Wikipedia:Reference desk/Archives/Computing/2006 October 14

Humanities Science Mathematics Computing/IT Language Miscellaneous 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 at one of the pages linked to above.

Frames Per Second - limits of conscious visual recognition

Question - what is the typical limit in Frames Per Second (or fractions of a second) of human visual recognition on a conscious level?

Background - trying to do an experiment in subliminal messages using home edited videos. Microsoft Movie Maker 2 appears to edit only at 7 FPS. I have found that when a frame is 1/7th of second, it is clearly visible to the conscious mind. I have scoured articles on how fast the image needs to be (including Wikipedia article on subliminal messages), but cannot find answer to above question. When shopping for a replacement consumer software package that allows editing in more detail, how many frames per second do I need to edit so that, when I insert the subliminal message frame, most of the viewers will not consciously perceive it?

For extra credit, any ideas on consumer software packages that would allow such editing?

Thank you in advance.

I'm in no way an expert on this, but from what I've read, when the numbers of frames per second are 24 or more per second, the human brain can no longer recognize them as individual images, only as a moving picture. That is why movies in movie theatres are at 24 fps (which curiously makes them a few mintues shorter when airing on PAL-television, which is 25 fps). However, I'm certain that the human brain can comprehend images at even higher fpses, whether conciously or uncounciously. That is, if a frame that is radically different flashes by in a 30 fps clip, you probably would notice it (even though you obviously couldn't recall what was actually in the frame). That is my guess anyway. I'd say try it with 24 fps and a few higher numbers and see the results you get. Oskar 01:42, 14 October 2006 (UTC)

This limit is known as the flicker fusion threshold, so see there. A similar question was recently asked on the Science reference desk: WP:RD/S#How many "frames" per second can our eye process?, so see there, too. We also have an article on subliminal message. Simon A. 08:35, 14 October 2006 (UTC)

I too am no expert, but I am certain that I saw a documentary many years ago that stated that the human brain precives "animation" at 13FPS.

That's probably referring to the minimum framerate at which we can detect motion, not the maximum limit. 13 FPS is defianately visibly different from 60 FPS; you'd just have to try playing Battlefield 2 on my computer to figure that out :P. -- Consumed Crustacean (talk) 23:20, 17 October 2006 (UTC)

Page views of websites

I want a few links in which I can see the total page views of top 10 English websites. If you dont know any links, I just want to know page views of three sites Yahoo, MSN and google ( all figures must be for global including USA)

The closest you're going to get are the alexa rankings. They are not very reliable however, but they will give you a ballpark number. See [1] [2] [3] Oskar 03:30, 14 October 2006 (UTC)

If, then, how, why

If integer(FOO/2) = FOO/2 then Print "EVEN" ElseIf print "ODD" EndIF Is FOO even or odd?205.188.117.12 14:38, 14 October 2006 (UTC)

I'm confused. How is "integer FOO/2 = FOO/2" a meaningful if statement? There's no way to know whether FOO is even or odd in this case, because there's no meaningful initialization for FOO. Even then, I'm not that fluent in VB, but I don't understand why someone would want to compare FOO/2 to FOO/2, because it should always print EVEN, regardless of the initialization. Something like "if (FOO % 2 == 0)" would be a better choice (not sure how to do that out of C). Wooty 20:31, 14 October 2006 (UTC)

This is an extremely bad way to do it, but in some languages (depending on how it uses types), it might work. See he's using the function "integer()" which extracts the integer value from a float (ie, it floors it). So integer(3.5) = 3. Then for instance if you have FOO = 15, you get integer(15/2) = 15/2, or 7.5 = 7 which is false, therefore the number is odd. I used a similar trick on my old TI-83+ because it didn't have a mod function and it had only one type (a float). Still, in more advanced languages, this is a mindnumbingly stupid way to do it. There are two major pitfalls: if FOO is of type integer, FOO/2 might return an integer answer, thus the statement is always true, also if the integer function returns an int, there might be type error. And it's hard to read.

Still I don't know how VB handles types, but if I remember my BASIC correctly it's pretty liberal. So it might work. That is, however, not an excuse for bad design. As you would say in a c-derivative language, (FOO % 2 == 0) ? "EVEN" : "ODD" is a much better way to do it. Oskar 21:14, 14 October 2006 (UTC)

The even/odd distinction depends on FOO being an integer. (Is 3.14 even or odd? No.) Thus we use a bitwise operator; test if FOO And 1 equals 0. It's cheap and readable.

