00:00:25 -!- Sgeo_ has quit (Read error: Connection reset by peer). 00:01:20 exit 00:01:24 BWAHAHAHAHA 00:01:25 -!- catseye has quit (Quit: leaving). 00:02:12 -!- Sgeo has joined. 00:09:10 i assume that means we are doomed. 00:09:45 oerjan, we always been 00:10:26 -!- catseye has joined. 00:11:00 Mounting NTFS filesystems read-write should be stable on Ibex? Do you think so, what is your opinion? 00:12:03 catseye, "no" 00:12:20 catseye, on karmic or later with ntfs-3g "yes" 00:12:53 catseye, but I wouldn't do it on ibex 00:12:55 Vorpal, "fuck" 00:13:06 catseye, just upgrade ... duh 00:13:21 DUH 00:14:00 no risk of losing the files I'm trying to copy onto the NTFS filesystem by doing that 00:14:01 catseye, do you have a large enough usb drive to hold a disk image of lucid or such? And an external hdd to back the whole disk to? 00:14:02 if so 00:14:06 boot usb stick 00:14:24 dd if=/dev/sda1 /media/external-drive/whatever 00:14:26 and so on 00:14:30 for each partition 00:14:31 I has a purry lap kitty! 00:14:48 Gregor: so has me! aw she left 00:14:49 catseye, once per partition since that is easier to loop-mount 00:15:16 wait wait wait 00:15:21 catseye: it is clear that you need to sacrifice the goat here 00:15:24 i can boot 10.10 live off the usb stick 00:15:38 and mount both partitions r/w under that boot 00:15:54 catseye, I would still be careful and back the partition up first 00:15:56 -!- augur has quit (Remote host closed the connection). 00:16:45 Vorpal: it's backed up. I made sure to back it up before restoring Windows. But I'd like to get the files onto DVD (I don't entirely trust external HDDs after my experience with my last one.) 00:16:46 catseye, and remember: do it using ntfs-3g, not using the in-kernel drivers 00:17:33 Vorpal: how do I tell the difference? in fstab (in Ibex) it just says 'ntfs-3g' 00:18:16 catseye, ibex has early ntfs-3g? 00:18:17 well 00:18:21 I would not trust it 00:19:00 catseye, I mounted ntfs with ntfs-3g using the ntfs-3g command, never by fstab 00:19:45 Some of the programs I have written in Enhanced CWEB, it seems it should have bibliography section. 00:20:06 zzo38, use bibtex then? 00:20:32 -!- augur has joined. 00:20:49 I would assume an ntfs-3g fstab entry is handled by the ntfs-3g command 00:21:13 catseye, same 00:21:24 catseye, I wouldn't trust such an old version as that of ibex though 00:21:29 ok, trying-this time 00:21:34 -!- catseye has quit (Quit: trying-this time!). 00:26:06 -!- catseye has joined. 00:26:14 llo llo 00:27:21 oh, too easy. 00:27:34 both drives are just available from the 'Places' menu. sweet. 00:29:30 Vorpal: BibTeX is one possibility. Or perhaps I can just write a TeX macro for bibliography, instead. 00:29:55 -!- SgeoN1 has quit (Quit: Bye). 00:32:23 mhm 00:32:25 night → 00:39:37 I GOT ACTUAL 24P OUT OF THIS 00:40:42 filmdint followed by phase=U. filmdint is a hybrid inverse teleciner and deinterlacer, and phase=U attempts to autodetect and correct interlacing that's at the wrong phase. 00:44:31 The *only* problem? Some of the frames are deinterlacer output, because for no good reason, some frames appear to only possess the bottom portions of the source 24p frame. 00:45:23 Ah, running phase=U before as well fixes some of *that*. 00:45:53 I AM VICTORIOUS 00:46:21 * Sgeo goes to install LyX 00:46:41 zzo38, what's your opinion of LyX? 00:47:00 Better-than-source encode FTW. 00:48:20 Sgeo: I have not used it, but from what I see, I would prefer typing TeX codes in a text editor instead, including macros and category codes and all of that other stuff. 