| Wednesday, December 7th, 2011, 00:09 UTC | ||
| [00:09:30] | sphery: | kenni: you around? Was wondering if you want http://code.mythtv.org/trac/ticket/9227 back? I'm going to put a note on it that says that the changes to add subtitle aren't usable, which leaves only the word changes (and I don't know enough about i18n/contexts to do that right). |
| [00:13:35] | reynaldo (reynaldo!~rverdejo@cpc3-cmbg14-2-0-cust186.5-4.cable.virginmedia.com) has joined #mythtv | |
| [00:25:28] | mmiller (mmiller!VAXVMS@pdpc/supporter/professional/mmiller) has quit (Remote host closed the connection) | |
| [00:37:38] | reynaldo (reynaldo!~rverdejo@cpc3-cmbg14-2-0-cust186.5-4.cable.virginmedia.com) has quit (Ping timeout: 240 seconds) | |
| [01:00:34] | stichnot (stichnot!~chatzilla@192.55.54.40) has quit (Quit: ChatZilla 0.9.87 [Firefox 8.0.1/20111120135848]) | |
| [01:11:51] | saintdev (saintdev!~saint@c-24-9-201-236.hsd1.co.comcast.net) has joined #mythtv | |
| [01:11:51] | saintdev (saintdev!~saint@c-24-9-201-236.hsd1.co.comcast.net) has quit (Changing host) | |
| [01:11:51] | saintdev (saintdev!~saint@unaffiliated/saintdev) has joined #mythtv | |
| [01:15:07] | gigem (gigem!~gigem@mythtv/developer/gigem) has quit () | |
| [01:15:45] | jpabq (jpabq!~jpabq@mythtv/developer/jpabq) has quit (Quit: jpabq) | |
| [01:18:30] | gigem (gigem!~gigem@mythtv/developer/gigem) has joined #mythtv | |
| [01:43:15] | rhpot1991 (rhpot1991!~rhpot1991@ubuntu/member/rhpot1991) has quit (Ping timeout: 268 seconds) | |
| [01:46:16] | rhpot1991 (rhpot1991!~rhpot1991@2001:4968:202:3:20f:eaff:fefc:ba0e) has joined #mythtv | |
| [01:46:16] | rhpot1991 (rhpot1991!~rhpot1991@2001:4968:202:3:20f:eaff:fefc:ba0e) has quit (Changing host) | |
| [01:46:16] | rhpot1991 (rhpot1991!~rhpot1991@ubuntu/member/rhpot1991) has joined #mythtv | |
| [01:54:36] | ybot (ybot!quassel@nat/nokia/x-avwfzcfvismlckpj) has joined #mythtv | |
| [02:05:53] | davide_ (davide_!~david@mythtv/developer/gigem) has quit (Remote host closed the connection) | |
| [02:06:17] | davide_ (davide_!~david@mythtv/developer/gigem) has joined #mythtv | |
| [02:15:08] | abqjp: | sphery, just updated my machine again. Started mythbackend with "/usr/local/bin/mythbackend --daemon --logfile /var/log/mythtv/" under Fedora 13, and still have the issue where it tries to ask permission to update the DB, even though it is running in the background. |
| [02:15:48] | abqjp: | I have NOT continued with the update, if there is anything you want me to try. |
| [02:28:14] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [02:29:24] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [02:34:46] | sphery: | abqjp: can't think of anything to try right now, but it should be easy to repro on a system that exhibits the problem |
| [02:34:53] | sphery: | so feel free to let it upgrade |
| [02:35:26] | NightMonkey (NightMonkey!~NightMonk@pdpc/supporter/professional/nightmonkey) has quit (Quit: Body blow! Body blow!) | |
| [02:36:17] | abqjp: | OK |
| [02:44:04] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [02:52:32] | toeb (toeb!~tob@HSI-KBW-078-042-104-026.hsi3.kabel-badenwuerttemberg.de) has quit (Ping timeout: 252 seconds) | |
| [02:54:13] | toeb (toeb!~tob@HSI-KBW-078-042-104-026.hsi3.kabel-badenwuerttemberg.de) has joined #mythtv | |
| [02:54:40] | stichnot (stichnot!~chatzilla@nat/intel/x-glbcngtamxjjxuce) has joined #mythtv | |
| [03:29:29] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [04:02:54] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [04:38:55] | kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has quit (Quit: kormoc) | |
| [04:57:22] | Freemason1977 (Freemason1977!4a613cd2@gateway/web/freenode/ip.74.97.60.210) has joined #mythtv | |
| [05:01:37] | Freemason1977: | dont suppose someone could help me setup a card, this isn't exactly my area of expertise :-/ |
| [05:02:03] | Freemason1977: | mtv 11.10 has got a heartbeat and fully updated |
