Wikipedia:Reference desk/Archives/Computing/2010 October 28

= October 28 =

3-D graphic video design
does any one know how are the graphics in films like UP created and what is the technology \ framework behind them? --Umar1996 (talk) 12:01, 28 October 2010 (UTC)


 * Pixar films are rendered using PhotoRealistic RenderMan, which uses the Reyes rendering scheme. Pixar typically write dozens of new plugins for that for a given movie (adding new features like better hair, clouds, skin etc.). I don't know what they use for modelling and painting (that is, what the animators use to create characters and paint them with colour). -- Finlay McWalter ☻ Talk 12:09, 28 October 2010 (UTC)


 * I believe they use Pixar Marionette, an in-house software suite. However they sell the tools to export to RenderMan from more standard 3D suites, like Autodesk Maya. --Mr.98 (talk) 13:59, 28 October 2010 (UTC)
 * If you're looking for software that you can use, consider trying Blender (software). Blender is free software designed for 3D modeling, animation, and rendering.  You can use for any purpose, including to make video-game characters, full-length feature-films, and explore the theory and practice of computer animation and rendering.  Though Blender lacks some of the advanced features that an expensive commercial software package may provide, it is still capable of producing impressive pieces of art - for example, Elephants Dream and other movies.  Some major commercial studios even use Blender for some or all of their special-effects work.  Nimur (talk) 14:51, 28 October 2010 (UTC)


 * Yes, and it will certainly get you into thinking about how to use 3D software, the concepts involved, etc., which are a little counterintuitive if you have not used them before (or are only used to 2D graphic programs like Photoshop). The downside of Blender is that it has a notoriously idiosyncratic interface (which is evident by the fact that people will tell you that once you've learned it, you wished every program worked that way — i.e., it's a huge investment of your time to use it, and it's too bad it has absolutely nothing to do with the interfaces of any other computer program you have ever used), and in my opinion pretty poor documentation (the "community" manual and tutorials often leave huge things unexplained or poorly explained), but with some perseverance you can make nice things with it, skills that are presumably (I have not tried to do this) transferrable to programs like Maya or other "industry standards". --Mr.98 (talk) 21:25, 28 October 2010 (UTC)


 * If you're wondering how they make the characters move like people, skeletal animation describes the basic idea. Pixar animators adjust the positions of various control points frame-by-frame, and the software positions the skin, clothing, and hair. It's a lot like stop-motion animation, though the computer dolls are more flexible. I think even Elastigirl from The Incredibles was animated this way, and I remember the commentary track mentioning that the principal and the teacher in this scene are actually the same doll ("universal man"). Sometimes the control points are placed by motion capture instead of manually by animators, but I think Pixar never uses motion capture. -- BenRG (talk) 05:01, 29 October 2010 (UTC)


 * thanks for the responses, i think i will try Blender --Umar1996 (talk) 09:33, 29 October 2010 (UTC)

Greasemonkey for IE interface
I don't have Windows/IE, so I would like to know from others if this is possible... I was asked if it would be easy to write a program to make a button in IE that a user can click on. When he does so, a modular window pops up where he can type in a keyword. When he does that, it hides all of the monstrous (50+ page) webpage he is viewing except the sections around the keyword he types. Optimally, he wants to be able to set how much before/after the keyword he wants to be seen. Can Greasemonkey for IE do the user interface thing? I know it can hide content on the fly. -- k a i n a w &trade; 17:24, 28 October 2010 (UTC)
 * Greasemonkey is essentially just a tool to automatically run a bookmarklet when the page matches a given URL. What you describe would be straightforward to do in Javascript, so could easily be made a bookmarklet, which would then be the "button" you want. --Sean 14:27, 29 October 2010 (UTC)
 * Actually, I see that the example on the bookmarklet article gets you most of the way there! --Sean 14:29, 29 October 2010 (UTC)

