Wikipedia:Reference desk/Archives/Computing/2012 May 12

= May 12 =

Lines appearing on zoomed-in photos from my camera
Whenever I take a photo with the optical zoom set at maximum, my Canon PowerShot A540 camera creates faint lines on the resulting images; see the full-resolution version of File:Glendale Ridge Archaeological Site.jpg for an example of this phenomenon. Why is this the case, and what (if anything) can I do about it, aside from taking photos at reduced zoom? It's definitely only happening at that zoom and not at others; you don't see it at File:Epsilon II Archaeological Site.jpg or File:Ice on washed-out old road in the Hoosier National Forest.jpg, taken a few days later, but I uploaded the Epsilon II panorama that I did instead of a more zoomed one specifically because the zoomed one displayed the same lined effect as the Glendale Ridge image does. Conversely, the phenomenon hasn't been going on forever; I've had the camera for four years (and bought it used from a friend), and I don't remember this "effect" showing up in pictures from when it was substantially newer. I took File:Kintner-Withers House from the road.jpg about a year before Glendale Ridge, and due to the extreme distance of the subject from the public right-of-way, I zoomed in all the way, but the lines don't appear. Nyttend (talk) 00:36, 12 May 2012 (UTC)


 * Are you sure you don't have some type of special effect turned on ? It rather has the look of being printed on rough paper, where you can see the grain of the paper, which might be what this special effect is supposed to do. StuRat (talk) 01:07, 12 May 2012 (UTC)


 * Truly amazing! If I had to posit a guess, I'd say you have a "defect."  What in the world might cause that?  Without tearing apart the electronics or connecting up a software debugger, or grabbing the raw data (which would invariably require proprietary tools)... it's nearly impossible to know... but if I had to hazard a guess, I'd gravitate towards these few possibilities:
 * The power-supply might be aging. It might be injecting noise into the photo when the lens is fully extended; perhaps when fully mechanically extended, the zoom motor is still drawing power and causing electromagnetic interference.  This may have changed since the last year, because mechanical parts and contacts tend to age poorly, compared to electronics.  Or, a capacitor may be dying; capacitors can slowly change over long periods of time, and that can manifest as a slowly-degrading PSRR.  A camera's imaging sensor, being "sensitive," will naturally pick up on even a very miniscule change in the power-supply noise.
 * The sensor might be aging, but why the symptom should only manifest when fully zoomed is a mystery. CCD sensors, like the one in your camera, read out interleaved lines (similar to interlaced video, but commonly implemented using between 2 and 6 fields); and it looks like you're dropping one entire CCD field.  It's possible that in telephoto mode, the firmware sends a new set of sensor configurations (generally that would be done to improve quality and tune the sensor for the optics); but if the firmware is buggy, that "tuned sensor configuration" might be going haywire ... but why don't we see this in your first photo??
 * A software (or firmware) bug is possible. Between Glendale Ridge and Kinter Withers House, your device changed its reported focal-plane resolution in its EXIF data (despite the same focal length, same focus point, and arguably the same sensor).  This suggests that you updated the camera's internal software.  Knowing anything about the complexity of image processing that goes on inside a camera, but lacking the specific details for your Canon A540, it's nearly impossible to diagnose... but "a software bug might be responsible" is almost tautological in this case; unfortunately this insight doesn't help fix it!  You could check for a new updated firmware from Canon.
 * Absolutely fascinating "glitch." Any other information or symptoms you can provide might hint at the root-cause, and possibly hint at a solution; but realistically, the probable fix will be "replace the entire unit."   Nimur (talk) 02:09, 12 May 2012 (UTC)
 * Oh, throw in one more possibility. The final CCD field might not be "dropped" - perhaps it's just exposed incorrectly.  If so, that could be caused by a timing bug, (a software race condition, for example - though why it occurs only recently is still unexplained); or it might be a defect in the autoexposure metering algorithm (which would also explain why it's only occurring at maximal zoom - autoexposure metering depends on the optical characteristics).  Nimur (talk) 02:12, 12 May 2012 (UTC)

Comparing your old (fine) and new (problematic) files, the following differences can be found in the exif data…

Digital Zoom: None

Digital Zoom Ratio: 1

Digital Zoom Ratio: 2.8

Digital Zoom: x5.5

Focal Plane Horiz Resolution: 12515 dpi

Focal Plane Horiz Resolution: 34417 dpi

Focal Plane Vert Resolution: 34366 dpi

Focal Plane Vert Resolution: 9372 dpi

Image Created: 2011:03:03 10:48:29

Image Created: 2012:02:06 11:33:15

Image Height: 1584

Image Height: 2112

Image Size: Large

Image Size: Unknown

Zoomed Resolution: 1024

Zoomed Resolution: 2816

…suggesting a comparison with another file would probably be more telling. ¦ Reisio (talk) 02:23, 12 May 2012 (UTC)