00:48:45 (Category codes is a feature of TeX which I use often.) 00:49:54 I suppose LyX is OK if you want to use LaTeX, though. But I don't use LaTeX. 00:50:43 (Because I think Plain TeX is superior) 01:12:42 -!- comex has quit (Ping timeout: 240 seconds). 01:14:30 -!- quintopia has joined. 01:14:51 it is the millennium+10, and there is no more time for Mandelbrot 01:15:27 the future is _so_ last decade 01:25:03 * pikhq wonders why 25i video doesn't use 2:2:2:2:2:2:2:2:2:2:2:3 pulldown for 24p display 01:27:11 -!- wareya_ has joined. 01:30:14 -!- wareya has quit (Ping timeout: 276 seconds). 01:36:43 Note to self: Don't accept defaults just because they sound vaguely reasonable 01:37:00 Asking me for each package whether or not to install it is NOT reasonable 01:37:15 * Sgeo unchecks something or other 01:41:03 elliott: I think we can safely assume NetBSD will not be happening tonight. 01:42:14 -!- FireFly has quit (Quit: swatted to death). 01:43:37 bac 01:43:39 *back 01:43:42 catseye: aww. why? :( 01:44:04 Sgeo: what are you installing? 01:44:26 LyX, which needed to install MikTeX 01:44:44 Sgeo: installing all latex packages is a good idea 01:44:52 hey i just realised, i'm on at the same time every day 01:44:59 since i wake up on weekends the same time i get back from school 01:45:20 You may have a sleep disorder. 01:46:12 pikhq: i deem it "teenagerism" 01:46:15 I may have a sleeping disorder too. 01:46:27 pikhq: The great thing is, I can't medicate it legally! 01:46:28 `addquote hey i just realised, i'm on at the same time every day You may have a sleep disorder. 01:46:37 pikhq: Melatonin is a prescription-only drug over here. 01:46:38 nah, I'm just not sleeping properly, it's not a disorder 01:46:45 oerjan: :D 01:46:52 And convincing a GP to prescribe melatonin... hahahhahahahahahaha never. happening. 01:46:59 dammit, hackego isn't here 01:48:24 * Sgeo should probably make some effort to do some homwork 01:48:51 * olsner makes an effort to sleep 01:51:29 -!- zzo38 has quit (Quit: zzo38). 01:51:53 slackego! 01:52:12 elliott: well, depends on how much longer it takes to copy, burn, etc. 01:52:47 * coppro wants to implement a generic tactical game engine in Haskell 01:54:38 MikTeX isn't LaTeX? 01:55:00 coppro: in the type system! 01:55:05 * oerjan runs away 01:55:23 ab-dab 01:55:31 oerjan: ofc 01:55:46 amusingly, it leads to a precise definition of a tactical game 01:59:49 elliott, is Computer Modern an acceptable font to you? 02:00:12 Sgeo: for what? 02:00:15 your question is beyond vague 02:00:27 coppro: in the type system! 02:00:32 I want to write a typechecker. 02:00:32 To look at/use/whatever 02:00:33 In the type system! 02:00:36 Sgeo: for *what*? 02:00:39 what document? 02:00:58 ...good point 02:01:15 Is it visually pleasing with Lorem Ipsum? 02:01:26 elliott: the AI Emancipation Proclamation 02:01:44 Sgeo: I refuse to answer typography questions about Lorem Ipsum... 02:03:54 Ipso Facto 02:04:01 * oerjan now with recycling 02:06:09 I assume that Yap is a DVI viewer? 02:06:58 pikhq: http://fbpanel.sourceforge.net/index.html 02:07:11 Sgeo: yes 02:07:17 "Yet Another Previewer" 02:11:56 pikhq: Somehow setting the resolution with xrandr and rebooting... works. 02:12:11 * Sgeo decides to redo the LyX tutorial 02:12:19 It's been years since I've touched LyX 02:13:01 elliott: ? 02:13:07 elliott: That.. And. Why? 02:13:13 pikhq: I don't even know. 02:13:18 pikhq: It... it remembers. 02:13:20 How does it remember? 