Wikipedia:Reference desk/Archives/Miscellaneous/2010 November 25

Welcome to the Wikipedia Miscellaneous 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 25

Establishing the Date of a Photograph

Is there a relatively easy way to establish the date of a photograph solely from its content (assuming no digital manipulation)? Including a newspaper headline will establish that the photo cannot have been taken before that date, but of course it could have been taken afterwards. Including a unique event, such as three identifiable Olympic winners standing on the winner's platform, would work, but such events do not come along that often, and probably never at the time and place required. I am looking for something that is easy to implement at an arbitrary time and place, using readily available objects. —Preceding unsigned comment added by Callerman (talk • contribs) 00:37, 25 November 2010 (UTC)

I'm confused. The first half of your question seems to be asking how to find out the date of an existing photo, whilst the second half seems to be asking how to include some proof of date in a photograph you are about to take. If the first half is what you are asking, then I would say it would very much depend on the photo and it would be impossible to give an answer that fits all. If the second part is what you are asking, then I would suggest adding a calendar? Or, the TV in the background, with a certain news item on? A computer? A sign on the wall with the date? Include the sun in the photo, so the angles can be measured to find the date? --KägeTorä - (影虎) (TALK) 01:13, 25 November 2010 (UTC)

You pose an interesting challenge. The most readily available object in my mind would be a handheld GPS unit, but you'd have to make sure the readout was visible in the photograph. As far as I know, Time/Date data on an unmodified unit is always accurate. I don't know if that would hold up to cross-examination in court, though. The Masked Booby (talk) 01:18, 25 November 2010 (UTC)

Depends heavily on the subject of the photograph, I'd say. Photographs of the sky, a certain tree, a certain building, person or street or a particular crowd will all have to be judged differently. --Ouro (blah blah) 05:06, 25 November 2010 (UTC)

You can't date a photograph precisely from subject alone if you don't know anything about the subject - but often you can reasonably place a photo within a particular era based on the relative technical quality of the photo. The more you know about the subject, the better chance you have of narrowing the date. For example, a picture of random action at a major league ballpark can sometimes be narrowed down to a specific game if the ballpark and players are identifiable. ←Baseball Bugs ^{What's up, Doc?} carrots→ 07:05, 25 November 2010 (UTC)

I suspect that some of the preceding answers are interpreting Callerman's "establish" more post-eventually than intended, and that the question, as KageTora surmised and The Masked Booby understood, is merely how to include something in a photograph about to be taken that will prove its time and date. Computer displays and domestic/office calendars could have been manipulated for the occasion, and a TV programme could be a recording; the Sun's position would likely be too difficult to measure accurately enough for a precise date. A public calendar display, such as might be found on a railway station platform or in major public venues, or an electronic news billboard in the latter, might be thought sufficient but would greatly constrain the available locations, contrary to Callerman's "arbitrary time and place" stipulation. An interesting problem. 87.81.230.195 (talk) 09:17, 25 November 2010 (UTC)

(EC) It really depends for what purposes... You could simply write the date and time on paper and hold it up if all you want is a record. But since you mention newspapers, I guess this isn't sufficient. I don't really see much point for a calendar, it doesn't see any real better then a newspaper and in fact in some ways worse depending on the nature of the calendar (even if you have control over a newspaper, unless it's some sort of silly tabloid where the content has little to do with what's going on in the real world people would expect the front page to reflect what's going on in the world so if you print something you constructed several years before it's going to seem odd, this is usually doable with a calendar within reason). Similarly for the GPS thing, it's surely relatively trivial to hack the GPS receiver display to show whatever you want and I can't see why there would anything on the display you can't predict beforehand. Of course in the modern era of easy digital manipulations you probably want several high quality photographs, whatever you do, so that they can be analysed for any signs of manipulation and even then I wonder if it will always be enough.

Proving something wasn't taken in the future is obviously the more difficult thing (since there are things about the future but we can know what happened in the past) a better bet may be not to prove it in the photo itself but someway demonstrate the photo existed when you say it did (you can combine this with newspapers or whatever to ensure it wasn't taken in the more distant past). Emailing a copy to 5 different email service providers just after you take it may be enough for some although I'm not sure whether it will be enough for a court. Perhaps also post various hashes of the file to usenet and other forums. Getting a few lawyers to hold on to copies (and who are willing to attest they've held them since date X) may also help (I don't know how easy this would be but I would guess if you have the cash it can't be that hard).

