Wikipedia:Reference desk/Archives/Science/2014 July 23

Science desk
< July 22	<< Jun | July | Aug >>	July 24 >

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.

July 23

Are all spontaneous processes exothemic?

According to the article Spontaneous process, a spontaneous process should be exothermic as said by the first lines

"A spontaneous process is the time-evolution of a system in which it releases free energy (usually as heat) and moves to a lower, more thermodynamically stable energy state."

But I think the statement is wrong. Because ice melting is an endothermic but spontaneous process at room temperature or even at or above 273 K, some spontaneous processes should be endothermic. And when glucose is dissolved the water becomes cooler. But glucose melting is spontaneous. Should enthalpy get confused with entropy? When ice melts ice absorbs some heat from the surrounding. Please give your comments. A user has already pointed out this in the article's talk page. But there were no response for 3 years. I asked this here to get some response.--G.Kiruthikan (talk) 10:47, 23 July 2014 (UTC)

Further down the article, it specifies "free energy" to refer to Gibbs free energy. Plasmic Physics (talk) 11:47, 23 July 2014 (UTC)

OK. I understood. Sorry.--G.Kiruthikan (talk) 06:04, 26 July 2014 (UTC)

No need to apologise. Plasmic Physics (talk) 06:08, 26 July 2014 (UTC)

Natural Selection

Could it be argued that those who get bullied or those who end up single for life, not by choice but as a result of shyness or women not finding them attractive etc, are experiencing a process of natural selection, favouring those who are confident and can stand up for themselves?

This is called sexual selection Graeme Bartlett (talk) 11:48, 23 July 2014 (UTC)

It ultimately benefits humanity, by ensuring a continued supply of Wikipedia editors. 80.43.213.39 (talk) 12:03, 23 July 2014 (UTC)

wouldn't that be called Wikipedia selection,

There is nothing good or bad about natural selection, it is just what happens in the circumstances. Our circumstances are that we have laws and a belief in rights and basically a complex society. What that will lead to is anybody's guess. For instance currently deaths by firearms and deaths from car accidents are at a rate that very probably will have an effect on our genome but we can't guess what it will be. For all we know the characteristics that lead to bullying may be preferentially removed by them being involved more in such fatalities. I'm not altogether sure though if that was the case that we should then say we should remove car and gun safety laws. Society is the environment for evolution now. Dmcq (talk) 12:48, 23 July 2014 (UTC)

I should apologize in advance for blatant speculation, but just thinking out loud... I recently saw a study in which the frequency of homosexuality was listed as only 1.6%, and only 2.5% even when bisexuality is included. The figures for this used to be much higher, starting with the famous Kinsey Reports figures of 10% and 37%, but every decade they seem to get lower. I would suggest that this may not be experimental error; rather, the effect of anti-gay oppression may have been to pretty much force people to reproduce who wouldn't have otherwise. (The irony of this effect should not much shock those who have seen obscure plants like Cannabis turned into household words over the past century) The present figure is still greater than the (somewhat inflated) figure for autism, so by this model I'd expect the frequency to drop several-fold more over the next few generations. Anyway, if any of this is true then I would expect any other genetic trait whose effect on reproductive success was concealed by social compulsions to reproduce would likewise be uncovered by the same change in attitudes, and to be undergoing a similar reduction in frequency. Wnt (talk) 12:54, 23 July 2014 (UTC)

it only takes a single exception, you could be a lonesome reject for life except meet 1 bookworm after 17 years of nothing post-puberty, and it's literally all it takes to procreate, get married, etc. Think about your parents :) Or their parents. In fact you come from a direct line of pushovers who all successfully procreated, going back tens of thousands of years. You'll probably end up married yourself (chances are). So zero evolutionary pressure here. 213.246.165.17 (talk) 14:42, 23 July 2014 (UTC)

• Anything that influences the probability of ultimately having children is going to create selection pressure. But it is important to realize that natural selection operates on a time-scale of thousands of years, whereas human social systems affect us on a time-scale orders of magnitude faster. The consequence is that natural selection is almost never a significant factor when thinking about how our social systems ought to be structured. Looie496 (talk) 14:53, 23 July 2014 (UTC)

That's not true though. Peppered moth evolution changed the color of the peppered moth from specked white to totally black within 50 years - and back again within another 50 years. Evolution can produce an effect within just one generation if the selection pressure is sufficiently harsh. A sufficiently decisive social movement could easily wipe out a particular single-gene variation in just a few generations.

