (UTC)

You can have high-resistance connections and arcing without sounds or smells. My money would be on either a transformer fault or a breaker fault (I had a similar issue with a 50A 240V breaker for a range - the oven wouldn't heat all the way and half the burners didn't work - replacing the breaker fixed it), but the possibility of a faulty connection is sufficiently dangerous to warrant immediate investigation and action. An electrician can provide ammunition if it's the power company's fault; I got my power company to add a transformer for my house and my neighbors after convincing them that six houses on a transformer was too many - we had serious voltage drops every time somebody turned on a load. Get it checked out immediately: it's potentially dangerous. Acroterion (talk) 13:06, 15 April 2011 (UTC)

Thanks so far. How would an electrician determine if it's the main breaker or the transformer ? StuRat (talk) 18:06, 15 April 2011 (UTC)

Measure the voltage before the breaker with a graphing multimeter. You could probably do it yourself if you like to take risks, the wires are accessible in the box. You could probably attach a light before the breaker, and one after and watch them. It's risky though - you have no protection against shorts while working on it. Also, ask your neighbors, if it's only you it's your breaker, or meter, or wires, or junctions. Ariel. (talk) 19:33, 15 April 2011 (UTC)

What Ariel said; the main feed lugs are normally shielded, but can be accessed by a very careful person, ideally one who is used to doing this. Since it's intermittent, some time will be needed, which are further grounds for leaving to an electrician, who will have the right clamps. If there's a load on the leg and the problem is internal, there may be visible damage or heat somewhere in the panel - the main breaker may be hot.. Bear in mind that conductive tools inside a live panel can be dangerous for the uninitiated; an arc flash (vaporized copper) in your face can blind you. Acroterion (talk) 19:41, 15 April 2011 (UTC)

Damn Detroit Edison! Perhaps they should never have been granted a franchise. Better to have a DC generating plant every mile or so. Edison (talk) 03:51, 16 April 2011 (UTC)

UPDATE: We inspected the wires outside the house. Where the service drop connects to the house via 3 insulators, the top connector is melted, and the tape around it has burnt off. The neutral and other leg look fine. I will call the power company tomorrow and hopefully they will agree that this is the problem and fix it. StuRat (talk) 21:03, 17 April 2011 (UTC)

Since it's on the power company's side of the meter socket, they'll have to take care of it. It might be intermittently grounding itself to the house, which would be bad, or the conductor might have been so damaged in the event (lightning strike?) that it has poor conductivity. Glad you've figured it out. Acroterion (talk) 01:56, 18 April 2011 (UTC)

It looks like they just twisted two wires together in the hopes that they would have enough contact to make a good electrical connection, then wrapped them in insulating tape. This is common practice in a light switch, but I'm surprised they would think this was good enough for the mains connection. Since it's exposed to the elements, probably water got at it and froze and forced the wires apart until we got to the intermittent connection we now have. The resulting arcing likely burnt off the remainder of the tape and melted the connector. StuRat (talk) 02:15, 18 April 2011 (UTC)

That certainly explains the problem. The appropriate repair, short of a new drop (which you should encourage them to install if the insulation on the hot conductors is doubtful) would be a clamp connection with the wire ends laid parallel - they have those clamps on the truck by the bucketload. If you have three separate wires running to individual insulators, the drop's so old that they would probably want to replace it with cable and make the connections in the drip loops. Acroterion (talk) 14:37, 18 April 2011 (UTC)

Yes, we would like a new drop, mainly because the old one is so low we are in danger of striking the wires. However, it seems they would want us to get permits and pay for that. Our current drop is a single cable, that splits into 2 legs and the neutral, right before they attach to the house. StuRat (talk) 23:43, 19 April 2011 (UTC)

Relativistic mass

Hello. Mass in special relativity#Controversy explains that some researchers have rejected the concept of "relativistic mass", but apart from saying that it is "a troublesome and dubious concept", the article doesn't go into any depth regarding why they reject relativistic mass. Could someone please tell me what is fundamentally wrong with this concept? Thank you. Leptictidium (mt) 06:17, 15 April 2011 (UTC)

It just seems like a fudge factor to me. That is, when the numbers didn't add up, they just decided to say that the mass was changing. Imagine if your tax accountant could do that: "well, the balance sheet doesn't balance, so I will say that this is 'dynamic cash' and changes quantity as needed to make everything balance". StuRat (talk) 06:40, 15 April 2011 (UTC)

Isn't this just what the central banks do all over the world? 95.112.143.65 (talk) 09:02, 15 April 2011 (UTC)

