Wikipedia:Reference desk/Archives/Science/2017 November 10

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.

November 10

Infrastructure vs building engineering

Why is it that engineers and construction specialists working on buildings tend to stick to buildings, and those working on infrastructure tend to stick to infrastructure? I see very few who cross between the 2. 00:23, 10 November 2017 (UTC) — Preceding unsigned comment added by 94.10.251.123 (talk)

Different organisations do different things. A person working for a municipality or city local government may design roads and bridges, but would not be involved with buildings. Also someone working for a large building contractor would just get buildings to design. An independent engineering consultant would build up a reputation in one industry, and then get work from that industry. Graeme Bartlett (talk) 05:56, 10 November 2017 (UTC)

The outcomes and detailing are very different. Buildings are inhabited spaces with a particular level of finish, space conditioning and unique functional requirements suited for continuous use and habitation by people. Infrastructure (bridges, tunnels, dams, water systems, waste treatment, highways, power systems) are generally not inhabited, at least not to the same degree, and involve a different kind of detailing suited to function, with habitation a secondary concern, or no concern at all. Habitable environments are subject to building codes, life safety codes, energy conservation codes and the like and are closely regulated by building departments. Infrastructure is mostly governed by engineering industry standards for function and durability. While there is certainly overlap, the two involve differing skillsets. As an architect, I've worked with both, and infrastructure (or "heavy construction") requires a different kind of information presentation and detailing than general construction. If you're not set up for it it's hard to do efficiently or well. The same applies to contractors, who have to organize on different scales with different trades and equipment.

In the building area, designers and builders tend to specialize in general construction (large commercial or institutional structures) or light construction (small residential or retail spaces) for the same reasons. Light construction is subject to a lesser degree of regulation and scrutiny and uses different or less complex construction techniques. Acroterion (talk) 13:04, 10 November 2017 (UTC)

→Civil engineering --Hans Haase (有问题吗) 19:52, 11 November 2017 (UTC)

"Beam me up, Scotty!"

OK, I think everyone knows what's the fundamental problem with teleportation -- because of the Heisenberg uncertainty principle, Scotty would come out the other side looking like scrambled eggs (in the literal sense!) However, suppose that instead of actually teleporting stuff, the machine worked by somehow temporarily creating a wormhole which could then be used as a shortcut through space-time -- would that work more-or-less like teleportation? 2601:646:8E01:7E0B:0:0:0:EA04 (talk) 03:04, 10 November 2017 (UTC)

Yep! And you could use it to build a galaxy-wide empire of fear and oppression until somebody else with starship technology destroyed your home-world from orbit. You'd be better off sticking with warp drives.

Nimur (talk) 05:24, 10 November 2017 (UTC)

Teleport has two common meanings, and probably a hell of a lot more uncommon ones. In one meaning, it means instant movement from one location to another. In most magical instances, that is how teleport works. Object instantly vanish and reappear elsewhere. In another, it means to move something from one place to another without physically moving it - the "instant" is lost. Star Trek uses the second meaning. An object is turned into energy, transmitted (at the speed of light, not instantly) to another location, and assembled again. Depending on the meaning of teleport you want to use, the answer to your question could be yes or no. You are physically moving an object one place to another - just taking a shortcut. That isn't teleporting in the Star Trek sense. But, you have the ability to instantly move an object from one place to another. That is teleporting in other (magical) popular sense. 209.149.113.5 (talk) 15:25, 10 November 2017 (UTC)

Wormhole models involving black holes have to deal with spaghettification and infinite values at singularities. The standard "follow the money" argument applies otherwise. If telepathy or precognition were possible, psychics would be rich, not hanging their shingles in the cheaper part of town. If teleportation works, "where are they?". μηδείς (talk) 22:49, 10 November 2017 (UTC)

That's not actually a disproof, because the preak would probably remember losing his last lone dollar in a last desperate trip to the casino, and then of course, that's what would happen. Wnt (talk) 15:08, 11 November 2017 (UTC)

I am quite aware that you can't prove a negative, but to quote Christopher Hitchens, "What can be asserted without evidence can be dismissed without evidence." μηδείς (talk) 21:38, 11 November 2017 (UTC)

We're nowhere near proving or disproving this kind of future tech, including direct teleportation without a wormhole. Look into some of W. G. Unruh's recent work - especially there are some papers by Qingdi Wang that seem absolutely mind-blowing if only I could understand a bit of them. [1] The nature of spacetime is much more ... fluid ... than we typically conceptualize, and right now it seems to take as much theorizing to explain why things don't teleport than why they could. Wnt (talk) 15:14, 11 November 2017 (UTC)

This is how they transfer people in The Culture series of novels. LongHairedFop (talk) 17:42, 11 November 2017 (UTC)

Septic shock

Are there any known cases of patients surviving and recovering from septic shock with no treatment? 193.240.153.130 (talk) 12:41, 10 November 2017 (UTC)