The thing that makes the "lonesome bookworm" type succeed is that there are other lonesome bookworms of the opposite sex who really couldn't stand to hook up with someone with a loud, outgoing, sport-enthusiast "Type A" personality. So there will always be enough pairings to ensure that the type survives. The question worth asking is how long it'll be before "homo sapiens geek" become a separate species from the "homo sapiens jock" ? SteveBaker (talk) 16:59, 23 July 2014 (UTC)

If the jocks get the pretty girls and the geeks do the technology then it should have been the Eloi in the Time Machine who were the brutals not the Morlocks. ;-) Dmcq (talk) 17:27, 23 July 2014 (UTC)

I'd probably think a bit more about generation time of moths vs. humans before claiming that Looie said something wrong about evolution. Nothing about natural selection has an absolute time scale-- the widely variable and natural unit of time is the generation, because that's where mutation and selection happen (for simplicity ignoring the case of horizontal gene transfer). Peppered moths have 1 generation per year at minimum, humans of course go much slower. So 50 human generations may well be enough time for some of us to get spots or whatever, but that's still a much longer time scale than most of our social structures have. You are of course correct that strong selection can have a decent effect in just one generation, but that would be very strong selection indeed, and I think we'd all agree that that is not what this question is about. SemanticMantis (talk) 18:25, 23 July 2014 (UTC)

I'm a geek. My wife is beautiful. HiLo48 (talk) 20:01, 23 July 2014 (UTC)

--Bowlhover (talk) 23:35, 23 July 2014 (UTC)

I'm a geek. My wife is also a geek. We both agree that intelligence, wit, honesty, gentleness - and most of the other thousand attributes that make for a good marriage - *all* trump beauty. (Although, as it happens, she's also beautiful!) SteveBaker (talk) 04:25, 24 July 2014 (UTC)

Is there any evidence that there a large pool of 50 year old male "geeks" who yearn to be fathers, but haven't succeeded in fathering offspring? Cullen³²⁸ Let's discuss it 05:10, 24 July 2014 (UTC)

Probably not - but I guess the real issue is whether there are a large pool of 50 year old geeks of either gender who don't yearn to be parents? SteveBaker (talk) 16:07, 24 July 2014 (UTC)

But realistically, for this kind of 'geeky' person to evolve out of existence, you'd have to assume that 'extreme-geekiness-leading-to-not-having-kids' has a genetic basis - and that it's a simply-inherited one that comes from just a few genes that aren't 'mission critical' for anything else. It could easily be that you turn out to be a geek because of some complex mix of genes, then maybe you have one great grandmother had blue eyes, one great grandfather had a club foot, at least one other great grandfather was NOT color blind AND another great grandmother carried the genes for both sickle-cell anaemia and lactose intolerance. If the situation is that complex then the trait of geekiness isn't going to pass directly from one generation to the next and it would be unlikely to vanish from the gene pool no matter how bad we geeks are at reproducing.

Sickle-cell disease is another good model to consider. If you have a sickle-cell gene from both mother and father, then (without modern medicine) you're unlikely to survive long enough to reproduce - you'd think that this defect would have vanished from the gene pool tens of thousands of years ago. BUT having just one of these genes confers immunity from malaria. So the benefit of being a 'carrier' of sickle-cell outweighs the risk of producing sickle-cell children that die early...so the gene stays in our gene pool.

Complex genetic outcomes are much more complicated. For example, if there is a genetic component to living on into old age - that couldn't directly affect the gene pool because the gene doesn't improve reproductive success because it only does something after you're too old to have kids. If that were all there were to this, then humans would drop dead the moment they were no longer able to reproduce (that happens with LOTS of other animals). But in fact, we're social animals and an entire village gains reproductive success if there are some older people around to pass on knowledge and so forth. So the presence of the gene for living longer has value to the reproductive success of the population, and it is selected for to some degree.