Nil Einne (talk) 09:34, 25 November 2010 (UTC)

(ec)Don't most digital cameras come with a feature to impose the current date and time in a corner of the picture? And maybe modern film cameras as well. I wonder if the OP is looking for definitive proof for legalistic reasons, or merely to remind himself of when the picture was taken? ←Baseball Bugs ^{What's up, Doc?} carrots→ 09:38, 25 November 2010 (UTC)

I suspect the motivations are as simple as I and 87. have stated - validating a photograph forwards AND backwards in time, "an arbitrary time and place, using readily available objects" is a simple premise which gives way to fiendish difficulty upon close consideration. None have yet proposed a reasonable solution. The Masked Booby (talk) 10:42, 25 November 2010 (UTC)

Using the timestamp feature would be sufficient in most situations. It would be nice if the OP would add to his grand total of 2 edits since July by explaining a bit more clearly what he's got in mind. ←Baseball Bugs ^{What's up, Doc?} carrots→ 10:47, 25 November 2010 (UTC)

I think that's the answer, now, after having given it a bit of thought. The OP said 'assuming no digital manipulation', and modifying the date on the time stamp could be considered to be digital manipulation. Then, there is also the added convenience of all the meta data that digital cameras tend to include with an image file - the date and time taken is usually part of this, whether the time stamp is included on the picture or not. --KägeTorä - (影虎) (TALK) 12:27, 25 November 2010 (UTC)

That doesn't help if it was a Brownie snapshot taken around 1920. Which is why the OP's question is so confusing. If he hasn't taken the picture yet, why is he worried about it? He can do anything he wants. He could have someone stand there with the current date on a piece of paper. He could position the camera so that something dated is in the background. And I would argue that posting the timestamp doesn't qualify as "manipulation" unless you add it later. And as you say, the metadata would cover it, IF the timestamp in the camera were accurate. The OP needs to be more specific about what he's got in mind. But at his present editing rate, it will be March before he gets back to us. ←Baseball Bugs ^{What's up, Doc?} carrots→ 12:33, 25 November 2010 (UTC)

I think both of these points have been covered in the discussion at some point, but my idea would be that including the front page of a well known newspaper dates it to no earlier than the date of the newspaper. Then developing it and getting it notarised, or certified by whatever is the local certifying authority dates it to no later than the time of certification. This combination should be as effective as you can get.

Electronic innovations on top of that might be to use a computer screen displaying an electronically published news story from a major news website instead of a newspaper, though I don't think most countries provide some kind of electronic notarisation/certification of documents just yet. --PalaceGuard008 (Talk) 16:37, 25 November 2010 (UTC)

If he's asking about this for a future photo, what's the issue? He can use whatever method he chooses to. Unless he has absolutely no imagination, which is the only explanation I can think of. Well, not quite. He might really be asking the opposite, namely can a picture be taken which is impossible to date via content-only. I look forward to his followup clarifications, around March of 2011. ←Baseball Bugs ^{What's up, Doc?} carrots→ 04:09, 26 November 2010 (UTC)

I'm not sure I understand you. On the first point, clearly he's not asking about "whatever method he chooses to". He's asking for a conclusive method, although he has not specified whether the criterion is scientifically conclusive or legally conclusive (or some other stsandard).

On the second point, of course a picture can be taken that is impossible to date via content-only. A picture of a blank piece of paper with no other content would be one. I very much doubt anyone would ask for that. --PalaceGuard008 (Talk) 10:05, 26 November 2010 (UTC)

I think the OP wants to take a photo in the future, which even a month later he took it, that photo will contain a proof beyond reasonnable doubt that it was really taken on the day he took it, so no one will argue that it wasn't. --Lgriot (talk) 10:09, 26 November 2010 (UTC)

The basic problem is that he has not given us enough information, and since he's made only 2 edits since being created on July 27th, I wouldn't hold my breath waiting for it. ←Baseball Bugs ^{What's up, Doc?} carrots→ 10:52, 26 November 2010 (UTC)