The article Septic shock reports that the mortality rate from septic shock is approximately 25–50%. It also states that sepsis has a worldwide incidence of more than 20 million cases a year, with mortality due to septic shock reaching up to 50 percent even in industrialized countries. There has been an increase in the rate of septic shock deaths in recent decades. Blooteuth (talk) 15:14, 10 November 2017 (UTC)

To be complete... There has been an increase in attributing deaths to septic shock in recent decades. Because septic shock often leads to stroke, heart failure, or respiratory failure, it is reasonable to attribute death to the result of septic shock rather than septic shock itself. 209.149.113.5 (talk) 15:28, 10 November 2017 (UTC)

Is the given mortality rate for treated or untreated septic shock, or is that difference not distinguished in those statistics? My understanding was that septic shock was pretty much always fatal if untreated, but that's a layman's vague memory. μηδείς (talk) 22:44, 10 November 2017 (UTC)

I cannot imagine a scenario where a case of septic shock would be known to medical authorities and not be treated. If a case where to occur and not be brought to medical attention I have difficulty imagining how that would be recorded as a survival. Richard Avery (talk) 10:56, 11 November 2017 (UTC)

@Richard Avery: Seek and ye shall find: [2][3][4] I'm not quite sure it's the accepted standard of care, but there are a lot more stories like this. Wnt (talk) 15:23, 11 November 2017 (UTC)

• I find no direct address to the issue of total non-treatment, but Septic_shock#Epidemiology says that survivability without treatment goes down 4% per hour, and that septic shock is otherwise normally fatal within seven days or less. I was looking for a source for treatment before the advent of antibiotics, but it seems the disease was poorly understood until the recent discovery that it is more a problem of immune response than bacterial toxins. μηδείς (talk) 21:33, 11 November 2017 (UTC)

Physical Photo Stenography

This animation shows how numbers are hidden in the Ishihara test of Color blindness within a circle of dots appearing randomized in color and size. Blooteuth (talk) 15:24, 10 November 2017 (UTC)

The same image viewed by white, blue, green and red lights reveals different hidden numbers. Blooteuth (talk) 13:04, 15 November 2017 (UTC)

I want to recreate something I've seen. I want to have three photos - actual photographs. I want to have a blue, green, and red plastic filter - just little squares of colored plastic. I want to see a number appear in the photo when I place the filter over the photo. I've seen this as a child. It is also used in color-blindness tests. However, when I change the color in any area of a photo, it is obvious with or without the filter. What is the trick to making it hard to see the number without the filter, but obvious with the filter? 209.149.113.5 (talk) 15:11, 10 November 2017 (UTC)

I see the animation. It doesn't appear to be helpful for what I want. I want a physical photo that I can sit in a frame on my desk. I don't see a number in it. Then, I place a blue sheet of plastic over it and I can clearly see a number. I'm looking into the red-blue 3-D effects to see if I can make it work. I've tried a lot of ideas, but either I can clearly see the number with or without the filter or I can't see the number at all either way. 209.149.113.5 (talk) 15:31, 10 November 2017 (UTC)

"The Trick"? There are lots of tricks involved in making a high-quality optical illusion like the one described here - but if I had to name the single most important trick, it would be understanding that the color filter selectively attenuates chroma noise while equally attenuating luma noise. For your illusion to work, you must hide a very low-power signal in the chroma-channel and bury it among powerful luminance noise. Pick a color-space that lets you do that - HSV or YUV - for your noise-generator; then, blend with your original image.

First, let's correct a typo: you probably mean to say steganography, rather than stenography.

There are two parts to your task: (1) to create a special image that appears noise-like, or invisible, when viewed in normal conditions, and appears to contain your numeral when viewed only in one color channel; and (2) to combine that image with your original photograph.

Task 1 is construction of a the "stenographic image." (sic). This is the image that contains the secret message you want to convey, but only when viewed through the color filter. You can rely on certain facts of visual perception, and try to make the data appear noise-like by capitalizing on facts like human visual perception biases pertaining to contrast, illumination, edges and contours, and so on: these can inform the choice of your noise-generator. There's a lot of art to this: it's actually much more subjective than any other part of your task. Bear in mind that you are creating a synthetic data set that is going to be combined with an image in later processing: knowing this, you have a lot of options to choose from when you represent your noise. For example, you can choose to make your noise zero-mean: that entails representing each pixel as a signed data type, which is not a common way to represent a finished image product. Our article on steganography tools list several existing pre-packaged software options - very few of them are any good. The sort of task you describe tends to be so customized that it would require a custom software process designed just for your needs.

Task 2 is digital compositing; it is the actual steganography, or the hiding of the previous "noise-like signal" inside another photograph. You can use a wide variety of methods to blend these images. You can also use the special knowledge about how your image will be composited to help you craft the noise-like data in the first task. Compositing is, in itself, an entire field of study: you can add the images; you can multiply them; you can mix them according to a nonlinear relationship. Once again, this is as much art as science. The classical paper is Porter/Duff (1984). It gives you a fantastic overview of what your options are - I dare say, it is "mathematically-complete" (you have no other options besides what they describe in that paper). In the last decades, academic research, commercial productization, and practical experience have developed compositing in to one of the most elaborate areas of image processing. Artists and software designers spend years studying how to do it so that it looks good. In your case - intentional injection of a noise-like signal - you have the extra difficult job of preserving a noise-like character without perceptually damaging the final image.