If small communities see reproductive benefits from having geeks around to figure out why their PC won't run Candy Crush anymore - then the genes for geekiness will remain in the population

even if no geek ever has a child.

SteveBaker (talk) 16:07, 24 July 2014 (UTC)

can a clear substance become opaque upon electrical impulse?

Hi,

Can a transparent substance become opaque upon electrical impulse? For example glass that can become frosted (or in any other way not-transparent) upon electrical impulse?/213.246.165.17|213.246.165.17]] (talk) 13:38, 23 July 2014 (UTC)

An LCD sort of does this. RJFJR (talk) 13:46, 23 July 2014 (UTC)

Why do you say "sort of"? Would this work to darken a room for example? (Like, instead of shades.) Is the level of transparency adjustable? 213.246.165.17 (talk) 13:53, 23 July 2014 (UTC)

See Smart glass#Electrically-switchable smart glass. I don't know whether any of these technologies can be used to make glass perfectly opaque (and I didn't read enough of the article to find out), but your mention of "frosted" implies at least some translucency. Deor (talk) 14:04, 23 July 2014 (UTC)

If this is just an LCD, what keeps someone from just bulding up thousands of layers of LCD screens for a true volumetric display? 213.246.165.17 (talk) 14:24, 23 July 2014 (UTC)

Most LCD screens (not the ones referred to above) use two crossed polarizing filters so there's no way I can see to stack them meaningfully. Transparent OLED is probably the closest to what you want but it is nowhere near transparent enough and can't give dark colors for points. Dmcq (talk) 14:54, 23 July 2014 (UTC)

Display device lists various 2D and 3D display types. Dmcq (talk) 15:05, 23 July 2014 (UTC)

I've thought of this idea myself, as a way to control solar heating of a home. Because of the cost and lack of total transparency I rejected LCDs as the way to achieve this goal. I believe they do make electrically operated miniblinds between panes of glass, and making on side reflective (either white or silver), and the other side black should provide the maximum difference between open (when you want both heat and light), closed with black side out (when you want heat, but not light), and closed with reflective side out (when you want neither heat not light). Certainly it wouldn't block 100% of the light, though. An advantage over LCDs is that it would only use electricity when moving the mini-blinds, and very little then, relative to LCDs which would use electricity the whole time they are on. Also, LCDs don't seem able to withstand wide ranges in temperatures, which you should expect at windows.

Another option might be electrically operated window shades. If those could also be placed between panes of glass, hopefully the edges could be held firmly in place, so light doesn't leak around them as it does through a mini-blind. StuRat (talk) 15:28, 23 July 2014 (UTC)

What would be a way to get a true volumetric 3D display using such technologies? (i.e. that changes depending on the angle you view them at, where one layer can occlude another layer, etc) 213.246.165.17 (talk) 15:55, 23 July 2014 (UTC)

Well, if you DID have a material where individual pixels could be turned opaque or transparent on command without doing polarization tricks - then you could make a bunch of layers and have some sort of voxel-based 3D display. But for even a very low resolution display (say 640x640x480) you'd need 480 layers! Unless these layers are amazingly cheap, you'll have a fairly crappy display costing a good fraction of a million dollars!

Worse still, a 640x480 2D image only looks reasonably acceptable because it's typically "antialiassed" to get rid of the jaggy edges by making fuzzy transitions from regions of one color to the next. You can't really antialias a volumetric display - so even a display with hundreds of layers would be very 'blocky' looking.

I'm not sure if it's currently possible to electrically opaque individual pixels in a panel - but no matter because unless there is some kind of major technological breakthrough it would be prohibitively expensive. Another issue would be the reproduction of color - and to make convincing 3D, you'd probably need programmable shininess, partial translucency and other surface attributes. The way that light interacts with a surface is critical - and these blocky 'voxels' aren't going to do it right!