Here's an old trick along the same lines. To decide if an integer n is a power of 2, test if n equals n And -n. (The latter expression selects the low-order 1-bit in n.) --KSmrq^T 23:32, 14 October 2006 (UTC)

This is a clever hack if you are sure that the variable is an int. This might not be the case; even though the value obviously is an integer, the type might very well be a float. You might, for instance, want to know if 14.00 is even or odd. The way to reliably test that number is to do the mod thing. You'll get part of the manitissa if you do the binary and thing. Oskar 01:42, 15 October 2006 (UTC)

If the number is stored in a float, how can we know it's an integer? First, store it in an integer, then test. In fact, VB might even do the desired conversion automatically, since the And operator is not defined for "fractional" data types. The power-of-two test is clever; the even-odd test is neither clever nor a hack, but only an obvious use of radix-2 notation. Is it a "clever hack" to observe that a number written in decimal is divisible by 10 if the last digit is zero? I don't think so. --KSmrq^T 08:12, 15 October 2006 (UTC)

There might be many reasons why you would want to know if an integer float is even or odd, and converting it to an integer first is just a nuisance. You might have an application with a lot of mathematical operations, and you wish to keep variables consitently floats. You might get the number as a result of series of operations on floats, and you knew that the result would be an integer. I called your way a hack because it utilized a special property in how the number was stored, not a special property in the number itself. If a solution is equally good (infact, I believe that if you do the foo % 2 thing on an int, clever compilers will do the binary and thing), there is no reason to go with the more obscure and less clear way. Using "mod 2" makes the code more readable, it makes the code more understandable, it makes the code more portable and it is a more stable solution. Oskar 12:29, 15 October 2006 (UTC)

Blocking ads in websites

Is it possible for websites to detect whether the browser is using ad blockers? And if the browser is using ad blockers, is it possible for that website to say a message saying ' you are using ad blockers; you cannot see this page' or something like that. Is this possible or is it ot possible for the website to do this? I would also like to know how many percentage of world Internet users block graphical ads in percent

Now why do I get the impression you want to spam the world with an infinite number of ads and keep them from blocking you ? StuRat 22:01, 14 October 2006 (UTC)

It is possible: use JavaScript to determine whether a popup window exists or not after requesting it to pop up. x42bn6 Talk 22:33, 14 October 2006 (UTC)

That would require the use of a popup window. And it what to do when the user removes it? And how would you distinguish between the user and the computer doing it? In either case, would you wish to send them such a message, basically telling them to fuck off? Not likely. You're talking about a potential customer (if StuRat's assumption is correct). Other than that, a user downloads a page and a browser then renders it. How it does that is something you ultimaltely cannot affect, once you've sent it. So you'd first have to get the viewer's cooperation and only then send the info. DirkvdM 07:05, 15 October 2006 (UTC)

I think most scripts do this action immediately after a window is loaded, so technically a user can't, er, click it that quickly. Basically, JavaScript initiates a window opening call and detects if it exists or not, then closes the window in the next statement. A quick Google search gives you a script, by the way. x42bn6 Talk 11:53, 15 October 2006 (UTC)

That is true if one clicks. However, I use the keyboard and often remove the popup before it is fully loaded, by hitting the easy to reach 'esc' key the moment I see the faintest flicker that looks like a popup. Of course, I don't presume to be faster than my computer. :) However, it would be easy to make a popup blocker that adds a little delay. DirkvdM 05:20, 16 October 2006 (UTC)

But you were not asking about popups, but about ad blockers. As long as there is no feedback by a script, you can't tell how a page is rendered by the user. And if a script asks for a reply, the computer (at some level or other) could of course lie. Browsers do so all the time, pretending to be IE. DirkvdM 05:23, 16 October 2006 (UTC)

Users of Mozilla Firefox get the luxury of Adblocker + NoScript extensions. Therefore, javascript detection won't work. And users of IE can disable Javascript as well, although they can't block ads directly (but via software). So... go away and try another idea, like threaten to crash web browsers if the users don't disable ad blockers ;) good luck...--inky 07:04, 20 October 2006 (UTC)

Printing PDFs in a Different Color

I'm taking a college course right now that doesn't have a textbook; instead, everyone has to print out a bunch of readings for every class. It's taking a toll on my black print cartridge. For Word documents I don't need to turn in, I change the font to a different color (say, blue) because I rarely if ever print anything that needs color. Is there a way to print black and white PDFs, which most of the readings I have to print are, in another color like I do for my Word documents? Thanks. --Maxamegalon2000 17:17, 14 October 2006 (UTC)