From a practical point of view, some of the tools you may wish to use include a layer-capable image editor, like GIMP or Adobe Photoshop; and a programmable mathematical tool like MATLAB or the python programming language to synthesize noise and arrange it in the form of a raster image. If you can afford the commercial tools like MATLAB and its image processing toolbox, you will have great advantages, especially in terms of the ability to rapidly iterate your efforts and get immediate visual feedback.

Your task is not trivial; there are no easy automatic ways to do it. You will also need great familiarity with your tools, and the ability to carefully control them (typically this means writing program-code). You must be aware of all the manual- and automatic- image processing steps that your software tools will perform on your intermediate and final products to ensure that your steganographic work is not lost, for example, by automatic post-processing, image compression, or exported image- or file-format changes at the last step.

Nimur (talk) 18:17, 10 November 2017 (UTC)

Follow up question on dark matter and stars

Perhaps most stars everywhere are sited at dense wisps of dark matter to give the contraction a head start? After all, there is considerably more dark matter than ordinary matter, so it could be the determining factor.144.35.114.188 (talk) 15:45, 10 November 2017 (UTC)

Perhaps; perhaps not. What possible further response is there, given that no one knows anything about dark matter other than that it appears from indirect evidence to exist? As star formation was thought to be reasonably explicable before dark matter was conceived of, there would seem to be no need of your hypothesis (with apologies to Laplace). {The poster formerly known as 87.81.230.195} 90.200.138.27 (talk) 21:29, 10 November 2017 (UTC)

Yes but calculations based on no dark matter wisps vs. many dark matter wisps could indicate respective rates of formation. It has always seemed to me the current theory of star formation has been thought of as necessary rather than satisfactory.144.35.45.72 (talk) 21:56, 10 November 2017 (UTC)

This is not a forum or the place for speculation. Questions about articles or sources are more relevant than anyone's personal pet theories. That being said, this article The Bullet Cluster Proves Dark Matter Exists, But Not For The Reason Most Physicists Think was an interesting read, and seems to imply that dark matter may only play a very indirect role in the collision of interstellar gas. μηδείς (talk) 22:41, 10 November 2017 (UTC)

medeis. Along with your good contributions, you have made several disgusting offensive irrelevant comments here, and made silly power trips like one you have just made over the years, and also made false accusations. You are scarcely one to advise on what is appropriate in this venue. Anyway, I had good reason in asking my question here: I asked my socalled "pet theory" question for valid reasons: there is a good chance that some astrophysicist who volunteers might know whether the suggestion is reasonable or know how to refute it, and because it has almost certainly has been considered, perhaps briefly, in the astrophysics literature, and a reference could be provided to me...It is a shame you that you are often so difficult.144.35.45.72 (talk) 00:31, 11 November 2017 (UTC)

This is not a forum or the place for speculation or personal attacks. Questions about articles or sources are more relevant than anyone's personal pet theories (pet not being a "disgusting" or "offensive" term). That being said, this article The Bullet Cluster Proves Dark Matter Exists, But Not For The Reason Most Physicists Think was an interesting read, and seems to imply that dark matter may only play a very indirect role in the collision of interstellar gas. Please read that article, as it is the only source anyone has given, and even in good faith, so far. μηδείς (talk) 03:44, 11 November 2017 (UTC)

okay, thanks I guess.144.35.114.29 (talk) 20:42, 17 November 2017 (UTC)

Maybe the stars make dark matter 71.181.116.118 (talk) 23:37, 10 November 2017 (UTC)

That's not a bad idea! If that were true, and if we also supposed stars were the SOLE way of making dark matter, then since currently there's more dark matter than ordinary matter like stars etc, it would imply that long ago there was much much more ordinary matter than there is now, and it has since mainly converted to dark matter. That might be testable(and already thought of and tested).144.35.114.29 (talk) 20:41, 17 November 2017 (UTC)

Please feel welcome to speculate when asking questions; that's what we (should be) here for. You are apparently not the first to wonder this about dark matter: Dark star (dark matter) is more than just a fun movie. See [5] which says that axions and maybe WIMPs are not suitable for this sort of thing, but neutralinos might be... anyway, that article is confusing, and is intentionally untrackable back to a real source, but if you look at something like [6] you can see references back to a bunch of Paolo Gondolo papers in JCAP and one of them from 2010 supposedly should be the one that touches on this. I don't know what he's toking but it's laced with a whole lot of heavy mathematics, so you're in for a ride... Wnt (talk) 01:01, 13 November 2017 (UTC)

Thank you.144.35.114.29 (talk) 20:41, 17 November 2017 (UTC)