SteveBaker (talk) 04:19, 24 July 2014 (UTC)

How about a technology where you have one solid block of normally transparent material, that reacts to a certain level of radiation by becoming opaque, then it fades back to transparent in, say, a hundredth of a second after the radiation level drops below the required level ? The device would then aim multiple narrow beams of the appropriate frequency of radiation at each voxel, in turn. Obviously you would want a form of radiation that's not harmful to humans. Creating electromagnetic interference could also be a problem, but you might put the whole thing in a Faraday cage to prevent that. StuRat (talk) 04:37, 24 July 2014 (UTC)

Our volumetric display article briefly touches on this method - using lasers to create points of plasma in air. Katie R (talk) 13:08, 24 July 2014 (UTC)

That only works because the air remains transparent enough to transmit the laser light. What you see with those displays is a ghostly glowing object...it doesn't look solid. When you demand that the voxels become opaque - that would (presumably) block the laser light - so you couldn't make solid-looking objects that way. This is the real problem with most 3D display technologies - it's easy enough to generate those kinds of ghostly glowing images that you can see right through - but they don't look real because you can see distant parts of the object through the nearer parts. There are plenty of ways to make those kinds of displays - holograms, vibrating mirrors, spinning 2D display panels...etc, etc. SteveBaker (talk) 15:13, 24 July 2014 (UTC)

Yes, I envision using a form of radiation other than visible light, and the activated voxels should then remain transparent to that wavelength. StuRat (talk) 17:25, 24 July 2014 (UTC)

Holy cow, I disagree with every single assumption you've stated. In the first paragraph, you assume that 480 layers and the electronics to drive them would cost a "good fraction of a million dollars", but even at $2000 for the kit (which I would not say is a "good fraction of a million dollars") that is $4 per layer, whereas an 640x480 LCD layer plus everything to drive it (zero economies of scale) costs around $5 at scale, there is no way you couldn't put 480 of them on top of each other at significantly less, at volume manufacturing scales. So this assumption is just totally wrong.

In your second paragraph you say that 3D voxels couldn't be antialiased, but why the hell not? Here is antialiasing with legos: http://brickplayer.com/blog/wp-content/uploads/2010/05/lego_mosaic_anti_alias_3.jpg - why couldn't the same thing be done in 3D? (I realize that most lego sculptures don't use this effect, as they like the blocky effect and don't use that many shades of color anyway - but as a long as the transparency was't binary but had degrees, you could do this.)

In your third paragraph, you assume that voxels have to have real-world properties to "look right". But no pixel has real-world shininess or reflection, and things still look OK on monitors, despite being reduced to a few brightnesses of a few colors. Sure it won't look like the real world, but that is no objection to making voxels. It will look like something.

Finally I will say that despite your analysis being totally wrong, based on clearly false assumptions, I will grant that layers of OLED's or similar is probably NOT the way to do it. Then again, I doubt anyone in CRT era would have thought physically putting 4 million physical pixels - each containing a complete mix of colors - into a 15 inch monitor at 227 dpi was in any sense viable - surely scanning across is the only viable method. And yet that's exactly what a Retina Macbook contains, in a tiny light form factor, as opposed to a phosphor screen and scanning laser. So just because something seems ridiculously infeasible does not mean it is not *the* solution in the future.

All that said, I don't necessarily think layers of LCD's or OLED's or anything else is the correct way forward in the future with 3D volumetric display. I was more interested in the theoretical possibilities, and bringing up technical objections at this thought-experiment stage is unwarranted, especially when the analysis isn't even correct. 213.246.165.17 (talk) 12:50, 24 July 2014 (UTC)

Sadly, my job is doing this kind of thing...so I happen to know that you're wrong. You can't just stack panels like that because they emit heat and the heat buildup in the inner layers would kill you. You need some way to communicate the heat from the inner layers to the outside world. Flat panel displays need bulky connectors and external drivers that don't stack up so well and DO add to the cost. Just the panel itself is only a small fraction of the cost here. I can't think of any consumer-grade gadgets with 640x480 displays that cost $5.

But let's say you're right...$5 display panels aren't THIN. If we want to build a 640x640x480 display that's a few inches across - then we need each layer to be about 1/100th of an inch thick. Do you have any clue how much a 1/100th inch thick LCD panel costs? You want to guess? Please don't - I know the answer - and that's why the cost of your layered display *WILL* be a substantial fraction of a million dollars. Sure, you could build a 640x640x480 display that's (say) 48" thick with 1/10" thick layers - but still, you don't get 1/10" thick panels for $5. With $5 displays, you find them to be around 1/4" thick...so with 480 of them stacked up, you have a display that's 120" thick. Well, if you want a roughly equal resolution/dimensions in X, Y and Z - you now need $5 panels that are 1/4" thick and 10' x 10' wide! I don't think so! So quit talking without thinking.