02:13:22 I DON'T KNOW 02:14:04 pikhq: Anyway, fbpanel looks like a nice dependency-free panel. 02:19:12 phase,filmdint,phase,telecine,phase,pullup,softskip 02:19:38 I can't *believe* that is the fucking chain I need to put it through to inverse telecine it without noticable artifacting. 02:21:12 pikhq: LeanDE (Kitten's totally approverated desktop environment) as it is right now: lwm (perhaps a forked, slightly-enhanced version), fbpanel, Midori, ? 02:21:23 And now I can stop stopping the encode with a better idea. Because this is actually perfect. 02:21:40 Might tweak fbpanel a bit so that the current window is actually de-pressed. 02:21:41 But yeah. 02:22:45 pikhq: ...fbpanel tries to use slock by default :D 02:22:50 suckless' locking program. 02:23:06 TOTALLY APPROVE 02:24:04 pikhq: Heh, LXDE's lxpanel is a fork of fbpanel. 02:24:09 Except made to be more RUBBISH LXDE CRAP 02:24:27 God dammit. It's mixed telecine and interlaced. Because it hates me. HATES HATE HATES. 02:24:37 pikhq: More better ideas, eh? 02:24:52 elliott: No, just hatred. 02:25:52 pikhq: Oh lawl, fbpanel opens applications you click as child processes. 02:25:56 pikhq: Totally needs some patching. 02:26:00 Fuck it. What's the best deinterlacing filter for animation? 02:27:55 Not telecining in the first place X-P 02:28:26 pikhq: You should totally install lwm and admire the awesomely-designed WM. 02:28:36 Elliott H. represent 02:28:38 yay i now have a new sound test for bytepusher which doesn't have an irritating female robot voice 02:28:47 now to get some gfx up 02:29:46 Linear blend appears to look best for animation. 02:29:47 TO ENCODING 02:30:06 And *still* needing the phase filter. 02:30:08 elliott: Mmm. 02:30:43 Also features a Sufficiently Snarky Developer: 02:30:49 "lwm is a window manager for X that tries to keep out of your face. There are no icons, no button bars, no icon docks, no root menus, no nothing: if you want all that, then other programs can provide it. There's no configurability either: if you want that, you want a different window manager; one that helps your operating system in its evil conquest of your disc space and its annexation of your physical memory." 02:30:56 (Note: It has a few X resources to configure it.) 02:31:02 Goodnight; bye. 02:31:03 Gregor: Oh, it's telecined. And then edited. And post processed. And then it has 30i cell-shaded 3D overlayed. 02:31:07 -!- elliott has quit (Quit: Connection reset by peer review). 02:31:16 Gregor: And for kicks, it's got the wrong interlacing. 02:32:07 pikhq: So presumably this is something from one source used in the background of a different thing? 02:32:46 (That's just the only way I can imagine it being that broken :P 02:33:08 Gregor: It's just Invader Zim, off of DVD. 02:33:17 They just *put it to disc like that*. 02:33:36 Sweet. 02:34:23 The screwed up interlacing absolutely astonishes me, though. I can't believe they managed to screw up the field order. 02:34:43 lol, seriously? That's ... pretty broken. 02:35:17 Yes. The mencoder phase filter was designed to fix that. 02:36:12 ... FOR PAL, BECAUSE THATS THE ONLY PLACE THAT HAPPENS NORMALLY. 02:36:23 (they use the opposite field order from NTSC) 02:38:03 Hm 02:39:01 phase filter documentation doesn't MENTION it only working in pal :P 02:48:46 Mmmmaybe Mmmmmaverick can burn an ISO to DVD for mmmmme...? 02:49:16 trying 02:50:42 Gregor: No, it's *meant* for PAL. 02:51:17 it's going! 02:51:19 But it's flexible. 