It's not a fudge factor and it is a occasionally useful concept. For instance, if you put a block on a scale to measure its mass, then you heat it up to a higher temperature increasing its internal energy, its relativistic mass also increases and that should be measurable by the scale in principle (In practice the effect is too small to be measured). The problem is that the concept causes more confusion than it's worth and the modern convention is to reserve the word mass for the rest mass as BenRG points out below. Dauto (talk) 15:08, 15 April 2011 (UTC)

What's troublesome and dubious is thinking that you can plug relativistic mass into equations that have an m in them, such as F=ma, and get something that makes sense. Generally, you can't. To the extent that relativistic mass is just energy divided by c², it's a well-defined concept, but there's no point having two names (energy and relativistic mass) for the same thing. The modern convention is to call it energy, and reserve the word "mass" for rest mass (which is often written in units of energy as well). -- BenRG (talk) 09:23, 15 April 2011 (UTC)

I think it's quite the opposite - if you don't include relativistic energy in equations such as f=ma you will get incorrect results. This energy has inertia and momentum and causes gravity. Which is why I prefer to call it mass. But the hard part is that the mass of an object is relative, it's not fixed. This can make calculations all but impossible. For example, what is the mass of a magnet? If I turn on an electromagnet on the other side of the planet, the magnet in my hand is now heavier (relative to that electromagnet anyway). Ariel. (talk) 19:30, 15 April 2011 (UTC)

If you use the proper form $F={\dot {p}}$ with the three-momentum $p=\gamma mv$, you can easily dispense with "relativistic mass". Here's a quote from a book by Taylor and Wheeler: "Our viewpoint ... is that mass is an invariant, the same for all free-float observers... In relativity, invariants are diamonds. Do not throw away diamonds!" Modern physics understands relativity as a geometrical theory, a theory of the structure of space-time, not as a dynamical theory as suggested by the concept of relativistic mass. In relativity, energy and mass remain distinct physical quantities; "relativistic mass" obfuscates that distinction. Having said that, if the energy is in internal degrees of freedom of a composite body (not center-of-mass motion), then this energy does indeed show up in the effective mass of that body. That's where the mass deficit of, say, the helium nucleus comes from. --Wrongfilter (talk) 20:38, 15 April 2011 (UTC)

Ariel, the point BenRG was making is not that The formula F=ma is correct if you use the rest mass. The point is that even if you use the relativistic mass the formula is still wrong which seems to have gone right over your head. Dauto (talk) 03:15, 16 April 2011 (UTC)

Triplet v.s. Singlet Oxygen

According to the formula for calculating the bond order for diatomic molecular species, both triplet and singlet oxygen species have bond orders of 2. This does not make sense to me - triplet oxygen cannot have a double bond and concurrently be a diradical. I think the reason for this is that the formula ignores electron spin direction. Is this true, and does triplet oxygen have a bond order of one? Or, do I have it wrong? Plasmic Physics (talk) 08:17, 15 April 2011 (UTC)

Sure it can. You cannot draw a proper lewis structure for O₂ with a bond order of two and still have it be a diradical, but that's just a limitation of lewis structures. The molecular orbital diagram at the lower right corner of triplet oxygen shows how it works. The diradical occurs in the two degenerate π* antibonding orbitals. The bond order is calculated as: (bonding electrons - antibonding electrons)/2, which (8-4)/2 = 4/2 = 2. The reason that the Lewis Structure doesn't work out is that the geometry of the molecular orbitals does not work easily in a 2D representation, and lewis structures ignore the whole "antibonding" thing all together. --Jayron32 12:17, 15 April 2011 (UTC)

I was not using the Lewis model approach. Each oxygen has 6 valence electrons, if it has a bond order of two, that means that each atoms makes a net contribution of two electrons to the bonding orbital. This means that each oxygen has four remaining valence electrons. If triplet dioxygen is a diradical which I believe it is, then each atom has only one electron pair and two unpaired elecrons, not just one (according to a two bond order system).