| [05:02:29] | Freemason1977: | working with a hvr1500 |
| [05:03:24] | wagnerrp: | this is the development channel, perhaps you want #mythtv-users |
| [05:50:13] | knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Ping timeout: 244 seconds) | |
| [05:50:17] | knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv | |
| [05:54:39] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has joined #mythtv | |
| [06:12:47] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has quit (Quit: Leaving.) | |
| [06:17:37] | saintdev (saintdev!~saint@unaffiliated/saintdev) has quit (Remote host closed the connection) | |
| [06:41:51] | skd5aner (skd5aner!~skd5aner@cpe-071-071-242-134.carolina.res.rr.com) has quit (Ping timeout: 248 seconds) | |
| [06:42:21] | skd5aner (skd5aner!~skd5aner@cpe-071-071-242-134.carolina.res.rr.com) has joined #mythtv | |
| [06:47:33] | Freemason1977 (Freemason1977!4a613cd2@gateway/web/freenode/ip.74.97.60.210) has quit (Quit: Page closed) | |
| [07:03:13] | Freemason1977 (Freemason1977!4a613cd2@gateway/web/freenode/ip.74.97.60.210) has joined #mythtv | |
| [07:09:51] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has joined #mythtv | |
| [07:17:51] | jeff999 (jeff999!~jeff@58-6-93-73.dyn.iinet.net.au) has joined #mythtv | |
| [07:26:07] | rhpot1991 (rhpot1991!~rhpot1991@ubuntu/member/rhpot1991) has quit (Ping timeout: 268 seconds) | |
| [07:33:16] | rhpot1991 (rhpot1991!~rhpot1991@204.8.45.14) has joined #mythtv | |
| [07:33:16] | rhpot1991 (rhpot1991!~rhpot1991@204.8.45.14) has quit (Changing host) | |
| [07:33:16] | rhpot1991 (rhpot1991!~rhpot1991@ubuntu/member/rhpot1991) has joined #mythtv | |
| [07:56:05] | saintdev (saintdev!~saint@unaffiliated/saintdev) has joined #mythtv | |
| [08:10:03] | willcooke (willcooke!~will@canonical/willcooke) has joined #mythtv | |
| [08:10:40] | markk: | stichnot: all of the current OSD painters use QImage – so there should be no issue rendering in another thread (assuming an appropriate level of lock protection). The big issue is coordinating that rendering between the threads. |
| [08:11:41] | markk: | I'm also guessing that the correct place to handle it is in libmythui itself – i.e. implement some sort of background text rendering in the same way background image loading is handled. |
| [08:15:03] | simonckenyon (simonckenyon!~simoncken@195.7.61.12) has joined #mythtv | |
| [08:17:58] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has quit (Ping timeout: 248 seconds) | |
| [09:00:08] | gregL (gregL!~greg@cpe-74-76-125-87.nycap.res.rr.com) has quit (Read error: Connection reset by peer) | |
| [09:02:00] | TandyUK2 (TandyUK2!~me@2001:470:68c7:1:e4e0:9e82:91f5:7c9c) has joined #mythtv | |
| [09:04:39] | TandyUK (TandyUK!~me@2001:470:68c7:1:e4b1:9ecd:98ef:5947) has quit (Ping timeout: 252 seconds) | |
| [09:17:22] | gregL (gregL!~greg@cpe-74-76-125-87.nycap.res.rr.com) has joined #mythtv | |
| [10:15:13] | Beirdo: | markk: you just made my life OH so much more fun :) |
| [10:15:31] | markk: | in a good way? |
| [10:15:35] | Beirdo: | heh. Merging that should prove interesting |
| [10:15:50] | ** stuarta chuckles ** | |
| [10:16:12] | Beirdo: | well, I already worked around some of the issues in getting external decoding in for GPU commflagging, but hopefully your incarnation's better :) |
| [10:16:13] | markk: | ah – ok:( You may want to hold off trying to merge – I have a few more changes on the back of that |
| [10:16:18] | Beirdo: | hehe |
| [10:16:57] | Beirdo: | no prob, I'm not quite ready to marge current master into my branch anyways. I'm maybe about 1/2 to 2/3 done this, no biggie anyways :) |
| [10:17:30] | Beirdo: | that merge will be fun, so I might as well wait until all the fun is ready :) |
| [10:18:06] | reynaldo (reynaldo!~rverdejo@bhaji-mythic.collabora.co.uk) has joined #mythtv | |
| [10:18:46] | Beirdo: | if somehow, VideoOutNull could not use VideoOutWindow (or at least be able to override the use of Xinerama (sp?) that would be cool, but if not, we'll live |