02:52:25 actually i should probably consider modifying it to play part of a rendering of cd_orbit.mod 03:00:17 tworked! I have a NetBSD 5.0.2 install DVD. Quite glad I did not purchase the external DVD burner; it was just some problem in intrepid. 03:03:21 D'awww. ATSC allows for h264, but most ATSC TVs can't display it, so nobody broadcasts in it. 03:03:41 (h264 requires about half the bitrate for the same quality as MPEG2...) 03:10:14 It's also crazy-expensive to {en,de}code. 03:10:51 Well, it's CRAZY-expensive to encode. It's merely slightly more expensive to decode than what people are willing to put into TVs. 03:12:34 - 03:12:45 Yes. Behold my minus sign. 03:13:15 would VP8 or theora be expensive to decode? 03:13:21 kinda wondering 03:13:51 then again, tv manufacturers and broadcasters can probably pay for patent royalties 03:14:13 IIRC Theora is worse than it ought to be, VP8 is pretty OK, both are more expensive than MPEG2 or MPEG4, and less than H.264 (it doesn't take much to be less than H.264) 03:14:42 i quite like theora 03:15:50 Gregor: VP8 is actually only *slightly* less expensive than H.264. 03:16:02 Gregor: Because it's fundamentally H.264 without the patent violating bits. 03:16:19 (at least, as far as is known) 03:16:22 Oh, is it? That's ... weird, I thought it was totally unrelated. 03:16:46 Nope. ffmpeg's VP8 encoder is a hack of its h264 encoder. 03:16:47 :) 03:17:21 Hm. 03:17:35 Also, it's only expensive to decode in software. With a hardware decoder, it's not at all difficult. 03:17:47 (they're fucking putting h.264 decoders in *phones* these days, man.) 03:21:56 * Gregor is failing to find any corroborating evidence of this comparison. 03:24:57 http://pics.spaceghetto.st/images/k9dnf.jpg sfw 03:26:55 Gregor: My fucking cell phone decodes h.264. 03:27:24 pikhq: I'm talking about corroborating evidence of the H.264-to-VP8 similarity. 03:29:12 * Gregor goes back to not caring :P 03:30:11 I know someone who works of ffmpeg 03:30:14 *on 03:30:34 He was suggesting that they rename the fastest test profile from 'veryfast' to 'plaid' 03:30:53 X-D 03:30:58 We've gone to plaid! 03:30:59 Gregor: Oh. 03:31:01 oerjan: I'm trying to figure out bits of category theory and bits of domain theory so I can understand what denotational arrows would be. I think I'm inching closer... 03:31:17 pikhq: My phone decodes H.264 too :P 03:31:31 (Probably in hardware) 03:32:42 Gregor: http://x264dev.multimedia.cx/archives/377 03:33:37 Heh, just searching for "H.264" makes it pretty clear :P 03:33:49 So it's not that it's "based on" H.264 per se, it just rips off almost every aspect of it. 03:33:54 "VP8 is simply way too similar to H.264: a pithy, if slightly inaccurate, description of VP8 would be “H.264 Baseline Profile with a better entropy coder”." 03:34:00 ok, brasero has spent twenty minutes "Creating image checksum" for a 650M image. I think it's broken. 03:34:46 On a slightly tangential note, it's a bit upsetting what happened to VC-1. 03:35:01 Microsoft tried to make a patent-free video codec. 03:35:15 Aaand patent trolls found some patents and set up a licensing group. 03:36:05 Same thing that happens whenever anybody tries to make anything patent-free. 03:36:28 Except when you go to great pains to make it unique or use old techniques. 03:36:35 (Vorbis and Theora, respectively) 03:36:36 patent trolls are the sweetest kind of troll 03:37:38 Well. Theora's not *entirely* without patents. It just has a perpetual license granted to all of humanity for those patents... 03:37:53 -!- quintopia has left (?). 