Thanks for all the comments so far - this has generated far more interest than I thought! I'm sorry my original question was ambiguous. Lgriot has hit the nail on the head. I'm looking for a method of including something (an object, an event) in a still photograph yet to be taken which will prove beyond reasonable doubt that the photo was taken on a particular day, or at least within a particular week. Metdata can be modified, so I need something in the image itself. Of course, digital images can be manipulated, but manipulation of image content can be detected by sophisticated software. It seems to me that it has to be something of a public nature which the photographer could not reasonably influence, such as a newspaper heading (although that won't work, as already discussed). Perhaps an airport arrival/departure board would do it, but I was hoping for something a bit more accessible. —Preceding unsigned comment added by 82.44.14.76 (talk) 11:09, 26 November 2010 (UTC)

It's best not to jump around between registered user ID's and IP addresses, for a number of reasons. I recommend that you use a film camera rather than a digital camera, and take some establishing shots on other frames in the role, both before and after your photo, which will pinpoint the date and time. That's in addition to having someone in the photo holding up a sign with the date and time. Then when you get it developed, tell them not to slice the negatives but to leave them as a single roll. That would likely stand up to close scrutiny much better than would a digital image. Although to solidify the situation, it might be best to bring both types of cameras, and shoot the same pictures with both. ←Baseball Bugs ^{What's up, Doc?} carrots→ 11:56, 26 November 2010 (UTC)

Include in the photograph an identifiable female in an advanced (visible) state of pregnancy. Cuddlyable3 (talk) 12:38, 26 November 2010 (UTC)

There are any number of possibilities, none of them foolproof. But it might help if the OP would be a little more specific about what he's got in mind. I don't think this is a "what if" kind of question, I think he's got a very specific scenario that he's planning. Knowing more about that plan might yield more specific and better answers. ←Baseball Bugs ^{What's up, Doc?} carrots→ 12:49, 26 November 2010 (UTC)

Bugs, I think the OP has been extremely clear in his post dated 11:09, 26 November 2010. Here is another idea, not very practical, but it may trigger other ideas, so I'll try. If you need to prove the date within 6 months only and if you are lucky enough to do this on a sunny day outside, you could use the sun shade of an unmovable object like the tip of the roof of a house. The position of the shade of that tip will not be exactly at the same place (at least not within 1 millimeter) any other day of the last 6 months. But the picture could have been taken 6 months later unfortunately.

Another idea is to split the problem in 2:

1. Use a newspaper in the picture, to prove that it wasn't taken before that date.

2. Post it on some publicly available storage like flicker, but make sure the server provide the public with an upload timestamp. This will prove that the picture was taken on or before the timestamp.

If the 2 dates coincide, then you have proven that the picture was taken on that date, since it cannot have been taken before (the news paper proves it) and it cannot have been taken after the flicker timestamp (unless you have access to the servers of Flicker, which I beleive beyond reasonnable doubt that you don't).

How's that, my dear OP? --Lgriot (talk) 17:24, 26 November 2010 (UTC)

Me and Palaceguard already suggested to split the problem in to 2 (albeit with slightly different ways). This was before the OP responded again so I'm not sure if this means it doesn't meet their purposes for whatever reason. Nil Einne (talk) 17:33, 26 November 2010 (UTC)

No, the OP is NOT being clear. He obviously has something specific in mind but won't tell us what it is. It's got to do with a picture he hasn't taken yet. He can use any method he wants to. If he literally cannot think of a way to do this for a picture he hasn't taken yet, he should file a complaint with God for failing to issue him a right-brain. ←Baseball Bugs ^{What's up, Doc?} carrots→ 18:38, 26 November 2010 (UTC)

I think the OP has been very clear; they want to prove that a picture was taken on a certain date, not before and not after, based "solely from its content" (meaning no digital timestamps or meta-data). What is confusing about that? Insulting the OPs intelligence is not helpful. 82.44.55.25 (talk) 18:55, 26 November 2010 (UTC)

What's confusing is why the OP seems to either have no imagination whatsoever, or is hiding some important fact that, if known, could more easily lead to what he thinks the right answer might be. ←Baseball Bugs ^{What's up, Doc?} carrots→ 20:00, 26 November 2010 (UTC)