Yes, antialiassing is a problem. You only want those semi-transparent voxels at the profile edge of the objects - and that's view-dependent. If this is a view-independent display then the objects would have to be 'coated' in all directions with semi-transparent voxels, blurring and fuzzing absolutely everything. When you actually look into this in detail, it doesn't work. Your link to a 2D lego mosaic proves nothing. Your lack of understanding about the issues of light reflection in a 3D object require a longer explanation than I really want to give here - but suffice to say that a 2D representation of a 3D object has already undergone lighting calculations (well, unless it's a very flat cartoon) - and therefore the 2D display surface itself doesn't have to undergo illumination. But a true volumetric 3D object would have to interact with the room lighting or it can't be view-direction independent because the way light reflects off of a surface depends on where you are viewing it from. Since it's really just an array of little cubes, their faces wouldn't align with the light direction the way a real object would - so the resulting lighting would be decidedly strange. Perhaps that would be OK for some kinds of limited applications - but for general 3D display, it would pretty much suck.

Your analogy with the growth of 2D display resolutions isn't exactly correct - when you double the resolution of a 2D display, you only need four times the number of pixels. But pushing the resolution of 3D displays would require not just 8x the number of pixels but it would increase the number of layers that you need...DECREASING the thickness of each layer in proportion to the required resolution. While flat panel displays have shrunk somewhat in thickness over the years - it's not anything like the shrinkage you'd need here. There are no display technologies that are thin enough to generate a small display with reasonable Z resolution - and no thicker displays that are large enough in the X/Y direction to make a large volumetric display. Since the demise of the CRT, the technological improvements to make displays THINNER or LARGER lag by far the technologies to push the resolutions up. It turns out that there are very good laws-of-physics reasons why that is.

So, no - you're the one who is wrong in almost every regard. SteveBaker (talk) 15:13, 24 July 2014 (UTC)

The other issue that stood out to me is the issue of light transmission through each panel. They aren't perfectly transparent, so voxels far away from the viewer will be foggy. Minor imperfections in the plane cause the layer to work like a lens, distorting the path of the light through it. It doesn't matter on an LCD display where the only thing behind the panel is a backlight, but a panel behind 100 layers will be distorted and blurry. Both of those can theoretically be engineered around to make things acceptable, but will obviously cost a fortune. Katie R (talk) 15:54, 24 July 2014 (UTC)

I don't understand why a view-independent object is supposed to have to interact with the room lighting (outside the 3D display) whereas a monitor (normal 2d one) doesn't... We're not trying to trick someone into thinking that a real object is there - you NEVER think your monitor actually has a real object inside it - maybe an ant crawling across it or a hair on it or something - we just want something there, clearly different from a real-world object. Basically, in the sense that a set of pixels is an "approximateion" (a very poor one) of a piece of paper, in that it has different splotches of color at various places, is there a set of voxels that is physically an "approximation" of a paper statue, in that it has different splotches of color at various places, but they also occlude each other? A piece of people is approximated by rows and columsn of pixels. Could a similar theoretical solution approximate a 3d statue, by literally occluding parts of the statue? (such as the inside - you can't see inside it). Even theoretically? Even with a 9x9x9 matrix? Even if you made it very large? And if so, by what chemical or electronics means? 213.246.165.17 (talk) 16:20, 24 July 2014 (UTC)

To be clear, we are talking about the theoretical basis, not current technical limitations. For example the "plasma created by lasers" clearly can't occlude other plasma points (the back of the object) - so that doesn't work. 213.246.165.17 (talk) 16:22, 24 July 2014 (UTC)

The thick (120") display might even be promising.

Look at the 16/9 inflation. All manufacturers joined the 16/9 craze, not because of the applications in computing (there are VERY few if you don't include watching movies on a PC) but because a 20" diagonal doesn't indicate actual display area. With a more skewed ratio, you can achieve the same diagonal with less area. Now imagine the potential of a 3D display with a 120" triagonal and hardly any cubic inches to go with it.