| [10:19:31] | Beirdo: | as I'm using VDPAU and OpenGL to get the decoded video into OpenCL, I have to have X running even though I never bring up a window :) |
| [10:20:21] | Beirdo: | not to worry, I'm sure your changes will make life simpler anyways. This is the risk of working in parallel, which I'm happy to deal with |
| [10:21:12] | Beirdo: | BTW, OpenCL is happy on both NVIDIA and AMD GPUs, which makes me happy |
| [10:23:16] | Beirdo: | tis all good, keep up the good work :) Ignore my muttering |
| [10:34:52] | stuartm: | these trac spammers are really persistent |
| [10:39:54] | Beirdo: | yeah, and annoying! |
| [10:44:13] | markk: | Beirdo: there's no way to avoid having X running – both vdpau and opengl need X. what's the problem with VideoOutWindow? |
| [10:45:11] | Beirdo: | Yeah, I'm sometimes getting errors from the startup during xinerama checking saying it can't open the display |
| [10:45:21] | Beirdo: | even though glut and vdpau both have already |
| [10:45:39] | markk: | btw – are you touching the video output classes or decoder setup? |
| [10:45:41] | Beirdo: | and... it's just odd that a *null* output even checks if X is there :) |
| [10:46:17] | Beirdo: | no, I have a custom player based on mythplayer, and a custom decoder based on avformatdecoder |
| [10:46:49] | Beirdo: | as I'm queuing the packets for decoding in separate threads so I can process audio in one GPU and video in another |
| [10:47:08] | Beirdo: | I believe I've made a total of 5 line changes so far in the current code |
| [10:47:30] | Beirdo: | make that 8 |
| [10:47:47] | markk: | Beirdo: sounds odd re checking X. If that was a problem – it would already be happening with existing code?? |
| [10:48:16] | Beirdo: | well, we only access X from one thread currently, correct? |
| [10:48:40] | Beirdo: | the VideoOutWindow is in the CoreContext thread, and my decoder is in a separate thread |
| [10:48:47] | Beirdo: | that might be somehow related |
| [10:49:21] | Beirdo: | so with two threads trying to setup X display stuff, maybe Xlib is getting bent? I dunno |
| [10:50:29] | markk: | are you actually using 2 GPUs? |
| [10:50:34] | Beirdo: | just found it odd anyways. I looked at how to decouple it further for maybe 20 minutes and decided it was way too invasive |
| [10:50:37] | Beirdo: | yup |
| [10:51:07] | Beirdo: | my test platform has a 550Ti (which I'm using for video) and an ION2 (which I'm processing audio in) |
| [10:51:43] | Beirdo: | and I have an AMD Zecate board (E-350) and OpenCL sees two devices, so teh GPU does video and the CPU does audio |
| [10:52:09] | Beirdo: | I have yet to put in VAAPI decoding, although, potentially that would work on the AMD |
| [10:52:30] | Beirdo: | the OpenCL kernels need a fair amount of... optimization... :) |
| [10:53:16] | Beirdo: | but at least the CUDA profiler will profile OpenCL use too, so I have some tools to work with |
| [10:53:46] | Beirdo: | I considered just using CUDA, but we have too many stubborn users who insist on buying AMD |
| [10:53:52] | Beirdo: | sigh |
| [10:54:33] | Beirdo: | and theoretically, Intel should be putting out OpenCL for the next chip after the SandyBridge if not the SB itself |
| [10:54:40] | Beirdo: | seemed like the right path |
| [10:57:03] | Beirdo: | and minimizing the data flow between GPU and CPU makes it so much nicer, so I'm trying to do as much as possible on the GPU... which is so bitchy if you get your buffer management wrong or write off either end of a buffer |
| [10:58:18] | Beirdo: | anyways, so far, I have blank frame detection in, scene change detection (via the use of color histograms), and aspect ratio change detection (on the video side) |
| [10:58:49] | Beirdo: | logo detection is maybe 1/2 done (and is super tricky to figure out how to do in one pass!) |
| [10:59:34] | Beirdo: | on the audio side, I look for 6dB above average (for compressed audio in commercials) and 12dB or more below average (for silence) |