The reference desk is here to help people, not criticize them for having "no imagination" and not being able to solve their own questions. 82.44.55.25 (talk) 20:16, 26 November 2010 (UTC)

Well, smart guy, maybe you could find an actual link to an actual article that would explain how to do what the OP wants, and thus supercede the speculation and guesswork here which is totally unsourced and about which the OP is being no help whatsoever. ←Baseball Bugs ^{What's up, Doc?} carrots→ 20:46, 26 November 2010 (UTC)

I don't know how to do what the OP wants, which is why I feel it's a very valid question and not deserving of your "no imagination" and "file a complaint with God for failing to issue him a right-brain" comments. 82.44.55.25 (talk) 21:11, 26 November 2010 (UTC)

Of course you don't know, because the OP won't communicate with those who are trying to invent answers to his question. I've offered some ideas. I don't see you offering ideas, just criticisms. Either offer some ideas of your own to try to answer the question, or shut up about it. ←Baseball Bugs ^{What's up, Doc?} carrots→ 21:19, 26 November 2010 (UTC)

My inability to give an answer does not mean the OP has written a confusing question. The OPs question here is very clear imo. If you are having trouble understanding a specific part of the question I can try and explain it to you. What part of it is confusing you? 82.44.55.25 (talk) 21:30, 26 November 2010 (UTC)

If you can't provide anything except wise-guy comments, then go edit somewhere else. ←Baseball Bugs ^{What's up, Doc?} carrots→ 21:32, 26 November 2010 (UTC)

You said the OPs question is confusing and you think for some reason they're deliberately hiding vital information crucial to understanding it. I disagree, and offered to help you understand the question if you explain what about it is confusing to you. How is that a "wise-guy comment"? 82.44.55.25 (talk) 21:38, 26 November 2010 (UTC)

See my comment farther below. ←Baseball Bugs ^{What's up, Doc?} carrots→ 21:50, 26 November 2010 (UTC)

There is the "Authentication by newspaper" trope, which gives some variations on "unique event" as mentioned in the original question. How much flexibility do you have, OP? If any time/place would do, maybe including some astronomical event or alignment in the picture would work (eg. you could take it at Stonehenge at daybreak on a certain day of the year, something like that). WikiDao ☯ (talk) 21:18, 26 November 2010 (UTC)

It's pretty obvious by the responses that there is no airtight, foolproof way to do what the OP wants. Which is why it would help if he would explain the specific scenario he has in mind. What level of proof is he looking for? Good enough to convince his friends? Or good enough to stand up in a court of law? ←Baseball Bugs ^{What's up, Doc?} carrots→ 21:36, 26 November 2010 (UTC)

Assume court of law, because that will include friends too. --Lgriot (talk) 21:59, 26 November 2010 (UTC)

There have been plenty of assumptions in this section, with very little useful input from the OP, hence all the guesswork. It would take some research to find what level of proof is necessary in a court of law in order to properly date a photo. And those rules might vary from state to state. We are getting close to being asked to give legal advice, which we don't do here. ←Baseball Bugs ^{What's up, Doc?} carrots→ 22:04, 26 November 2010 (UTC)

Take the photograph with a newspaper headline in it, and then go to a lawyer and get them to certify that they recieved the photo on a particular date. In Britain you could send the photo to yourself as a postcard by registered delivery and get the date certified that way. 92.15.11.45 (talk) 12:59, 27 November 2010 (UTC)

That or variations thereof has been suggested by 4 different people now (including you). It's not clear if it meets the OPs requirements. Nil Einne (talk) 18:11, 27 November 2010 (UTC)

OK, then get X witnesses to all swear that the photo was taken on a particular day. 92.15.14.132 (talk) 22:54, 28 November 2010 (UTC)

I'm a little late to this but how about: Indelibly print the headline (or image of front page) of the day in question's newspaper onto a material with a recordable rate of decay. Then, you will have the newspaper proving the photo wasn't taken prior to this date, and a material which can be proved by analysis, that proves the photo couldn't be created after this date, due to its current state of decay. Now we just need to find a material with these properties... Muzzam (talk) 06:33, 30 November 2010 (UTC)

Thanks for all responses som far. However, responses suggesting supporting evidence such as witnesses are missing the point. To repeat, I'm looking for a method of including something (an object, an event) in a still photograph, yet to be taken, which will prove beyond reasonable doubt that the photo was taken on a particular day, or at least within a particular week, solely from the content of the image itself, assuming no digital manipulation. Some sort of public calendar would presumably do it, or an airline arrivals/departure board, but these are not readily available at all times and places and I was hoping that someone would be able to think laterally and suggest some more available object/event. It is not necessary that the photo directly show the date, but the date must be derivable from the image. For example, a photo of a royal wedding or a political assassination would clearly be datable but they are not events under the users control. Perhaps there is no straightforward way of doing it.