 * ... or try disabling digital zoom completely, and try to reproduce. Nimur (talk) 02:37, 12 May 2012 (UTC)


 * I see a problem there. It says it's doing digital zoom, which often produces crap, as it must interpolate between pixels.  You need to find a way to prevent it from doing digital zoom. StuRat (talk) 03:13, 12 May 2012 (UTC)


 * See Digital zoom for a comparison (although it doesn't have the horizontal lines). Bubba73 You talkin' to me? 03:40, 12 May 2012 (UTC)


 * I'm sorry for providing confusing information; I thought my camera did digital zoom by reducing the image size (simply cropping the edges), so I interpreted the identical file sizes as an indication that the house photo was only optically zoomed. I can testify that I didn't knowingly change anything with the camera's software; as far as I know, the settings haven't been changed, as both landscape and default modes (which are the only things I use for long-distance shots) have very few changeable settings with which I'm familiar, and I do my best to keep them the same, except for issues such as flash and image resolution.  I'm not a fan of digital zoom, so I avoid it when possible, but whenever I go to the maximum optical zoom, it both moves the lens and gives me a little notice on the viewfinder telling me what the zoom is; this appears when I'm unambiguously on digital zoom and doesn't appear on reduced optical zoom levels.  This is precisely what happened with this image, which I took today; you can see the lines somewhat against the fenceposts on the left side, but they're not as clear elsewhere; perhaps that's because the camera wanted to do flash, as it was cloudy and in a rather wooded neighborhood near dusk.  Most of the more clearly "lined" photos have a greater contrast between light and dark; for example, you can see it more clearly at File:Epsilon II Archaeological Site zoomed.jpg, which I've just uploaded.  Nyttend (talk) 04:46, 13 May 2012 (UTC)


 * OK, so are the lines purely a problem with digital zoom ? If so, you can be careful to avoid using that and we can mark this Q resolved, right ? StuRat (talk) 03:11, 14 May 2012 (UTC)
 * No, it happens when I do my best (1) to zoom in all the way optically and (2) to avoid digital zoom. The "little notice" is something telling me how much it's zoomed in (I think 4x, but I'm not sure), and if I go into digital zoom and then get out of it, the little notice still appears once I'm done with digital zoom.  Since the physical lens doesn't move when you go from maximum-optical-zoom-only mode to digital-zoom-mode, I expect that the notice is there to allow users to know that digital zoom is no longer there.  Nyttend (talk) 03:36, 17 May 2012 (UTC)

Unusual behavior by Agama Navigator 535
Hello there, I am having trouble with my mouse pointer for the past two months. The mouse has suddenly become unusually fast. It's clicking speed, scrolling and moving is way too fast that I can't work with it properly. For example, if I drag and drop a file/ multiple files, instead of moving the files, mouse pointer forces them to open immediately. In browsing, I can't easily select texts or sentences by moving pointer over them. It takes multiple strike to select that text and sentence. If there is a web link in sentences, it opens that link instead of selecting it. I lowered the pointer overall speed from its driver (dpi range is 800 to 2000). Even reinstalled the driver. But it did not work at all. I tried updating its firmware but failed due to the unsupported hardware version. I went to control panel to lower its speed, unplugged it from usb port and re-plugged it. But no changes seemed to appear. I am using Windows 7 ultimate 64 bit version. How can I rectify this problem?--103.10.77.226 (talk) 11:10, 12 May 2012 (UTC)


 * At the risk of sounding obvious, have you tried a different mouse old boy? Could be a hardware rather than a software problem. You should also try a different USB port if you haven't already!. Quintessential British Gentleman (talk) 22:16, 12 May 2012 (UTC)


 * Sorry, I don't have different mouse. I tried in different USB port with the same result.--180.234.194.218 (talk) 23:44, 12 May 2012 (UTC)


 * Well, get one. They're under $20.  I'm almost certain that will solve the problem.  If not, return it.  StuRat (talk) 00:03, 13 May 2012 (UTC)


 * I can't return it. Because I bought it 1 and 4 months ago. The warranty is expired.--180.234.87.61 (talk) 19:39, 14 May 2012 (UTC)


 * I was talking about buying a NEW mouse and returning that mouse if it doesn't work. StuRat (talk) 20:35, 14 May 2012 (UTC)


 * What is the best brand and mouse for FPS and third person shooter/action game? I am not worrying about price. thanks--180.234.87.61 (talk) 21:10, 14 May 2012 (UTC)


 * I won't recommend a specific brand, but do have some general recommendations:


 * 1) Don't use a wireless mouse. They use batteries, and nothing is more annoying than having to change the batteries mid-game.  Many also have a very short range, maybe 2 feet, and go nuts at the range limit.