| [11:00:07] | Beirdo: | The correlation of the events is soon to be worked on. And is scary |
| [11:00:09] | Beirdo: | :) |
| [11:00:39] | Beirdo: | also want to add on the audio side detecting a change in the number of audio channels |
| [11:01:00] | Beirdo: | as there are some places apparently where the commercials dip back to stereo in a 5.1 program |
| [11:01:27] | markk: | I can't honestly see how you'd get logo detection in one pass – but then I don't know how it works at the moment |
| [11:02:35] | Beirdo: | yeah, it's based on edge detection, so my basic plan is... after a blank frame or massive scene change, train over a number of frames with accumulating the edges in the corner regions |
| [11:02:43] | Beirdo: | then match that to each frame |
| [11:02:53] | Beirdo: | I'm hoping I can get it to behave :) |
| [11:03:16] | Beirdo: | you'd just have a delay of X frames after the transition. |
| [11:03:51] | Beirdo: | might not work well, and if not, I'll go back to the drawing board, but I really don't want to decode twice if I can help it |
| [11:04:24] | Beirdo: | and I've read far too many research papers lately on "novel" commercial detection algorithms |
| [11:04:27] | Beirdo: | heh |
| [11:04:50] | Beirdo: | math gets to hurt the brain after a while |
| [11:05:03] | 18WAACYDE (18WAACYDE!~mike@c-76-115-119-121.hsd1.or.comcast.net) has quit (Remote host closed the connection) | |
| [11:05:51] | Beirdo: | anyways, I think I should go to bed. 3am... why do I do this to myself :) |
| [11:06:00] | mike (mike!~mike@c-76-115-119-121.hsd1.or.comcast.net) has joined #mythtv | |
| [11:06:26] | mike is now known as Guest9718 | |
| [11:06:30] | Beirdo: | night. :) Keep up the fine work, don't worry about my merge, I'll get to it eventually when you've finished your round of fun changes |
| [11:06:53] | Beirdo: | and this will sit for quite a while before ever getting merged back to master, I'm sure |
| [11:06:56] | Beirdo: | I'm in no rush |
| [11:07:27] | Beirdo: | we'll have lots of time to smooth out any wrinkles that come up and so on |
| [11:41:04] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [11:54:16] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [13:05:53] | davide_ (davide_!~david@mythtv/developer/gigem) has quit (Remote host closed the connection) | |
| [13:06:14] | davide_ (davide_!~david@host70.16.intrusion.com) has joined #mythtv | |
| [13:06:15] | davide_ (davide_!~david@host70.16.intrusion.com) has quit (Changing host) | |
| [13:06:15] | davide_ (davide_!~david@mythtv/developer/gigem) has joined #mythtv | |
| [13:18:06] | jmartens (jmartens!~jmartens@109.232.42.33) has joined #mythtv | |
| [13:28:53] | ElmerFudd (ElmerFudd!~le@0x5737a682.cpe.ge-0-1-0-1101.hsnqu1.customer.tele.dk) has quit (Ping timeout: 252 seconds) | |
| [13:32:51] | ElmerFudd (ElmerFudd!~le@0x5737a682.cpe.ge-0-1-0-1101.hsnqu1.customer.tele.dk) has joined #mythtv | |
| [13:57:01] | j-rod|afk is now known as j-rod | |
| [13:59:31] | j-rod is now known as j-rod|afk | |
| [13:59:33] | j-rod|afk is now known as j-rod | |
| [13:59:38] | TandyUK (TandyUK!~me@2001:470:68c7:1:e4e0:9e82:91f5:7c9c) has joined #mythtv | |
| [14:02:23] | TandyUK2 (TandyUK2!~me@2001:470:68c7:1:e4e0:9e82:91f5:7c9c) has quit (Ping timeout: 252 seconds) | |
| [14:20:05] | MythBuild: | build #356 of master-vista-mingw-32bit is complete: Exception [exception git] Build details are at http://code.mythtv.org/buildbot/builders/mast . . . t/builds/356 blamelist: Mark Kendall <mkendall@mythtv.org > |
| [14:20:49] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [14:29:43] | iamlindoro: | markk: Very cool changes! Is commflag/transcode decode via VDPAU in the cards, or does it not exactly lend itself to that kind of thing? |
| [14:30:01] | iamlindoro: | (I have CHD and VDA capable hardware, but harder to get up and running as a BE) |