The mere absence of these displays is proof that they are not that easy to make. - ¡Ouch! (^{hurt me} / _{more pain}) 08:24, 25 July 2014 (UTC)

Steve and Katie, if we back away for a moment, one of the major things in this thought experiment is that it started by assuming that the polarization problem could be fixed from a modular-2 system to a higher dimension of modular arithmetic so that the different layers could all be visible. But of course polarity doesn't work like that. So for starters - if we go back to this assumption: 1) is there a competing technology that lets you go from opaque to transparent? 2) can "that" be filled into a cavity, so that it has some true volume as well? What are our options here, theoretically... (not practically.) Even at 16x16x16 pixels, I haven't seen a demonstration of this (true occlusion), even with terrible optical properties on the logically 'transparent' parts. 213.246.165.17 (talk) 16:02, 24 July 2014 (UTC)

It obviously could be done. Imagine that the display is the size of a large building, each voxel is a 1'x1'x1' glass fish-tank with a thin, clear plastic hose leading into the bottom of it out to a large array of pumps and storage tanks. You could pump black ink into the tank to make it opaque or drain it out to make it be transparent. With sufficiently fine hoses and sufficiently clear glass - this might make a low-res opaque/clear voxel display - with a refresh rate of maybe once a minute and a pile of several million ink pumps to keep working! That works (at least as a thought-experiment). Instead of ink, you could use water and mix in colored powders with four different grain sizes (one size for each of Cyan, Magenta, Yellow, and White) are infused into water as it's pumped into the tank - then as you drain the tank for the next update cycle, you filtere the powders out by four filters with different hole sizes so you can recycle the powders and the water for the next refresh cycle.

Using microfluidics, maybe you could somehow make this work at a small scale - possibly small enough to make it usable for a more sane size of display...maybe. So at least in theory, there is a way to do it.

Stacking current-generation panels doesn't really work because the panels have bezels around them to contain the scanning electronics, etc - so you wouldn't be able to see through four out of the six sides of the display cube...so that's not going to work...but we can certainly imagine at least one thing that could - at least in theory.

SteveBaker (talk) 16:25, 24 July 2014 (UTC)

SteveBaker - your thought experiment is fascinating. I like this solution. Basically, my original question is, can we start with filled but transparent fish-tanks, filled with something that electrically becomes opaque? Then you don't have to drain it to empty it, you just cut power to a tiny filament. To fill back up, you just turn power back on. Does such a materials exist? What is the closest approximatino?

Also, Stevebaker, you seem to have a really good grasp of optics. If we return to your super-slow, totally "elevator in space" (like relativistic thought experiment) level of impractical building-sized voxels-of-fish-tanks... viewed from a distance, what difference is there between that and a true 3D object? (supposing you could pump any material into the fish tanks)? Where would it look "off", other than needing exceptionally transparent glass, and being unable to have real-world reflective and refractive and transparent properties, except if you pump such a material into the voxels? Any other clear differences from a true 3D object, as seen from afar? (again, purely at the thought experiment level, nothing practical here. just wondering what physical properties come up.) 213.246.165.17 (talk) 16:45, 24 July 2014 (UTC)

Reflectivity. In the real world, the light coming from an object to your eye is composed of several kinds of reflection. A simple approximation that's used in computer graphics is to split it into three parts:

• Ambient - light which hits the object more or less equally from all directions (like the sky) and is reflected off more or less equally in all directions - typically after the object has absorbed some of it - so it takes on the color of the actual object itself.

• Diffuse - light which bounces off of the object in an amount that depends on the orientation of the surface to the light source - but which is scattered off equally in all directions. This is what gives objects their characteristic three dimensional appearance. Without diffuse light, a cube would look exactly like a flat hexagon. Diffuse reflection also takes on the color of the object itself.

• Specular - light which bounces off of a smooth surface without being absorbed by it and reflects off of it in a fairly narrow cone of directions. It keeps the color of the original light, not the object itself. Specular light changes in appearance depending on where the observer is positioned relative to the object.

(This is a gross over-simplification - real materials are much more complex than that - but even getting our fishtank display to do this is impossible).