03:39:07 trick for bytepusher programmers: if you're doing just audio and no graphics, set the framebuffer to point to code where stuff is happening or something 03:39:37 also it's not a dumb idea to overlap the video and audio buffers 03:41:39 that's... not always what i want to see when listening to audio, but, point taken 03:45:47 Hmm. MP3 should be patent-free soon. Awesome. 03:47:09 Though it's always ambiguous, it's *likely* that it'll be patent free by December 2012. 03:47:48 Likewise for MP1 and MP2. 03:48:18 http://pubacc.wilcox-tech.com/~greaser/stuff/bytepusher/cd_orbit.BytePusher.gz 03:48:24 -!- bsmntbombdood has joined. 03:48:29 sample not mine 03:50:44 There should be a video format that's actually just a bytecode format. 03:50:57 You include both the video and the code to decode and play it. 03:51:27 That would be wonderful for certain things. 03:51:43 Say, recordings from an emulator. 03:51:56 The "video" consists of the emulation for the video & sound chips and their inputs. 03:52:00 Vector video. 03:52:07 And vector video, yes. 03:52:20 From a vector arcade game. :P 03:52:32 And if you want conventional video, you stick a $YOUR_FAVORITE_CODEC decoder in there and you're done. 03:57:26 -!- augur has quit (Ping timeout: 264 seconds). 04:03:41 -!- augur has joined. 04:05:58 And now I am truly, 100% satisified with how this encode is going. 04:06:10 IT TOOK ALL FUCKING DAY just to figure out how to make it not suck. 04:08:30 -!- catseye has quit (Quit: leaving). 04:09:41 Gregor, malicious videos1 04:09:45 Well, I guess not 04:10:09 Or, actually. How much memory can the language request? 04:10:40 Way to get back your malicious videos through our good friend DoS :P 04:11:29 Sgeo: I'd imagine "as much as it can take before the video player comes to cockpunch the asshat that made the video." 04:12:18 Man, if I knew 100% more about video encoding but also knew what I do about PL, I'd totally make the procedural video codec. 04:12:26 Oh god it's PVC agaaaaaaaaaaain 04:12:49 again? 04:13:05 PVC got mentioned as a coincidental initialism a couple days ago :P 04:14:02 Gregor: Probably the hardest thing about video encoding is realising that the patent trolls have made it nearly impossible to do anything in the field. 04:14:32 -!- augur has quit (Ping timeout: 265 seconds). 04:15:13 That's the whole point of PlasticVideo :P 04:15:28 Alternately, flee the country. 04:15:44 PlasticVideo? 04:16:00 Sgeo: Procedural Video Codec -> PVC -> PlasticVideo 04:16:11 Somehow, I doubt that I'm interested in learning about how plastic gets.... Oh! 04:20:22 I guess it's not really a codec at all, is it :P 04:20:32 Well, maybe it is ... sort of. 04:20:45 It is to codecs as Fythe is to languages. 04:20:59 FyThE 04:21:16 Why is that sort of capitalization horrible everywhere else, but acceptable for TeX and LaTeX? 04:21:31 because Knuth is special 04:22:44 KnUtH 04:22:57 DoNaLd KnUtH 04:23:04 You mean KNuTH. Vowels are lower-case. 04:33:34 -!- catseye has joined. 04:33:57 Oh, huh. MP2 is actually patent free. 04:34:20 the audio codec? 04:34:25 Yeah. 04:34:36 Not by design, just coincidence. 04:36:06 oerjan: maybe, if you have a notion of reduction in your language, you can have one object per term in the language, and morphisms to all the terms it would reduce to. 04:36:46 *one morphism for each term it could reduce to. 04:36:51 this is rather simplistic 04:37:26 Isn't MP2 audio kinda ... garbage? 04:37:52 Gregor: No. It's actually in common use in broadcasting still. 