| [14:40:23] | markk: | iamlindoro: VDPAU, VAAPI and DXVA2 all require more extensive changes. e.g. vdpau requires a MythRenderVDPAU object that is currently tied into/assumes an onscreen window (and requires a running X server) |
| [14:41:27] | markk: | iamlindoro, Captain_Murdoch: does live streaming currently run from the master backend? or whichever backend hosts the file? |
| [14:42:06] | iamlindoro: | It runs from whichever backend Services API you contact, and streams the content from remote backends as necessary |
| [14:42:19] | iamlindoro: | The best choice is generally the MBE, though, for a number of reasons |
| [14:42:55] | Captain_Murdoch: | markk, I'd like to get it integrated with the JobQueue at some point if we can have 'run now' jobs. |
| [14:42:57] | markk: | there is also the possibility with VDPAU etc to do more than decode – e.g. deinterlace and scale. |
| [14:43:15] | ** iamlindoro heads to work ** | |
| [14:50:08] | Captain_Murdoch: | markk, that might be nice. I did modify mythtranscode to use a decode thread recently but scaling still occurs in the main thread right before the encode. libx264 HLS streams are encoded in separate threads since it's configured to use slices. it probably makes sense to have a mythtranscodeplayer.cpp at some point. |
| [14:54:32] | justinh (justinh!~justin@cpc10-salf5-2-0-cust50.10-2.cable.virginmedia.com) has quit (Quit: leaving) | |
| [14:56:31] | danielk22: | Beirdo: The Null video decoder shouldn't need X11. The only thing I can think of is that there is some query for display size. Also reading back from the framebuffer isn't as painful with pci-e as it was with AGP. Not as fast as on an SGI, but pretty fast. |
| [14:58:54] | danielk22: | Beirdo: For the audio are you using 1770 ? It's not too complex to implement and much more accurate. I'd love to have a loudness measure for each program so we could adjust volume at playback start so that programs don't have drastically different volumes depending on source... |
| [15:11:17] | justinh (justinh!~justin@cpc10-salf5-2-0-cust50.10-2.cable.virginmedia.com) has joined #mythtv | |
| [15:18:35] | ** stuarta waves to justinh ** | |
| [15:19:46] | abqjp: | Beirdo, Just as a data point ---- National Geographic channel often does not show a logo until well after the return from a commercial. With the current code, I often have to rewind as much as a couple of minutes after each commercial. NatGeo does consistently show the "rating" coming out of each commercial (TV-PG), but the current code does not look for that. |
| [15:34:48] | Freemason1977 (Freemason1977!4a613cd2@gateway/web/freenode/ip.74.97.60.210) has quit (Ping timeout: 265 seconds) | |
| [15:38:12] | zombor_ (zombor_!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [15:41:15] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Ping timeout: 252 seconds) | |
| [15:42:35] | Freemason1977 (Freemason1977!4a613cd2@gateway/web/freenode/ip.74.97.60.210) has joined #mythtv | |
| [15:43:12] | zombor_ is now known as zombor | |
| [15:47:14] | jmartens (jmartens!~jmartens@109.232.42.33) has quit (Quit: Leaving.) | |
| [15:48:35] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Read error: Connection reset by peer) | |
| [15:49:02] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [15:50:07] | joemyth (joemyth!~joe.peled@63.117.104.178) has joined #mythtv | |
| [15:56:40] | Freemason1977 (Freemason1977!4a613cd2@gateway/web/freenode/ip.74.97.60.210) has left #mythtv () | |
| [15:57:53] | Mkaysi (Mkaysi!Mkaysi@TouchLay/Member/Mkaysi) has quit (Ping timeout: 252 seconds) | |
| [16:01:26] | Mkaysi (Mkaysi!Mkaysi@TouchLay/Member/Mkaysi) has joined #mythtv | |
| [16:07:08] | stichnot (stichnot!~chatzilla@nat/intel/x-glbcngtamxjjxuce) has quit (Ping timeout: 240 seconds) | |
| [16:09:59] | Chutt (Chutt!~ijr@cpe-24-29-225-175.neo.res.rr.com) has quit (Read error: Connection reset by peer) | |
| [16:10:47] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has joined #mythtv | |