The problem we have with our "fishtank" display (neglecting the glass walls of the tank because this is a thought-experiment) is that the colored water is shaped into little cubes. Each face of those cubes is oriented in either the X, Y or Z direction. But if you're trying to represent (say) a sphere, then none of the faces of the voxels are ever pointing in the right direction to catch the light the way a sphere would. Ambient light will be OK - but diffuse light won't because there will be places on the sphere where the true spherical surface was at right angles to the light - and should be shining brightly in diffuse light - but isn't because the little cubes are all turned at around 45 degrees to the light where it's not sufficiently strongly reflected. Specular light will be even worse - as you move your head around, those shiney highlights should 'follow you around' on the surface of the sphere - but they won't because the only specular reflection is heading out in one of three distinct directions that's determined by the orientation of the display volume to the light source.

Now, one of our earlier posters is probably about to jump in and say "BUT THAT'S NO DIFFERENT FROM A 2D DISPLAY!!!"...which demonstrates a lack of understanding of what's going on here.

We humans know when we're looking at a photograph or a painting or an image on a TV screen or computer monitor - and it doesn't look real. No matter how cleverly it's done, it doesn't look like looking out of a window does unless you VERY carefully allow for where the viewer is standing and what the lighting in the room is (such as is done in Trompe-l'œil paintings and such). One of the reasons people love the Oculus Rift display system is that it tracks the position of your head and the software can adjust the lighting of the 2D scene to accomodate that. We could do calculated lighting on our 3D fishtank display - but getting people to accept that as a real thing would be difficult without doing all of the viewing in a darkened room with the voxels themselves emitting a very uniform light - and we wouldn't have a viewer-independent display anymore because you have to recalculate the specular light component as the view direction changed - which really defeats the object of having a 3D display in the first place!

SteveBaker (talk) 17:10, 24 July 2014 (UTC)

Even if it worked, it wouldn't give a realistic impression to more than one viewer, either. - ¡Ouch! (^{hurt me} / _{more pain}) 08:24, 25 July 2014 (UTC)

G2 vs Sgr A*

In April this year, the gas cloud G2 was due to collide with the black hole Sgr A* (or at least, when we would see the collision). However, the Sgr A* article has not been updated, and I can't (from a quick search), find anything about it, post-event. Was there anything to report about it? CS Miller (talk) 18:30, 23 July 2014 (UTC)

According to Sagittarius A*#Discovery of G2 gas cloud on an accretion course, observations during the expected time of the perinigricon in March found that G2 remained intact, most likely due to G2 hosting a central star. So it turned out to be kind of a non-event.[1] Red Act (talk) 19:03, 23 July 2014 (UTC)

Thanks Red Act. The paragraph Astronomers from the UCLA Galactic Center Group published observations obtained on March 19 and 20, 2014, concluding that G2 is still intact, in contrast to predictions for a simple gas cloud hypothesis and therefore most likely hosts a central star. does say that. However, the paragaphs before jump between the past and future tense, and reads like it has yet to occur, which is what threw me. Perhaps a copyedit is in order. CS Miller (talk) 19:40, 23 July 2014 (UTC)

Go for it. Red Act (talk) 19:46, 23 July 2014 (UTC)

Carbon auditing

Does carbon auditing, within a system boundary, normally require complicated mathematical calculations? — Preceding unsigned comment added by 176.254.33.42 (talk) 22:13, 23 July 2014 (UTC)

That depends a lot on the system. A coal-powered power station, should be pretty easy - you know the tonnage of coal going in - you know the tonnage of ash and debris coming out the other end - the difference will overwhelmingly be carbon - and you can calculate the amount of CO2 produced very simply.

However, if you wanted to figure out (say) the carbon footprint of a car factory - it's a horrifically complicated calculation because cars are made from a bunch of different materials, there is much processing of those materials and energy is consumed in hundreds of different ways.

It's not likely to be difficult from a strictly mathematical perspective - but the data gathering and the science involved may be a huge undertaking.