Scrap
I remember in Windows ME, if you copied an image or text or whatever to the clipboard, then tried to paste it like a file, a "scrap" file would be produced with the contents. What was this called? Why did Microsoft disable the feature on later versions of Windows? 82.44.55.25 (talk) 17:41, 28 October 2010 (UTC)


 * Here on Wikipedia, our coverage of this is pretty poor; it is touched on at Microsoft Windows library files. That article does have an old reference from microsoft.com entitled "What is a Scrap File?"  Comet Tuttle (talk) 17:44, 28 October 2010 (UTC)


 * It is called scraps (or shell scraps). As of why the feature was removed, this article might bring some light. 118.96.156.101 (talk) 23:12, 28 October 2010 (UTC)

Audio playback skipping when a program crashes
A strange question that you all might be able to help me with. Under some circumstances, when a program or system crashes, the audio that was being played back will repeat the same ~1/4th second of playback repeatedly. This can occur, for example, when a game freezes, or the OS freezes while playing back audio in Winamp/iTunes, etc. Typically, when such a freeze occurs, the video will also freeze, but it will only display a single frozen frame (instead of repeating the last ~5 frames). What causes audio "freezing" to differ in this regard?

I suppose I would expect the audio to playback the single note that was being played at the time of the freeze, but it will instead repeat that same 1/4th second. Does this have something to do with the way that audio cards handle audio playback? Does it depend on the amount of audio data stored in RAM? Thanks! Sutarion (talk) 21:04, 28 October 2010 (UTC)


 * Devices like audio cards and ethernet cards have a buffer (often a ring buffer) which stores the data they're currently sending. When streaming data, the hardware takes data from this and squirts it out the relevant port and decreases a counter. When that counter (which counts how much data is available in the buffer) gets too low, the hardware signals the operating system (and thence generally a software application) that more data needs to be written into the buffer. That application should respond reasonably promptly by adding more data. If the application is defective, or has crashed, then it might not be available to refill that buffer.  This condition is often called a "buffer underrun". Different hardware deals with an underrun in different ways - ideally it'd detect that there's no more data, and just go quiet (that's usually what an ethernet card will do); you'd really hope sound hardware would also just stop doing stuff, but some seem to just replay the ring buffer's contents over and over, until someone (often the operating system) closes them down.  The duration of that repeat will vary (for a sound device) depending on the size of the ring buffer and the format of samples stored in it (if the audio is using high-resolution samples at a fast sampling rate, it'll run out of samples quickly - if the samples are smaller and less frequent, you'll hear a longer fragment of audio before the repeat). -- Finlay McWalter ☻ Talk 21:13, 28 October 2010 (UTC)


 * If you're wondering why a buffer is used (why the program just doesn't play the sound data as it's needed) consider this analogy - say you owned a small shop, and you employed a worker to stock the shelves with cans. You could give the worker a can, have them go put it on the shelf, and then come back to you. But they have to interrupt you constantly, because they need another can, so you can't easily get on with your own work.  This used to be the case with simple UARTs (the chips that drove serial ports). As the ports got faster (the worker got quicker) the pace of interrupts built up so the boss was overloaded just giving the worker one can each time.  The solution is to give the worker a box of cans, and the worker only comes back when the whole box is gone. This makes for far fewer interrupts, but requires that the worker be a little bit smarter. -- Finlay McWalter ☻ Talk 21:18, 28 October 2010 (UTC)


 * Thanks for the quick response! That makes perfect sense to me. Thank goodness the ethernet ports typically go quiet, as that could be quite messy if the application wasn't prepared for it. I suppose audio repeats (while annoying) don't cause any major problems. Sutarion (talk) 21:22, 28 October 2010 (UTC)


 * If the software instructed the hardware to play notes then you probably would hear the most recent note played forever. You might hear that if you were playing a MIDI file and the player crashed. But most software these days sends PCM audio (like CD audio) to the sound card. Sound effects, background music, voiceovers etc. are mixed together in software, not sent to the card as instruments.