| [16:42:17] | joemyth (joemyth!~joe.peled@63.117.104.178) has quit (Read error: Connection reset by peer) | |
| [16:46:15] | freemason1977 (freemason1977!4a613cd2@gateway/web/freenode/ip.74.97.60.210) has joined #mythtv | |
| [17:01:49] | kenni: | sphery: Sure, or if the word changes for improved consistency looks ok, just accept them as they are (with no change in context). I don't feel strong about contexts currently, as the context isn't defined for most of the strings in the current codebase anyway. |
| [17:03:49] | kenni: | sphery: if we're going to fix the translation context, it's probably better to do it for all of the codebase at once instead of fixing 3 strings out of 4000. |
| [17:04:32] | sphery: | ok, thanks for the feedback... |
| [17:05:23] | sphery: | kenni: also, was wondering if yours was the same issue as http://code.mythtv.org/trac/ticket/8864 (i.e. were you able to reproduce it)? If so, any chance you could run a test with current -fixes or master? |
| [17:13:22] | kenni: | sphery: no, the issue I had was fixed when janne reverted a commit and closed http://code.mythtv.org/trac/ticket/8927 – The issue I had in #8927 was that a recording was unplayable by MythTV, but playable in other players. According to the OP in #8864 his recording was unplayable by other players |
| [17:13:50] | danielk22: | kenni: as a matter of my own use of contexts, I really only use them when a string is in QObject and hence doesn't have an implicit context & it is something like "A" which is meaningless without a context. What is the value of contexts in other situations? |
| [17:16:26] | sphery: | kenni: ah, ok, thanks... didn't realize yours wasn't actually corrupt nuv |
| [17:19:27] | willcooke (willcooke!~will@canonical/willcooke) has quit (Ping timeout: 240 seconds) | |
| [17:26:09] | NightMonkey (NightMonkey!~NightMonk@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [17:27:32] | kenni: | danielk22: When translators do their translations, strings within the same context are presented visually in the same "group". That makes it easier for the translator to use the same wording for the translations. If everything is within the generic QObject context, there'll be a race between the translation files for identical strings and a translation of a string in mythmusic might end up being used for the same string in mythfrontend. |
| [17:27:47] | stichnot (stichnot!~chatzilla@192.55.55.37) has joined #mythtv | |
| [17:29:10] | danielk22: | kenni: So if "Open File" is in multiple places you want to be able translate it separately for each 'context' ? |
| [17:30:47] | kenni: | danielk22: yep, I can't think of any good examples right now, but I have come across them before |
| [17:32:01] | willcooke (willcooke!~will@canonical/willcooke) has joined #mythtv | |
| [17:33:42] | kenni: | danielk22: a slightly better example could perhaps be "item" – if a direct translation doesn't exist (or isn't widely used), the translator might be tempted to translate "Play item" into "Play track", "Play video", etc – which will result in funny results when a race occurs |
| [17:35:31] | danielk22: | kenni: Yeah, I'm convinced that a context will be necessary in cases like that. Not yet convinced we need to add context for all the strings. |
| [17:35:46] | kenni: | danielk22: I agree |
| [17:37:09] | danielk22: | got it, thanks for the explanation |
| [17:47:55] | Chutt (Chutt!~ijr@cpe-24-29-225-175.neo.res.rr.com) has joined #mythtv | |
| [17:48:02] | Beirdo: | iamlindoro: yes, commflag via VDPAU is in the works. I've been working on it for more than a month now. |
| [17:53:19] | iamlindoro: | Beirdo: I actually was referring to the generalized decode he put in, which would help with live streaming transcode |
| [17:53:43] | iamlindoro: | ie, making it available for all purposes rather than just commflag |
| [18:01:08] | stichnot_ (stichnot_!~chatzilla@192.55.54.38) has joined #mythtv | |