04:38:39 quite so. 04:39:25 And it is most comparable to AAC, Vorbis, or other such modern codecs, starting at about 128 kbit/s. 04:39:45 MP3 took off because it has better performance at lower bit rates. 04:41:49 (MP2 was designed to have very high quality sound at 192 kbit/s and transparency at 256 kbit/s with the reference encoder, MP3 was designed to have *acceptable* performance at 128 kbit/s. Aaand then encoders improved.) 04:42:58 Oh, and MP3 is actually incapable of transparency. Literally incapable of it. 04:43:46 * Gregor wonders vaguely what it means for audio to be transparent. 04:44:09 Gregor: Perceptually identical to the uncompressed source. 04:44:24 Err ... that seems ... really difficult to define. 04:44:29 AB testing. 04:44:47 But different people have different sensitivities. 04:44:48 MP3 always has artifacting, it's just a matter of how annoying it is. 04:45:05 same as JPEG 04:45:08 * coppro twitches 04:46:48 coppro: It's actually possible to have lossless JPEG, I *think*. 04:47:20 There is a /modification/ of JPEG to be lossless. 04:48:08 Gregor: No, the original JPEG standard defines a lossless scheme. 04:48:35 Yeah, which is totally unrelated to the non-lossless one and isn't implemented by anyone :P 04:48:43 transparency? 04:48:48 lawl 04:48:49 Oh 04:49:30 And I'm pretty sure ordinary JPEG can be made lossless-modulo-rounding-errors just by setting the quantisation matrix right. 04:49:35 How can it be impossible for something to have transparency? Is it always known that artifacts will be of a certain size? 04:50:02 oerjan: but if your arrows take programs (representations) to their meanings (denotations), those are really two different kinds of object, so how is that a category, unless you do some sleight-of-hand? that's where I'm stuck at atm 04:50:02 *at least of a certain size 04:50:09 Sgeo: Yes, this is the case for MP3. The best it can have is not having very bad artifacting on certain sources. 04:51:23 maybe make it a functor and programs and meanings different categories 04:51:55 oerjan: But "denotational functor" just doesn't have the same ring to it! oh well. 04:52:30 no but the arrows could be the arrows in the meaning/denotation category 04:53:07 oh, er, ok... i suppose they could... 04:53:22 * catseye tries to visualize what those would mean 04:54:06 compositionof meanings, i suppose 04:54:32 ok, i need to appreciate more about functors for this to work itself out in my head now 05:01:34 -!- augur has joined. 05:14:52 My God. I'm getting transparency at 450kbps. 05:15:27 I LOVE X264! 05:15:47 -!- augur has quit (Ping timeout: 276 seconds). 05:19:31 Your input is cell-shaded, yes? 05:21:46 Yes. Still quite impressive. 05:21:53 Well, cell-shaded, animated on cells, etc. 05:22:14 You'd think that'd be easy to compress. 05:22:26 The problem is, most encoders fail horribly at this. 05:22:29 Except that hard borders are the WORST things to compress. 05:22:41 Yup. 05:23:03 Everything else is quite nice about it, but hard borders require a lot of pain. 05:23:07 Which is why I'm finding 450kbps semi-dubious, since I usually go for 1200 :P 05:23:30 I'm telling x264 to go for a target quality instead of a bitrate. 05:24:00 And it popped out a 450kbps average for this video. And it looks wonderful. 05:24:04 Hm. 05:24:39 What's the native format? 05:24:49 s/native/source/ 05:24:51 MPEG2 complying to DVD specs. 05:24:52 I'm assuming video has to be stored as something 05:25:18 Hmm 05:25:26 Which needed fiddling to come out well. 05:25:35 How would PlasticVideo handle Worms replay files? Watching them requires having the game :( 05:26:00 Might be a bit massive even to include a stripped down copy of the game with each file 05:26:24 Sgeo: You would have to convert them to some layers-and-sprites format. 