I used the MO diagram for my argument, from it I used two factors - the total number of valence electrons, and the total number of unpaired electrons. Plasmic Physics (talk) 12:50, 15 April 2011 (UTC)

You can't consider the electrons of each atom separately when you're looking at their arrangement in the combined molecule. Electrons that are unpaired for the independent atoms are not required to remain unpaired in the final molecule. Do you understand how the orbitals are filled in the MO diagram at triplet oxygen, and how to calculate bond order based on filling of bonding and antibonding orbitals? TenOfAllTrades(talk) 13:51, 15 April 2011 (UTC)

@ Plasma Physics: To expand on what TOAT said above, when considering the organization of the electrons in the O₂ molecule, you cannot consider each atom as retaining any individual characteristic. The whole thing with Molecular orbital theory is that you treat the entire molecule as a single entity, and calculate the quantum states of the electrons based on that presumption. The system cannot be accurately modeled (for these purposes) as individual atoms which are merely sharing a few electrons (which is how both valence bond theory and hybridization theory model bonding). Molecular orbital theory models the O₂ molecule as a 16-electron system with two nuclear charges of 8+, and calculates the shapes and populations of the various orbitals that way. For convenience, the molecular orbital diagram referenced above only looks at the valence electrons, but the actual calculations are based on all of the electrons. Its the organization and shapes of these molecular orbitals that gives rise to the particular properties of triplet oxygen, that being that it has a second order bond (or "double bond") and is diradical. You can easily arrange all of the electrons in orbitals to get this result; in the case of triplet oxygen you have a total of 8 valence orbitals: five of them have 2 electrons in them, two have one electron each, and the last one is empty. That's 8 orbitals, a double bond, and a diradical. This is empircally confirmed by things like the bond strength and length of the O=O bond in O₂ (compare the bond lengths of O=O with the average peroxide bond length of O-O here, which gives O=O a bond length of 121 picometers, and O-O a bond length of 148 picometers. This is on par with the 20 picometer difference between C-C and C=C), and with the magnetic properties of ground state(triplet) O₂, which is experimentally confirmed to be paramagnetic, as would be expected of a diradical. To sum up; the experimental evidence indicates that oxygen is BOTH second bond order (double bond) and a diradical. Molecular orbital theory predicts both facts about oxygen (again, check the MOT diagram at triplet oxygen), so it is the model which is perhaps best in describe the organization of the oxygen electron cloud. There's actually even MORE evidence that confirms this model of triplet oxygen (vis-a-vis chemical reactivity and spectroscopy), which I'll not go into in the interest of not extending this discussion to the TLDR point. If you are using a model that leads you to a structure which contradicts the empirical, observed properties of oxygen, then simply put, the model is inadequate and needs to be discarded for this purpose. --Jayron32 14:33, 15 April 2011 (UTC)

Well, it's just that the molecular modelling program I'm using does not like the idea of a double bonded dioxygen diradical. It seems to have the least problems when the unpaired electrons are on the same atom. Yes, I have to specify where the unpaired electrons are. One more thing, why do I find lewis structures of single bonded triplet oxygen on google images? Plasmic Physics (talk) 20:47, 15 April 2011 (UTC)

The usual skeletal diagram notation doesn't work so well for complex MO situations. The line between two atoms represents a covalent bond, which means a shared pair of electrons. But the diradical isn't a shared pair making one bond--that's impossible if they have the same spin. Instead (again, exactly as the MO diagram illustrates) it's two shared lone electrons each in separate orbitals that are orthogonal to each other. The limitation is in your modelling program and meaning implied by the diagram style (or at least in the the modelling program's interpretation of the diagram style). Drawing it as a single-bond with a single electron on each atom at least gets the idea of "electrons are not paired, and therefore easily have same spin" correct, at the cost of error in the estimated bond-length. Drawing it as a double bond would imply (to those that don't know the details) that ³O₂ would undergo reactions characteristic of other pi bonds, which is not true. DMacks (talk) 21:03, 15 April 2011 (UTC)

OK, helpful. What do you think, should there be two bond order formulae, one for calculating the total net bond order, and one for calculating spin matched bond order? In the spin matched bond order, only electrons of the same spin cancel, this kind of formula can distinguish between a true bond, and a quasi-bond, which is really what triplet oxygen has. Triplet oxygen has an effective bond order of one and two halves. Plasmic Physics (talk) 21:34, 15 April 2011 (UTC)

"One and two halves" is a pretty good description! One way might be to have the "second" bond written as two dots (but in the bonding position) rather than as an actual connecting line (like the sigma bond is)? The minimum technical notation is to include a superscript "3" to the left of the whole diagram to indicate the spin (see my last sentence in my previous comment), just like you can write the net charge of a structure as a superscript to the right.^† I would like a more detailed way (analogous to "formal charge" on atoms vs "net charge" on a structure) to indicate where the unpaired spins are rather than just the fact that there are two of them, but I don't know a notation for that. But using the "individual dots" vs whole bonds comes close perhaps. A more technical way is to write the spin at each electron (up/down vector), not just its location (dot) or bonded/nonbonded state. DMacks (talk) 15:31, 16 April 2011 (UTC)