Osama bin Laden as a name

Could it be that many Arabs are called Osama bin Laden and do not have any relationship with the real one, nor with his family? I know that Osama is a pretty common name and bin Laden seems at the first glance to be relatively common (but I am not sure). --Quest09 (talk) 12:44, 25 November 2010 (UTC)

If there are (and as you say, it seems likely), then they are as 'real' as the notorious one. I'd imagine this could make for problems in some circumstances. AndyTheGrump (talk) 12:54, 25 November 2010 (UTC)

Yes, like taking a plane? Actually, Al-Qaida seems to be composed by low-rank people (who could be suicide bombers and the like) and higher rank people (who would never do it).--Quest09 (talk) 12:59, 25 November 2010 (UTC)

Folks who work at airlines are aware that there are many people who have famous names, and are not easily spooked by namesakes. If you worked at a ticket counter and an Osama bin Laden came by, would you immediately assume he is the internationally wanted criminal? Of course not, and people who work at ticket counters have real experience of this phenomenon, and are even less likely to freak out than you are. Yes, there are occasional reports of people who get detained for having famous names. This is so rare that when that happens, it can make world wide news.Captain Hindsight (talk) 00:24, 26 November 2010 (UTC)

This is true. The real problem is if you share a name with a criminal/suspect/terrorist/watchlisted individual who isn't particularly famous or well-known. At a quick glance, they aren't going to recognize that "Oh, he's obviously not him, so no worries". See also No Fly List#False positives. TenOfAllTrades(talk) 16:56, 26 November 2010 (UTC)

Yes. In 2005, the Canadian documentary series The Passionate Eye aired a show titled Being Osama. It interviewed six Arabs who shared the name Osama bin Laden; the director apparently identified at least seventeen such individuals in his research: [1]. TenOfAllTrades(talk) 15:04, 25 November 2010 (UTC)

Actually, the six persons in the documentary only share the first name "Osama", but even that was problematic enough for them in the wake of 9/11. The surname "Bin Laden" is a lot less common than the first name "Osama", and seems to be geographically restricted to Saudi Arabia and nearby countries, whereas "Osama" is given all over the Arab and Muslim world. --Xuxl (talk) 16:28, 25 November 2010 (UTC)

Arabic name may be relevant here. Nil Einne (talk) 19:14, 25 November 2010 (UTC)

geography India

when looking for towns, city, district and state in india, the information is segregated

can the below information made available as a summary on click of button

state - district - city —Preceding unsigned comment added by Anky2211 (talk • contribs) 13:52, 25 November 2010 (UTC)

What do you mean? Can you give us an example? ←Baseball Bugs ^{What's up, Doc?} carrots→ 13:55, 25 November 2010 (UTC)

The categories Category:Geography of India and its subcategory Category:Subdivisions of India would give you some places to start researching. --Jayron32 15:07, 25 November 2010 (UTC)

sterling silver

i have an old silver metal tray it has sterling stamped on the outer edge how do i tell if this is real silver —Preceding unsigned comment added by 85.211.239.101 (talk) 20:15, 25 November 2010 (UTC)

It's most likely to be sterling silver, an alloy containing 92.5% by weight of silver and 7.5% by weight of other metals, usually copper. Here are a couple of tests for silver you might be able to perform at home. --Tagishsimon (talk) 20:18, 25 November 2010 (UTC)

If it was made in Britain then it should have hallmarks on it - see Silver hallmarks. 92.28.251.194 (talk) 20:54, 25 November 2010 (UTC)

Maximizing expected utility

What is the process one should use to maximize one's expected utility ? 220.253.217.130 (talk) 22:15, 25 November 2010 (UTC)