05:26:50 It seems to be varying between 450 and 650 kbps average for these episodes. 05:28:20 Depending on the amount of cell-shaded 3D stuff in the episode, it seems. 05:29:16 pikhq: How does one make x264 do constant-quality? :P 05:29:40 Gregor: crf option. 05:30:00 Ah, there 'tis. 05:30:03 What crf do you use? 05:30:28 The default, which is 23.0. 05:30:42 I hate Chrome's hatred of Reddit 05:30:48 So... Just don't pass a bitrate option. 05:30:53 -!- oerjan has quit (Quit: Good night). 05:31:06 I'm currently encoding with -x264encopts tune=animation:preset=slow 05:31:32 The preset & tune options are by far my favorite recent feature. 05:32:17 * Gregor triest to remember why he added bitrate= to his script in the first place... 05:32:44 The crf rate control thing is somewhat recent, I think. 05:33:13 I know my mencoder.conf was written some 3 years ago according to what was recommended practice *then*... 05:33:49 Mmm. 05:33:50 And I only changed it today because 2-pass encoding stopped working with my settings. 05:34:09 And then I found out that I can completely omit 2-pass encoding now. :D 05:38:06 Oh? Doesn't 2-pass still do better even with constant-quality? (Better bitrate at the same quality) 05:39:17 2-pass is mutually incompatible with that. 05:39:33 Erm. Just incompatible. 05:39:34 Anyways. 05:40:21 All 2-pass does is try and get better quality from the video while meeting the *precise* bitrate you asked for. 05:40:49 Which makes it both look better while shooting for a specific bitrate and makes you actually *get* the bitrate you asked for. 05:41:37 Really handy if you want to try and make the video fit onto, say, a Bluray disc, sucky if you just want it to look good and don't care about file size consistency. 05:43:41 Right. 05:51:22 Is Codu frozen upgrading GHC? :P 05:54:32 Did I make an idiot out of myself, or is does my attempted statistics reference make sense? 05:54:33 http://www.fark.com/cgi/comments.pl?IDLink=5695862 05:55:15 s/or/xor/ 05:55:28 s/xor/fuck you/ 06:04:11 Y'know what I love? Things that segfault unless you run them under gdb. 06:04:12 -!- sftp_ has quit (Remote host closed the connection). 06:16:07 Hmm. There is one noticable artifact here. I can see artifacts from the 4:2:0 chroma subsampling. 06:17:18 But, that's in the source. Can't do anything about that. 06:18:06 Artifacts on DVDs = fail. 06:18:48 Artifacts from the chroma subsampling are actually pretty impossible to avoid if the original video is just right. 06:19:09 (read: animated) 06:19:28 That's happening long before any information is actually thrown away by the codec. 06:19:52 chroma subsampling? 06:20:22 Sgeo: The chrominance on many digital video signals is stored at a lower resolution than the luminance. 06:20:46 chrominance? Anything like hue? 06:20:48 36/8 and 42/8 allocated to APNIC. 7 /8s to allocate (+5 to distribute) remain. 06:20:53 -!- augur has joined. 06:20:53 That's the color information. 06:21:21 luminance is how dark/bright it is? 06:21:24 Yup. 06:21:50 The reason that's done is that the human eye is more sensitive to luminance than to chrominance. 06:21:56 Notebook's yelling at me 06:22:04 And it was first done in NTSC... 06:23:33 * Sgeo turns off compy 06:23:37 Nightynight 06:24:29 Anyways, it's somewhat noticable on animation because the hard edges between regions can sometimes result in showing an off color on those edges. 