^†(later/expanded comment on this point). That's actually the standard way for quantum modelling programs: you declare the spin of the system not of specific electrons or parts of the model. The whole Lewis/skeletal-diagram idea of covalent bonds as specific pairs of electrons at specific atom-pair locations is of a bunch of crap (to use the scientific terminology) at that level anyway. So ³[O=O] vs ¹[O=O] would be fairly accurate for bond-lengths, bond-order, and spin, at the cost of the less important (for many readers) idea that one of those lines is "two half-bonds" rather than a normal covalent pair. But it also leads the reader to think the only difference is one of an overall or hidden detail, even though it really is a major difference in the whole nature of the bonding and physical and chemical properties. DMacks (talk) 18:27, 16 April 2011 (UTC)

While I'm on topic, while I studied the consequences of particle spin at Univeristy, I have no idead what spin actually is. All I know about it, is that it is some kind of description of a particle's kinetics. Does an electron cycle around an actual locus? If it does, what are the constraints? Does the locus describe a point, line, or a surface? Does the locus lie within the electron? Plasmic Physics (talk) 00:07, 16 April 2011 (UTC)

If a charged particle (such as an electron) were to spin, that would cause it to have a magnetic moment and other magnetic effects. These particles do have the properties that result from spinning, therefore they are described as having "spin". But it's an intrinsic property itself, not quite that the particle really "spins". To my mind, since a quantum particle doesn't have a definite position, I can't see how that would be consistent with it having a definite physical rotation as the actual underlying property. See spin (physics) for the gory details. DMacks (talk) 15:31, 16 April 2011 (UTC)

Period food

Are there any studies into what foods women particularly crave on their periods? I know sugary food is pretty common, and iron-rich food, as well as the more specific chocolate, but I'm interested in if anything has looked in more detail or more broadly. I don't have access to most of the paid-for literature at the moment, so public-domain or summaries would be appreciated. 86.164.75.102 (talk) 09:48, 15 April 2011 (UTC) Brain-fart, I meant freely available, not actually released into the public domain. I think. 86.164.75.102 (talk) 15:22, 15 April 2011 (UTC)

Binge eating occurs in a minority of menstruating women. This may be due to fluctuation in beta-endorphin levels. Source: Price WA, Giannini AJ (November 1983). "Binge eating during menstruation". J Clin Psychiatry 44 (11): 431. PMID 6580290. See also Premenstrual dysphoric disorder.

Thanks, that's an interesting start, although I cannot read the paper (even its summary). The Wikipedia article is an interesting read, as the more recent findings on variable response to hormones fits with some of the stuff I've been reading about different women's responses to different versions of the pill, suggesting years more profitable research to be done. So thanks. For this question, I suppose I'm looking more for research into food cravings experienced by women in the 'normal' range, rather than eating associated with various interesting disorders. 86.164.75.102 (talk) 15:22, 15 April 2011 (UTC)

Pinion?

what is racken pinion —Preceding unsigned comment added by 175.157.66.79 (talk) 14:09, 15 April 2011 (UTC)

Did you mean to start a new question? (I've reformatted this, because I think you did..) Are you perhaps thinking of Rack and pinion steering in an automobile? SemanticMantis (talk) 14:54, 15 April 2011 (UTC)

See also rack railway.--Shantavira|^{feed me} 16:57, 15 April 2011 (UTC)

Design pattern: both for software and art

Is there a design pattern which can be applied both to works of art and software? Quest09 (talk) 15:40, 15 April 2011 (UTC)

You may be interested to read the works of Lawrence Lessig et al. The so-called "free culture movement" inherited its philosophical inspiration from the free software movement. Nimur (talk) 17:18, 15 April 2011 (UTC)

I'm not sure that I necessarily understand your question correctly but you may also be interested in Carlos Amorales' modular "Liquid Archive" approach to making his art. Sean.hoyland - talk 17:46, 15 April 2011 (UTC)

I also am not sure what your question means. could you be thinking of fractal geometry? as in the Mandelbrot set.190.148.136.166 (talk) 19:32, 15 April 2011 (UTC)

Software is often created by iterating the cycle Run-Crash-Debug. Most artists and composers follow a similar cycle of successive improvement e.g. Compose-Listen-Make better. Cuddlyable3 (talk) 12:19, 16 April 2011 (UTC)

Cuddlyable3 has the precise answer: something that can be applied both by an artist and by a software developer. The rest is also interesting, in a more broad context.Quest09 (talk) 12:34, 16 April 2011 (UTC)