| [18:02:20] | stichnot (stichnot!~chatzilla@192.55.55.37) has quit (Ping timeout: 244 seconds) | |
| [18:02:31] | stichnot_ is now known as stichnot | |
| [18:06:03] | davide_ (davide_!~david@mythtv/developer/gigem) has quit (Remote host closed the connection) | |
| [18:06:07] | freemason1977 (freemason1977!4a613cd2@gateway/web/freenode/ip.74.97.60.210) has quit (Quit: Page closed) | |
| [18:06:24] | davide_ (davide_!~david@host70.16.intrusion.com) has joined #mythtv | |
| [18:06:24] | davide_ (davide_!~david@host70.16.intrusion.com) has quit (Changing host) | |
| [18:06:24] | davide_ (davide_!~david@mythtv/developer/gigem) has joined #mythtv | |
| [18:08:28] | Beirdo^2 (Beirdo^2!~Beirdo@mythtv/developer/beirdo) has joined #mythtv | |
| [18:18:38] | Beirdo^2 (Beirdo^2!~Beirdo@mythtv/developer/beirdo) has quit (Quit: Bye) | |
| [18:21:19] | kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has joined #mythtv | |
| [18:26:10] | JoeyJoeJo (JoeyJoeJo!~brian@pool-173-71-223-79.clppva.fios.verizon.net) has quit (Quit: Leaving) | |
| [18:29:40] | kenni: | danielk22: Thinking about it, the main issue with missing contexts is that it makes it much harder for the translators to make good translations. The logical way to make good consistent translations would be to take one logical section in MythTV at a time (for example the Live TV OSD menu or the recordings screen) and check it frequently while translating. Currently if you want to do that, you'll have to switch to the screen on your MythTV installation |
| [18:30:02] | kenni: | When I did the full Danish translation for 0.22/0.23 with a friend, we took this approach and today we're STILL struggling with weird translations (usually on screens which aren't used frequently), which could be a correct translation in one context, but isn't in the current context. If the strings were grouped under individual contexts which corresponded to the screens/sections in MythTV, we could probably have nailed many of the bad translations whil |
| [18:30:52] | kenni: | anyway, need to get some lunch :) |
| [18:33:47] | knightr: | danielk22, When possible use of QObject should be avoided, as Kenni mentionned the using QObject make it harder to make context-sensitive translation and it also makes it difficult to know what the translation is for.. |
| [18:34:16] | knightr: | We have way too much stuff that was sent to QObject which should have been put in its own context... |
| [18:36:23] | knightr: | Anyway when we run lupdate if a translation for a string that was already present in another context is found it is offered (actually recopied but not marked as complete) which we can replace with another or accept (mark as complete) |
| [18:40:36] | Mousey (Mousey!~wtfisme@ross154.net) has joined #mythtv | |
| [18:45:13] | Beirdo: | danielk22: the NULL video out right now includes VideoOutWindow class via VideoOutBase, and it's a static member of that class, so the constructor always fires up. In VideoOutWindow's constructor, it checks for X usage regardless of what top-level class included it. |
| [18:45:50] | andreax (andreax!~andreaz@p54BF2C02.dip.t-dialin.net) has joined #mythtv | |
| [18:46:07] | Beirdo: | as for audio, I'm not aware of "1770", you'd have to give more details. Currently, I can give you the RMS average level of the entire file, and also the DC offset (if any) |
| [18:46:51] | Beirdo: | the problem with tweaking volume level based on that is that RMS average does not cover the peaks, which are likely to get clipped during above-average periods (or even in average) |
| [18:47:14] | Beirdo: | abqjp: yeah, that's part of the fun of logo detection. sigh :) |
| [18:47:48] | Beirdo: | danielk22: you mean ITR-R BS.1770? |
| [18:48:38] | Beirdo: | wow, they actually allow downloading that one for free |
| [18:49:38] | Beirdo: | looking briefly at it, the measurement is similar, but no prefiltering |
| [18:50:38] | Beirdo: | hmmm, and I'm working with 32ms windows, not 400ms overlapping ones |