06:25:34 -!- Sgeo has quit (Read error: Connection reset by peer). 06:32:10 Huh. It is possible to donate to reduce the US national debt. 06:33:06 Just write a check to: Attn Dept G / Bureau of the Public Debt / P.O. Box 2188 / Parkersburg, WV 26106-2188. (source: http://www.treasurydirect.gov/govt/resources/faq/faq_publicdebt.htm#DebtFinance) 06:33:43 -!- augur has quit (Quit: Leaving...). 06:33:44 Perhaps Republicans should know about that. 06:33:44 :P 06:34:02 -!- augur has joined. 06:39:45 When "drop in a bucket" meets "single molecule of water on Jupiter" 06:44:15 Yeah, but if they're poorer they'll have less influence on politics. 06:44:15 :P 06:49:34 My crf=23 gave me 1700kbps :P 06:49:39 Higher than the 1200 I specified. 06:50:00 (I figured you might want me to inform you that 1700 is in fact greater than 1200) 06:51:03 The resulting bitrate is quite dependent on the source quality. :) 06:51:13 Erm, the source. 06:51:20 BTW, were you doing that with preset=slow? 06:54:39 Nope. 06:54:42 Just no options at all :P 06:54:59 And the source is really low-quality, it's a video from my PowerShot X-P 06:55:44 Setting it to preset=slow should improve the quality/bitrate ratio. 07:00:02 -!- augur has quit (Ping timeout: 255 seconds). 07:02:42 -!- tombom has joined. 07:05:45 I think Codu's hard disk access got super-slow. 07:05:53 And that's why all the blehness. 07:13:15 -!- FireFly has joined. 07:29:42 -!- Gregor has quit (Ping timeout: 240 seconds). 07:36:31 -!- Gregor has joined. 07:36:54 Yup, Codu is definitely self-destructing X-D 07:38:04 -!- Gregor` has joined. 07:39:08 HI ME! 07:50:37 -!- augur has joined. 07:50:42 -!- tombom has quit (Quit: Leaving). 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:09:01 -!- FireFly has quit (Quit: swatted to death). 08:25:23 -!- augur has quit (Ping timeout: 265 seconds). 08:35:37 -!- augur has joined. 09:17:31 -!- wareya_ has quit (Ping timeout: 240 seconds). 09:19:42 -!- wareya has joined. 09:23:17 -!- catseye has quit (Ping timeout: 276 seconds). 09:50:22 -!- ais523 has joined. 10:00:35 wtf is going on on http://esolangs.org/wiki/BytePusher? 10:00:46 occasionally you get a page submitted to Wikipedia which is more like an advert or press release 10:00:50 I think that's happened there, but ontopic 10:01:07 (I'm inclined to just leave it, it's fun to have the page around and it is actually useful) 10:05:47 heh, it does kinda look like that 10:06:12 btw it's quite easy to make a compliant VM for 10:06:19 like, *scarily* easy 10:09:43 also yeah it needs a subleq OISC VM 10:10:36 seems doable, have a sub table and a carry table 10:11:03 the carry table having $00 on no carry and $FF on carry 10:11:07 that's two 64KB tables 10:22:08 -!- augur has quit (Remote host closed the connection). 10:27:36 -!- GreaseMonkey has quit (Quit: Welcome honored guest. I got the key you want! would you like onderves. of Yourself). 11:23:51 hmm, amusing changelog entry (Ubuntu bumping the version of Wine): "Many more applications work, especially those from companies named after particularly cold weather events" 11:39:31 -!- wareya has quit (Ping timeout: 240 seconds). 11:41:42 -!- wareya has joined. 11:42:01 -!- augur has joined. 12:21:42 -!- augur has quit (Ping timeout: 265 seconds). 12:24:01 -!- atrapado has joined. 12:29:14 ais523: ? 12:29:19 ais523: what does that mean? 12:53:23 -!- augur has joined. 12:55:23 -!- augur has set topic: RIP Benoît Mandelbrot | (logs: http://is.gd/g4uID). 12:58:52 -!- oerjan has joined. 12:59:51