Live vs. recorded sound

Why do they sound so different? Quest09 (talk) 15:42, 15 April 2011 (UTC)

You mean Acoustic music vs. electronically recorded music played back through electronic amplifiers and loudspeakers? Pfly (talk) 15:50, 15 April 2011 (UTC)

Live sound reaches your ears as each sound source sends out waves which bounce around rooms and off surfaces and arrive at your ears. Recorded sound can only approximate all of these relationships, and generally (unless you are an audiophile with a shitload of cash and time) most people listen to recorded sound from a set of earphones/headphones or speakers which are at a fixed location, and so do not accurately model the actual source sounds. They do a passable job for most people who just want to listen to the latest Lady Gaga song, but there is a noticable difference in the sound of hearing her come from the speakers on the stereo in your living room, and hearing her sing in your living room. --Jayron32 16:01, 15 April 2011 (UTC)

Also, consider a piano, for example, vs. a recording of a piano played back through loudspeakers. There's a huge difference between what is making the sound. Loudspeakers are nothing like pianos. A piano's sound board alone is a resonating surface far larger and different in design and function than a loudspeaker. The amazing thing is that loudspeakers can even come close to sounding like so many different things, from pianos to people singing to cymbals and so on. Still, loudspeakers do a poor job of reproducing the experience of sitting close to the front of an orchestra, or listening to a powerful pipe organ playing full bore. Pfly (talk) 16:14, 15 April 2011 (UTC)

But if the question is about live, as in performed in "real time" vs. recorded, as in created in a studio, then the answer is something to do with live performance vs. the process of building something slowly and in pieces. Something like theater vs. film. Pfly (talk) 16:12, 15 April 2011 (UTC)

No, I asked in the sense that you pointed to at your first reply. But, what's the difference in the wave that hits my ears when coming from a recording or coming directly from an instrument? Which properties are different? Quest09 (talk) 16:33, 15 April 2011 (UTC)

The waveform is different; and to be strictly accurate, the only real reason is because the wave field is different. It is difficult (but not impossible) to completely re-synthesize the entire wavefield; instead, electronic recording only resynthesizes a sampled waveform. For most purposes, your human ears are a stationary set of two single points that sample the pressure level of a multidimensional acoustic wave field; a more complete description of the total wavefield must account for its extent in 3 spatial dimensions, plus the pressure, velocity, and other non-linear acoustic properties of sound waves in air. Most electronically recorded sound waveforms only seek to sample the pressure as a function of time at one (or maybe two or more locations). (They accomplish this by recording pressure fields, placing a microphone as a pressure transducer at one or more fixed locations). For most purposes, this totally replicates the "99th-percentile" of the audible experience; but it is a known fact that the perception of sound waveforms is actually more complicated. A good friend of mine worked on a project to synthesize a multidimensional surround-sound experience; you can read "artistic" and technical descriptions at the technical specifications page. Nimur (talk) 17:27, 15 April 2011 (UTC)

From the list of sound wave properties, frequency, wavelength, and wavenumber are basically the same in this context, as are amplitude, sound pressure, and sound intensity. "Speed of sound" doesn't seem relevant in this context. So that leaves three basic properties, frequency, amplitude, and direction (Nimur addressed wave shape issues above I see, which curiously isn't included in that properties list). A few quick comments:

• Frequency: typical consumer-grade loudspeakers have trouble with very low and very high frequencies. Not uncommonly they drop off below 40-60 Hz and fall off somewhere around 15,000 Hz or perhaps around 20,000 Hz, depending on quality. Subwoofers can handle very low frequencies, but typical consumer-grade, not-super-expensive subwoofers are not very good at producing strong stable pitches, tending more toward booms and rumbles. Compare pipe organs, which can produce extremely low pitches--sometimes so low you feel them more than hear them. No subwoofer can make a sound like a big 32' pipe on a pipe organ. I was lucky enough to get to watch this pipe organ being installed. The sound of the 32' pipes was intense. If I recall right, that beast can accurately and loudly produce pitches down to 8 Hz or so. There's further issues regarding the recording medium, which limit and distort the frequencies in various ways (see Sound recording and reproduction).
• Amplitude: With electronic amplification you can make acoustic instruments much much louder than they could ever be themselves. It is common, I think, to listen to recorded music at louder volumes than would be normal for acoustic instruments (at least the quieter ones like guitar). Plus, when music is put together in a studio it is typical to mix the amplitudes of different instruments in "unnatural" ways--a quiet singer overpowering a drum kit, for example. In olden times you needed a operatic voice to overpower loud instruments. Today on a recording a whisper can overpower a marching band, if desired.
• Direction: The sound from stereo loudspeakers comes from two relatively small, usually non-moving places. Surround sound systems up that to five or so. Compare a live modern orchestra, where the sound comes from well over 100 sources. Even a single acoustic instrument, say a piano, produces sound from multiple sources—the strings, the sound board, etc. Pfly (talk) 18:05, 15 April 2011 (UTC)