 * 2) Use a laser mouse, not a ball mouse (I probably didn't need to say this, as ball mice are quite hard to find these days, in any case).


 * 3) Don't get a mouse bristling with buttons. I find it entirely too easy to hit the wrong button on those.  Stick to the basics, maybe 2 buttons and a wheel.


 * 4) Make sure it's supported on your O/S.


 * 5) Use a proper mouse pad, not your pant leg or chair armrest. StuRat (talk) 18:11, 15 May 2012 (UTC)

Capablanca Chess (10x10 version) vs Go:Which is more complex?
Capablanca experimented with at least two larger boards...I'm asking about the 10 by 10 board here. The snap (and perhaps correct)answer is that Go is more complicated than Capa chess, but I'm not so sure. Thanks in advance for your responses.Rich (talk) 11:30, 12 May 2012 (UTC)
 * I just now put this in hopefully more appropriate Miscellaneous reference desk.Rich (talk) 12:03, 12 May 2012 (UTC)

Apple's "Numbers" Spreadsheet
I find this a very unsuccessful and complicated programme which replaced Apples AppleWorks spreadsheet. Is there a basic spreadsheet programme, (free or cheap!) that anyone can recommend please?--85.211.154.241 (talk) 12:15, 12 May 2012 (UTC)


 * LibreOffice Calc / Category:Spreadsheet software ¦ Reisio (talk) 14:22, 12 May 2012 (UTC)
 * For Apple, this should probably be NeoOffice, the marriage of OpenOffice and Aqua. I don't know how far native MacOS-X interface support for the other OpenOffice variants has come. I don't, as a rule, use spreadsheets.1 But from my experience with other iWork programs, I always found them useful, and much easier to use than the corresponding tools from other office suites. I can actually write something in Pages without cringing all the time, and Keynote's consistent rendering of everything, including PDF, is beautiful. So I'd be interested in what is wrong with Numbers. --Stephan Schulz (talk) 15:16, 12 May 2012 (UTC)
 * 1 Either they are just lists of numbers, or incredibly error-prone spaghetti programs. In either case I prefer simple ASCII files and my own error-prone spaghetti code that is at least written sequentially, and does not consist of a two-dimensional global array and code smeared out over all the individual cells.


 * Another alternative is to just ask us what you're having trouble with in Numbers. The odds are that there is an easy way to do what you want, but you just don't know how to do it. (Which I sympathize with.) The only trouble I've had with Numbers is that you can't really customize charts to the degree that I'd like — in trying to make it "easy", they remove a lot of control (like many Apple products). --Mr.98 (talk) 17:30, 12 May 2012 (UTC)

Thanks for the sympathy! My problem mainly relates to printing. I have quite large spreadsheets which will only print on the one page, VERY small requiring a magnifying glass to read, or over a number of pages which are then difficult to read since part of the information is on separate bits of paper. The old AppleWorks never gave the the same problem. I am trying 'OpenOffice' but that seems no better.--85.211.154.241 (talk) 06:16, 13 May 2012 (UTC)
 * Hmmm. That seems to be a fundamental problem - if you need to cram too much content onto one page, things will necessarily get smaller. But you can try to hand-optimize the sheet. Select all columns, then go to Table->Resize columns to fit content to automatically make the columns as small as the data suggests. If that is not satisfactory, go to Table->Allow border selection, then rearrange the borders to fit onto one page. File->Show print view will show you how Numbers will currently distribute content to print pages. For more advanced tips, use the Inspector, as described in this Mac Observer article. --Stephan Schulz (talk) 06:49, 13 May 2012 (UTC)

Great stuff, thanks for the ideas, shall give them a try.85.211.154.241 (talk) 08:03, 13 May 2012 (UTC)

Tools to sort and filter academic papers
I'm subscribed to e-mail alerts for several subjects at arXiv and am receiving more than I can keep up with. What text mining and text classification tools, if any, are available to help sort and filter the new papers and help decide which ones to read right away? Neon Merlin  12:47, 12 May 2012 (UTC)


 * Why not read the abstract of each, and use that to decide ? If you get so many you don't even have time to read those, then I suggest you unsubscribe from some of the less interesting ones. StuRat (talk) 22:21, 12 May 2012 (UTC)