Go to school and university. The Masked Booby (talk) 22:27, 25 November 2010 (UTC)

Why would anyone want to? AndyTheGrump (talk) 22:33, 25 November 2010 (UTC)

To learn to know what is the process one should use to maximize one's expected utility. Captain Hindsight (talk) 00:14, 26 November 2010 (UTC)

Er, that wasn't what I meant. I'm all in favour of education (for those that want it), but I can see little point in 'maximizing one's expected utility', whatever that means. This might possibly be something one should do with light bulbs, but not with people. AndyTheGrump (talk) 00:21, 26 November 2010 (UTC)

Keeps your lights on when you leave the house. schyler (talk) 00:49, 26 November 2010 (UTC)

I suspect he's trying to say "make the best use of one's potential". Given that, I'd recommend he start with a journalism class, so he can learn to use less obscure metaphors. ←Baseball Bugs ^{What's up, Doc?} carrots→ 01:49, 26 November 2010 (UTC)

Marginal analysis, specifically looking at Marginal utility. Total utility is maximized when marginal utility is zero (assuming the law of Diminishing returns holds true). Buddy431 (talk) 03:11, 26 November 2010 (UTC)

...And assuming that economic hocus-pocus applies to people (it doesn't, as is self-evident). AndyTheGrump (talk) 03:20, 26 November 2010 (UTC)

The phrasing of the question suggests that Mr. 220.253... is in a microeconomics class, in which case my answer is probably the one that the teacher is looking for. "Utility" is a term rarely encountered outside of economics. If the OP is in a more advanced course, they could be looking for something along the lines of the Expected utility hypothesis, which deals with maximizing "expected utility" given unsure payouts (i.e. bets) dependent on things like probabilities of success, potential payouts, and risk aversion. Buddy431 (talk) 05:16, 26 November 2010 (UTC)

And our standard response to requests for help with homework is...? ←Baseball Bugs ^{What's up, Doc?} carrots→ 10:55, 26 November 2010 (UTC)

...to not berate people who answer homework questions in good faith, even though we prefer not to answer them at all. --Jayron32 15:28, 26 November 2010 (UTC)

On the other hand, should we respond with what we think is the answer the teacher is looking for, or with what we consider to be the correct answer, which is that 'maximizing one's expected utility' is not something that people actually engage in? AndyTheGrump (talk) 15:47, 26 November 2010 (UTC)

You might not try to maximize your expected utility, but many people certainly do. We may not always be rational actors, but we do a good enough approximation of it enough of the time that microeconomic theory does a lot to describe personal decisions. If you have not read it, Freakonomics has some entertaining (and interesting) applications of economics, including microeconomics, to situations that we might not ordinarily associate with the field. Buddy431 (talk) 00:17, 27 November 2010 (UTC)

No, the point is, as general practice, we don't answer homework questions. I never do, I always tell the person to look in their textbook or look through their lecture notes for the answer. Period. However, I hold no prejudice if someone comes along after me, and in good faith, answers their question. That's fine too. There's no need to make snide comments against people who DO answer them, even if I don't. --Jayron32 15:57, 26 November 2010 (UTC)

(e/c) We should respond by providing helpful and informative resources (preferably WP articles) related in some way to the question. No homework was done for anyone in response to this question so far, so that's okay then. WikiDao ☯ (talk) 16:01, 26 November 2010 (UTC)

Oh, just read closer that "Total utility is maximized when marginal utility is zero (assuming the law of Diminishing returns holds true)" which I guess would be the answer here... :| WikiDao ☯ (talk) 16:05, 26 November 2010 (UTC)

If you are evaluating alternative investment projects, then you would calculate the expected-NPV (ENPV) by using NPV with decision trees. However this does not take account of utility theory, where the value of an extra £1 is a lot less if you already have £1M.

If the question is about maximizing one's utility as a person, then is this utility for oneself or for others? 92.29.115.8 (talk) 00:40, 27 November 2010 (UTC)

Try expected value (especially the example about the coin toss) and utility then expected utility hypothesis202.173.162.101 (talk) 22:26, 28 November 2010 (UTC)

If the OP means the question in a personal sense, then self-efficacy and self help may be relevant. 92.15.14.132 (talk) 23:20, 28 November 2010 (UTC)