To clarify: if you are only modeling the sound pressure level, you are only modeling p wave; this is the straightforward "acoustic wave equation" that is suitable for describing sound in a gas (like air). But in reality, air is not ideal; it has some viscous properties; and you should use a suitable elastic wave equation, such that the velocity of any individual particle is not equal to the velocity of the wavefront. It is unlikely that most human ears can perceive this difference, but we absolutely can measure this phenomenon using sophisticated equipment. A typical condenser microphone only responds to pressure, not to velocity; but a scientific-grade acoustic transducer will be able to record pressure and at least one component of the 3-dimensional acoustic velocity vector. Sampling and recording this information would be critical to a total and exact re-synthesis of the wavefield. And as always, I will re-emphasize the caveat; the goal is to mathematically model the physical effect to some specified level of accuracy/detail. We can always make a more complicated model to account for the 99.999999...th percent effects. Most human ears only "hear" or perceive stereo sound (i.e., two pressure-waveforms) sampled at 40 kHz, so it's not necessary to record more data. Nimur (talk) 18:12, 15 April 2011 (UTC)

Nimur, your technical understanding is way beyond mine. This brings up a question I've wondered about. Human hearing range doesn't normally reach much beyond 20 kHz (if that), and one often hears how it is therefore unnecessary to record or reproduce frequencies above 20 kHz or so. But I wonder--is it possible, or even common, for frequencies above 20 kHz bouncing around a room after, say, a gong is struck, to interact with other waves bouncing around, with the room itself, etc, such that lower, audible frequencies are produced and heard? Something akin to a resultant tone (or better, combination tone), but at the high end? I've long wondered if this is possible and if so, how common it is. And if so, whether it would have an effect on music with lots of very high overtones/harmonics--making it sounding richer live, with all those ultrasonic waves existing and bouncing around, than recorded, with no ultrasonic frequencies in the first place. ? Pfly (talk) 18:55, 15 April 2011 (UTC)

Certainly I feel that way, but my preference is actually for the recording. The clash of real cymbals in the highest frequencies is just so painfully loud that it cuts up the rest of the song. Wnt (talk) 23:37, 15 April 2011 (UTC)

Regarding technical understanding... Anyone can understand mathematical descriptions of physical phenomena if they spend enough time analyzing them. I have spent, and continue to spend, a lot of time thinking quantitatively about daily mundane things. It also helps to have academic or formal training in physics or mathematics. Anyway - regarding wave mixing. It is a fundamental assumption of linear system theory that frequencies are preserved by systems. In other words, the simplest model forbids a high frequency from "reflecting" and producing a lower frequency. On the other hand, a nonlinear wave model (such as the elastic wave model) does permit nonlinear, frequency-altering interactions. For most purposes, the amplitude of such an effect is very small, so we safely ignore it for day to day ordinary sound waves. I can think of at least one case where we cannot ignore these effects! A broken loudspeaker results in a nasty rattle that is nothing like the intended waveform; what is happening is that energy is coupling nonlinearly into the torn paper cone of the speaker, which is then buzzing and sounding awful! Nonlinear acoustics are therefore not merely a theoretical concoction of bored physicists! Nimur (talk) 02:25, 16 April 2011 (UTC)

I question the premise that they sound different, if the sound is recorded and reproduced by high-fidelity media. If they are all that different, then perhaps you need to buy better microphones, recorders, amplifiers, or speakers. Edison (talk) 03:46, 16 April 2011 (UTC)

They certainly aren't exactly the same. The question is if they are enough different that you can tell. StuRat (talk) 03:49, 16 April 2011 (UTC)

And that's the entire premise behind sampled audio and lossy sound compression technology - when recorded properly, essentially no human can perceive the difference between the live and the recorded sound. Nimur (talk) 05:59, 16 April 2011 (UTC)

...except for those blessed with one or more Golden ears, such as everyone who writes in an Audiophile magazine. Years ago a loudspeaker amnufacturer (was it Quad Electroacoustics?) gave a stage demonstration showing a string quartet. Midway through a music piece the musicians got up and left the stage, leaving only the loudspeaker that had really been producing the music all the time. Cuddlyable3 (talk) 12:13, 16 April 2011 (UTC)