Unless the PDF is protected, a lot of the time you can select all, copy and paste into a word document, then change the color your self. Even if "select all" is not available from the edit menu, try Ctrl-a, Ctrl-c then Ctrl-v into a blank word document. Vespine 05:55, 19 October 2006 (UTC)

I wan't as much Information as possible on:

• World of Internet & unique websites,IT Buzzwords,Acronyms.
• Personalities-International,National,Local.
• Advertisements of IT and Communication companies.
• Software products,companies & brands,History of IT,humorous side of IT.
• Areas where computers have made an impact such as education, entertainment, books, multimedia, internet, banking, advertisements, sports, etc.—Preceding unsigned comment added by 59.95.5.64 (talk • contribs)

For all that info I suggest you read Wikipedia, one article at a time. If you have a SPECIFIC question, then come back here. StuRat 20:42, 14 October 2006 (UTC)

Best way to send large files over internet

What's the best way to send a large file (~100-500MB) over the internet if you don't have access to a reliable server with that much space? The tech-saavy could set up an ftp/http server on their home computer, configure the hardware firewall, and send a link. Most people aren't going to be able to do this, however. What are the other options? Sending the file through AIM? Is there some good free third-party server for this sort of thing-- an Imageshack for half-gig files, for example? --Alecmconroy 19:34, 14 October 2006 (UTC)

Best way? Learn to configure your router (http://portforward.com) and install a server (Apache web server). If there's anyone out there too stupid to look that up, they don't need to be sending their 500 megs of crap.--Frenchman113 on wheels! 19:59, 14 October 2006 (UTC)

www.megaupload.com goes up to 250mb, there are ones out there that go up to a gig. Wooty 20:26, 14 October 2006 (UTC)

You could always use bittorrent, even if you just want to send a single file to one person. Create a torrent from the file and register it to a free tracker and start your torrent-program and act as a seeder on that file. Then you can send the torrent (which is fine, since they're tiny) to every computer you want to download the file to and they can fire up their torrent software and download from the seeder. This method has a number of advantages: easy pause/resume functionality, it's free, it's very simple, as many people as you want can download it, don't have to worry about IPs and such (the protocol takes care of that), and the files can be as big as you want. If you want to send a really large file to someone, this is most likely the best way. You could use something like AIM, but the torrent protocol is much more sophisticated in transferring large files. For instance, what if you need to reboot while it's downloading? With bittorrent that isn't a problem, but it certainly might be with aim. Oskar 21:24, 14 October 2006 (UTC)

Easiest and fastest for files over 100MB would probably be these temporary hosting sites. This list is from the good folks at 4chan/b/; I can't vouch for any of them being non-evil. Here they are, grouped by maximum upload size. grendel|khan 05:06, 15 October 2006 (UTC)

2 GB:
http://www.sendover.com/
http://www.snaggys.com/

1.5 GB:
http://www.megashares.com/

1 GB:
http://www.transferbigfiles.com/

700 MB:
http://www.depositfiles.com/

500 MB:
http://www.filefactory.com/
http://www.zupload.com/
http://www.spread-it.com/
http://www.mooload.com/

300 MB:
http://www.uploading.com/
http://www.sendspace.com/
http://www.bigupload.com/
http://www.rapidupload.com/
http://www.sharebigfile.com/

250 MB:
http://www.megaupload.com/
http://www.updownloadserver.de
http://www.xtrafile.com/
http://www.bonpoo.com/
http://www.filecache.de/upload

200 MB:
http://supload.com/sendfile

I would like to point out, in defence of my bittorrent solution: it will take the same amount of time to upload to one of these sites as it will to transfer the file directly (it's capped by your upload speed), and it is much more "stable". if you will. What if it takes 3-4 hours to upload and you accidentally close your browser? What if a cable is accidentaly yanked out, or your internet connection suddenly goes out? The bittorrent protocol was invented to be able to easily transfer large files over the internet. That was it's purpose. Oskar 12:32, 15 October 2006 (UTC)

Don't know if there is a attachment size cap but gmail has a 2 GB size limit. Vespine 05:52, 19 October 2006 (UTC)

Yes, there's a 10 meg cap on individual emails.  freshofftheufoΓΛĿЌ  06:30, 19 October 2006 (UTC)