SteveBaker (talk) 04:07, 24 July 2014 (UTC)

I agree in concept that no single calculation is terribly complicated from a mathematics point of view, and that it depends a lot on what is being audited. For example, these people [2]. Have done cradle to grave life cycle analysis on the carbon cycles of their energy farm. There are several publications listed at the link above if anyone is interested in seeing how this works in modern practice. To sum up carbon accounting is conceptually relatively simple, and does not intrinsically rely on advanced math. However, in practice, it can become incredibly complicated. SemanticMantis (talk) 14:47, 24 July 2014 (UTC)

You can cheat a little. The cost of the final product is proportional to the energy consumed, and energy consumed is proportional to CO2 released. So to a first approximation you can just see how much the product costs - the cheaper it is the better for the environment (*ONLY* from a CO2 point of view). Major caveat: This does not include money spent to prevent or treat waste released to the environment (pollution), so this only works is you have similar environmental regulation. Ariel. (talk) 03:30, 25 July 2014 (UTC)

Beer cans are easier to squeeze when half frozen. Why?

At room temperature, you can squeeze a can of beer a little. Fully frozen, you can't squeeze them at all. Halfway though, when there's like 50% ice, 50% liquid, squeezing is a lot easier than when it's completely liquid. The biggest component of beer is water which expands while cooling, so why? Joepnl (talk) 23:25, 23 July 2014 (UTC)

It probably has something to do with Ductile–brittle transition temperature (DBTT) –(good luck making sense of the article).   —71.20.250.51 (talk) 00:20, 24 July 2014 (UTC)→[Nevermind]  —I sort of misunderstood the question. Rather than the can itself, it is the change in the resistive force of the contents. Hopefully somebody (else) can provide a better explanation. (?)   —71.20.250.51 (talk) 00:27, 24 July 2014 (UTC)

Well, it's almost certainly something to do with the pressure inside the can right? Water is incompressible either frozen or liquid so my guess is that the answer lies in the gas. The warmer the liquid is, the less gas it can hold, is that right? When you warm up fizzy drink, the gas escapes. So, the most gas is soluble when the liquid is at it's coldest, but not yet completely solid, then obviously mechanical forces of the ice prevent squeezing the can. Does that make sense? Vespine (talk) 01:29, 24 July 2014 (UTC)

I'm trying to find a ref for the claim that cold liquid holds more gas, which does indeed seem to be the case, but the top of my search is "straight dope" which might be mostly correct but not really what I'd call a scholarly source..Vespine (talk) 01:33, 24 July 2014 (UTC)

Wikipedia covers it at Henry's law#Temperature dependence of the Henry constant. Red Act (talk) 02:27, 24 July 2014 (UTC)

You can do an experiment too! Put on your official white lab coat, freeze one can of beer/soda and leave another out of the fridge. When you open the cans, you won't get the same initial "woosh!" of escaping gas from the frozen can - which is because the gas is staying dissolved in the ice and low-temperature water. Also, if you leave a can of soda in a hot car, the base of the can will dimple outwards because the pressure inside the can goes up so high as the CO2 comes out of solution at the higher temperatures (you might not want to test that assertion because the can may actually crack open and fill your car with a foamy, sticky mess). SteveBaker (talk) 04:03, 24 July 2014 (UTC)

Use a can of carbonated water instead of a sugary soda, whether trying either the freezer test or the hot car test. Back to the original question, water expands dramatically when it freezes, but not until then. Chilly liquid water is almost exactly the same volume as warm liquid water. Cullen³²⁸ Let's discuss it 04:52, 24 July 2014 (UTC)

And furthermore, the OP's condition where you have a mix of liquid water and solid ice happens at exactly the freezing point. However, if this is beer then the small amount of alcohol will act as an antifreeze and lower the freezing point of the water...so we're actually likely to be seeing temperatures below zero celcius without yet seeing the large expansion of the water as it changes state - yet still seeing the increased ability to dissolve CO2. So I bet this effect is even more pronounced in beer than in soda. However, soda has a complex mix of ingredients - I have no idea what (if anything) that does to the freezing point - so I could easily be wrong in that guess. But it does mean, for sure, that using canned carbonated water won't demonstrate the effect as well as beer. SteveBaker (talk) 14:43, 24 July 2014 (UTC)

The gas head at the top will also reduce in pressure, by about 1/3% per degree Celsius at room temperature, assuming the gas follows the ideal gas law. ---- CS Miller (talk) 09:07, 24 July 2014 (UTC)