 * With video there's also usually a ring buffer of frames, but the software explicitly signals the hardware to switch frames. There are a few reasons for the difference. The video frame rate is probably 120 Hz max, and different frames may take significantly different amounts of time to draw. The audio sample rate is probably 44,100 or 48,000 Hz and needs to be kept constant, since a variation of even a few percent would be noticeable as a change of pitch. Occasionally a software malfunction will lead to the video ring buffer being visible. An example is the hall-of-mirrors effect in DOOM. -- BenRG (talk) 21:26, 28 October 2010 (UTC)

Mail.app sending phantom duplicates
My mail.app on OS X (10.5.8) is behaving strangely. Sometimes, *but not always*, when I use the reply button to respond to an email (to a single address), the mail activity panel shows two messages being sent in rapid succession. The 'sent mail' folder only shows one. Can anyone explain this? I think it might have something to do with the way nested/threaded replies are handled, but could it be some type of malware? (None of my contacts have complained about spam from my address.) Any thoughts are appreciated; thanks! SemanticMantis (talk) 22:20, 28 October 2010 (UTC)


 * I've seen this as well: I always put it down to IMAP synchronization or cross-checking (Mail.app's efforts to make sure that everything on the remote machine is duplicated properly locally, and vice-versa). It's unlikely to be malware (OS X still doesn't have any native malware in the wild, and the problem's been around for a couple of years now, so someone would have screamed about it if it were).  It might be (insert your favorite conspiracy/space-alien theory here), but if it is there's not a damned thing you can do about it, so why worry?    -- Ludwigs 2  22:50, 28 October 2010 (UTC)


 * Thanks! Good to know it's a known issue. SemanticMantis (talk) 00:07, 29 October 2010 (UTC)


 * You could always use wireshark to monitor the SMTP, IMAP4, and POP3 protocols between your Mac and your ISP's mailserver. Are you using IMAP or POP3 to receive email? Do you have autoreply to 'receive receipts' enabled (most mail clients enable it by default). It could be your client fetching new emails after sending anything and possibly autoreplying. CS Miller (talk) 09:56, 29 October 2010 (UTC)

Installation of x8 PCIe 2.0 Card in x8 PCIe 1.0 (x4 Electrical) Slot
Hello. I have an Intel S5520SCR Workstation Motherboard, which has the following expansion slots. Both of the x16 PCIe Slots are occupied, and I want to install an Intel RS2BL080 RAID Controller Card, which has a x8 PCIe 2.0 interface, in the motherboard's one x8 PCIe 1.0 (x4 Electrical) Slot. My question is whether will this configuration work - will the motherboard enumerate the card, will I be able to set up a RAID 6 array using this card, and is there likely to be any problem? Any and all help is appreciated. Thank you very much. Nasi.Lemak.Kastari (talk) 22:44, 28 October 2010 (UTC)
 * two x16 PCIe 2.0 (x16 Electrical) Slots
 * one x8 PCIe 1.0 (x4 Electrical) Slot
 * one x4 PCIe 1.0 (x1 Electrical) Slot
 * one PCI 32/33 5V Slot

PPT font
I would like Garamond to be my default font in PPT but no matter how many times I select it, the default remains Ariel. What am I to do?  DRosenbach  ( Talk 23:38, 28 October 2010 (UTC)


 * Edit the design template (slide masters) by following these instructions from Microsoft. Then (and this is critical) ... specify your new template as the default template by following these instructions from Microsoft.  This is the trick - you have to save your template as blank.pot. Nimur (talk) 23:53, 28 October 2010 (UTC)
 * Woohoo!!! Thanx so much!  DRosenbach  ( Talk 03:25, 29 October 2010 (UTC)


 * Also notice the subtle difference between Arial and Ariel. --Andreas Rejbrand (talk) 09:21, 29 October 2010 (UTC)