 * I am indeed only reading the abstracts (about 100 per month of them). The reason I don't unsubscribe is that I'm concerned I might be "the last to know" about a paper that totally changes the course of my work (like this one, which I only heard about because someone happened to report it on Engadget, which my stepdad reads). Neon  Merlin  03:26, 14 May 2012 (UTC)


 * Skimming 100 abstracts a month doesn't seem like much, more like a day's worth. I suppose you could use some type of keyword search, but that really isn't as good as reading the abstracts yourself and deciding which academic papers are worth reading in their entirety.  (A keyword search might fail with false negatives because they refer to something by another name, or you forget to enter some keywords, etc.  And false positives occur if your keywords are too common, have multiple meanings, etc.)  If reading all the abstracts is too much, how about if you read all the titles, and only read those abstracts where the title sounds relevant ? StuRat (talk) 03:35, 14 May 2012 (UTC)

Battlefield 3, 32-bit v 64-bit
Battlefield 3 is a 32-bit game (on task manager, is has a *32 and it installs in the x86 Programs files folder). Is there any advantages on installing it on a 64-bit Win 7 machine vs a 32-bit Win7 machine? Thanks Acceptable (talk) 18:16, 12 May 2012 (UTC)


 * Everything else being equal, I don't think so. Bubba73 You talkin' to me? 18:20, 12 May 2012 (UTC)


 * 32-bit Windows has a 4 Gb limit on address space (excl some server variants which can make use of PAE). Of this address-space, some is used for memory-mapped devices, then Windows itself uses a chunk then so does anything else running (e.g. Antivirus, etc) - so there may be 2-3Gb available for the game.
 * In a 64-bit OS the 4Gb limit still applies to 32-bit apps, however each 32-bit process can address 4Gb.
 * Whether or not this improves real-life performance depends on many factors, not least whether the machine has enough physical memory for this to make a difference, and whether the game itself uses that much. Cheers,  d avid p rior  t/c 20:45, 12 May 2012 (UTC)
 * Just to explain this more clearly, your 32-bit win7 install only allows your computer to see about 3.5G of memory. So if Windows and other processes are taking up a lot of this memory, this leaves less available memory for your game to run in (and Windows will have to start swapping between available memory and disk). The 64-bit install will ensure that your game uses all the memory it needs to run efficiently, provided you have enough RAM in your computer. Shortage of RAM is one of the worst performance hogs in computing. Sandman30s (talk) 12:17, 14 May 2012 (UTC)
 * Just for further clarification, AFAIK, there are actually 2 limits here. One is the amount of physical memory addressable by Windows which is limited to 4GB on non server x32 variants of Windows.. But another is the amount of virtual memory available to user mode processes. By default this is split 50/50 meaning only up to 2GB is available to user mode processes and 2GB is available to the kernel. You can adjust this to make 3GB available to user mode processes, at the risk of causing problems because of the kernel being limited to 1GB. Even if you have PAE which isn't available to none server variants, this limit still applies. So if you have a single process that needs to use more then 2GB, even PAE doesn't help on Windows x32 unless you enable the 3GB switch and all that comes with it (or the process supports Address Windowing Extensions which is unlikely for games]]). Of course if our need to use more then 3GB you're SOL. In other words, with a potentially memory intensive game, if you have at least 3GB RAM and definitely 4GB, there is a potential advantage. Of course if the program is not large address aware, then there's minimal advantage (will still likely be some advantage without PAE since you can't remove other processes completely). But I think many modern demanding games are anyway.      (Besides memory, I guess there's also a possible advantage due to the x64 drivers but it's unlikely to be significant.) Nil Einne (talk) 11:43, 18 May 2012 (UTC)

video editing
anyone know any programs I can download to speed up a video of mine, such that when I change the frame rate the sound gets faster too?

As it is, I'm editing in avidemux, which doesn't change the sound, so I have to export the sound file, speed that up too, then join the two together in windows live movie maker, since for some reason if you import a sound file to avidemux it reverts back to the original as soon as you close the menu. it's worked so far, but just this one time, movie maker refuses to open this file, saved in the exact same format as all the others.

Kitutal (talk) 23:40, 12 May 2012 (UTC)


 * Question: Do you just want to speed up the sounds, which will increase the pitch and make everyone sound like chipmunks, or do you want to compensate for that and bring the pitch back down to the original ? StuRat (talk) 00:06, 13 May 2012 (UTC)

same pitch, just taking less time. except it turns out I can do that in windows movie maker, for as long as that keeps working, somehow missed it all this time... Kitutal (talk) 10:00, 13 May 2012 (UTC)