The "live vs recorded" demo would be more impressive were it not for the fact that Thomas Edison did the same thing with acoustically recorded "Diamond Disc" records and his phonograph 1915-1925. It was reported that the audience could not tell when the musician was performing and when the phonograph was playing[1], [2]. The repertoire was generally limited to solo voice or solo instrument, so there was no need to reproduce the high sound pressure level of a brass band or orchestra, and the soloists may have avoided loud sounds. For an overview of "live vs recorded" tests over the years, see [3]. Edison (talk) 19:50, 16 April 2011 (UTC)

My impression is that a lot of people who think/claim they have golden ears don't, they just don't ever actually test their alleged ability to hear differences properly (e.g. with an ABX) but often still make the claim they can hear these differences. I can't say if this applies to audiophile magazine writers in particular, but again my impression is a lot of the magazines don't publish or use ABX or other more scientific testing methods and are willing to promote things like Monster cables usually with questionable technical advantages (worse of course are those that promote cables and other things that are supposed to provide better output despite the fact both are capable of providing the exact same digital signal). I'm not of course denying there are people who can hear differences that many others can't and there are of course audiophiles who do approach their hobby/whatever from a more scientific viewpoint, e.g. those at Hydrogenaudio Nil Einne (talk) 21:40, 16 April 2011 (UTC)

A string quartet plays quite low frequency sound - only a few sorts of things go over 20,000 Hz. One thing about very high pitched sound is that it's almost line of sight. Back in the 1980s it was easy for people to tell because all the TVs and monitors and videocameras consistently made a lot of this sound when on (sometimes uncomfortably loud in the case of "hidden" anti shoplifting cameras) - I don't think they do this nowadays, though I'm not sure because I'm also losing high-pitched hearing with age. See ultrasound. Wnt (talk) 04:08, 17 April 2011 (UTC)

A cathode ray tube screen still makes the high-pitched hum, but these are becoming rarer. Flat-screen TVs make a lower-pitched whirring sound, as far as I can tell, but I may be picking up the sound from the digibox that switches on at the same time. 86.164.75.102 (talk) 13:23, 17 April 2011 (UTC)

Even before flat screens the monitors seemed to get progressively better about the noise. I think the scan rate, especially frames per second was related to it, but I don't think it was the only factor. Wnt (talk) 00:55, 19 April 2011 (UTC)

two questions about electromagnetic waves

1.I have no imagination of photons as particles. I mean we can think of electrons as particles moving (or present) in regions called orbitals. do photons have a certain measurable "area of influence" or something like that?

2.In the emission or absorbtion spectrum of different elements,there are lines that show different transitions in energy levels.well, according to the formalae, each kinds of transition must produce a single kind of electromagnetive wave(E(n2)-E(n1)=hf). but when we look at the spectrums there are neighbours of wavelengths, not "exactly one" wavelength.I mean although the neighbourhood may be small, but it is still a neighborhood, otherwise we couldn't see it.

It's REALY hard for me to ask such questions in English, so I'm sorry and I hope you understand what I mean.Thanks! —Preceding unsigned comment added by 178.63.158.171 (talk) 16:55, 15 April 2011 (UTC)

For question #1, a single photon's "area of influence" is probably best thought of by defining a particular fall-off in the electric field intensity (or, for one single photon, the probability of measuring an electric-field at a particular intensity). There's going to be a "soft edge" for the photon; for any radius r from the photon's "center location," there's going to be a lower probability that the photon can interact with anything. This is described mathematically as a wave packet, which has finite extent in space and time. (We also have wellenpaket in German).

For question #2, see hyperfine structure (also available in German at Hyperfeinstruktur). Nimur (talk) 17:57, 15 April 2011 (UTC)

(EC) The answers to both of your questions involve the Heisenberg uncertainty principle.

Similar to how you can't say with certainty exactly where within an orbital an electron is, there is also a "fuzziness" as to where exactly a photon is. For example, if you shine a monochromatic light on a circular aperture, the light intensity behind the aperture will form a diffraction pattern known as an Airy disk, rather than a solid circle as one would expect if each photon traveled along an infinitely thin line.

Spectral lines not having zero width is due in part to excited states having a finite lifetime, which causes an uncertainty in the emitted photon's energy, as explained in Uncertainty principle#Energy-time uncertainty principle. Red Act (talk) 18:18, 15 April 2011 (UTC)

The German articles on the topics I mentioned are Heisenbergsche Unschärferelation, Beugung (Physik), Beugungsscheibchen, Spektrallinie and Linienbreite. Red Act (talk) 18:34, 15 April 2011 (UTC)

An other reason for widening of spectral lines is the Doppler effect due to the relative motion of the atoms that emit the photons due to thermal motion. Dauto (talk) 03:24, 16 April 2011 (UTC)

Discovery of rimantadine and references

In the past few days a history section was added to the rimantadine article. An anon-IP editor has added several references to verify the claims, but I am having difficulties verifying the claims, mainly of two reasons: 1) Chemistry is not something I am familiar with. 2) English is not my native language. Can someone with knowledge in chemistry please have a look at the references, and explain whether the references really verifies the claims or not, and perhaps explain why – since I, as a layman am not able to understand the references given at first sight. Thanks in advance. ^Talk/♥фĩłдωəß♥\^Work 18:02, 15 April 2011 (UTC)

Head rotation

Is it theoretically possible by surgical means to make the human neck safely rotate the head at 360 degrees or so (bearing in mind that it's not fatal in some birds)?--89.76.224.253 (talk) 18:16, 15 April 2011 (UTC)

No, this is not possible. The structure of our spine and neck is different from birds. Nimur (talk) 18:20, 15 April 2011 (UTC)

I'm pretty sure birds don't rotate 360 - they probably do 180. Ariel. (talk) 19:23, 15 April 2011 (UTC)

OP, what birds are you referring to? The usual example of a bird rotating their head to an extreme is the owl. And according to our article on the owl, "Owls can rotate their heads and necks as much as 270 degrees in either direction". Or do you mean 360 from one extreme to the other? Dismas|^(talk) 19:41, 15 April 2011 (UTC)

Screen-width Problem in computer

The ratio of width-to-height in new laptops is considerably different than traditional desktops. Does that mean that everything will appear deformed i.e. stretched along horizontal axis ? —Preceding unsigned comment added by 124.253.130.232 (talk) 21:52, 15 April 2011 (UTC)

Please don't ask the same question on multiple ref desks, especially when it has already been answered on one of them (Computing). Looie496 (talk) 22:16, 15 April 2011 (UTC)

Seafloor river bed in Toyama Bay, Honshu from Google Earth

Hi. On Google Earth, a long river valley appears carved onto the seafloor leading out from Toyama Bay in Japan, extending toward the bottom of the Sea of Japan. It is most likely not volcanic or tectonic in origin, and also does not extend from the mouth of the Shinano River, which lies closer to Niigata farther northeast. It appears to wind itself around the eastern edge of some continental shelf and steadily drops off in elevation on the seafloor, tracing a path from a reverse-delta toward the north. In fact, the main undersea channel appears to be a merger of the Shō River and the Jinzū River. The channel appears to have a lower sea bottom elevation on Google Earth than its surroundings and stretches for about 590 km before ending at the edge of a deep basin, at which point it is closer to Sapporo than to Tokyo. At the point equidistant between the two cities, the elevation at the bottom of the channel is about 2800 metres below sea level, It is noticeably longer than any other apparent seafloor river channel in the area and may be one of the longest in the world. Also, the channel appears to bypass the edges of a few underwater volcanoes, suggesting that sea levels were lower or the river has enough power to carve this deep channel even when underwater. At various points, the channel itself lies an average of 145 metres lower than the surrounding ocean floor. Any information on what was responsible for creating this channel, whether it was an ice age event or a phenomenon similar to the conditions today, and how this particular locale compares to any similar channels worldwide? Thanks. ~AH1 ^(discuss!) 22:59, 15 April 2011 (UTC)

Apparently it's called the Toyama Deep-Sea Channel.^[4] Wnt (talk) 00:02, 16 April 2011 (UTC)

I don't know about that one specifically, but our article on submarine canyons give information (or at least speculation) about how such things form. Looie496 (talk) 00:05, 16 April 2011 (UTC)

"Submarine canyons are well developed around the Japanese Islands. Three major large-scale submarine canyons are the Kushiro Canyon, the Toyama Deep-sea Channel, and the Boso Canyon. The Kushiro Canyon greatly encroached on the continental shelf and deeply eroded the continental slope. This characteristic is markedly different from other canyons. The Toyama Deep-Sea Channel is characterized by the length of over 500 km, considerably meandering, a vast submarine fan, and well-developed submarine natural levees. The Boso Canyon has significantly incised meander 100 km in length."^[5] Looks like 3.6 million articles just aren't nearly enough. ;) Wnt (talk) 00:10, 16 April 2011 (UTC)