| [18:50:47] | Beirdo: | it's something that can be considered though |
| [18:55:29] | knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Ping timeout: 240 seconds) | |
| [18:55:46] | knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv | |
| [18:58:09] | smoothifier (smoothifier!smoothifie@gateway/shell/anapnea.net/x-fihzdbslcamlubcm) has joined #mythtv | |
| [18:58:22] | smoothifier (smoothifier!smoothifie@gateway/shell/anapnea.net/x-fihzdbslcamlubcm) has left #mythtv () | |
| [19:07:44] | rsiebert_ (rsiebert_!~quassel@g224248007.adsl.alicedsl.de) has joined #mythtv | |
| [19:08:07] | rsiebert (rsiebert!~quassel@g231187089.adsl.alicedsl.de) has quit (Ping timeout: 240 seconds) | |
| [19:10:13] | stichnot: | markk, danielk22: on further testing, it looks like the subtitle-related stuttering I mentioned is coming from font initialization/caching, since it doesn't happen if I skip back and play the subtitle again. If this becomes a problem, the simpler solution is to just preload the common font/attribute combinations, rather than a complex multithreaded subtitle mechanism. |
| [19:10:39] | stichnot: | in other words, I won't be working on the multithreading patch. :) |
| [19:14:32] | jpabq (jpabq!~jpabq@mythtv/developer/jpabq) has joined #mythtv | |
| [19:18:32] | knightr: | danielk22, I had read what you wrote a little fast... Use of context a context different from QObject is not necessary for all strings but there are a lot of strings where one should have been used... |
| [19:21:14] | sphery: | stichnot: wasn't http://code.mythtv.org/trac/ticket/7913 (admittedly, on the pre-mythui, ttfont-using OSD implementation) basically saying the opposite (that preloading fonts resulted in a playback startup delay, especially on slower systems, like your Atom) |
| [19:22:11] | sphery: | or at least saying that preloading puts several little delays on subtitle creation together to create a long playback startup delay |
| [19:30:41] | stichnot: | sphery: here I'm talking about the possibility of preloading just one or two fonts, not 48 fonts like in ticket 7913. |
| [19:31:55] | stichnot: | but I don't think it will be needed because in practice, I only ever see italics and a very occasional color change. |
| [19:33:41] | stichnot: | also, one could hide latency by preloading one font every few frames after playback starts |
| [19:34:30] | stichnot: | again, realistically, I don't expect to hear any complaints since I only saw problems during torture testing. |
| [19:39:47] | sphery: | ok, cool... was just hoping you weren't suggesting a return to the old way of preloading all fonts for osd--and I agree that most probably won't really notice issues |
| [20:06:08] | willcooke (willcooke!~will@canonical/willcooke) has quit (Quit: We will learn more of his wisdom later) | |
| [20:26:53] | TandyUK: | video playlists... surely i cant be the only one who is interested |
| [20:27:12] | TandyUK: | the current play everything / play a folder is woefully inadequate |
| [20:27:24] | TandyUK: | note videos, not recordings |
| [20:38:01] | markk: | stichnot: have you actually profiled the code or similar to see what is taking the bulk of the time? my gut reaction to the skip back would be that the text has already been rendered – hence the actual text image is already in the cache (i.e. I wouldn't assume it's the fonts) |
| [20:38:54] | andreax (andreax!~andreaz@p54BF2C02.dip.t-dialin.net) has quit (Read error: Connection reset by peer) | |
| [21:08:08] | iamlindoro: | TandyUK: Surely you're not-- feel free to implement a universal playlist functionality |
| [21:08:37] | Jordack (Jordack!~jordack@h69-131-44-221.plmomi.dedicated.static.tds.net) has joined #mythtv | |
| [21:18:35] | andreax (andreax!~andreaz@p54BF2C02.dip.t-dialin.net) has joined #mythtv | |
| [21:23:46] | TandyUK: | after xmas, once i have a tv, and tv license, i will be buildingmyself a myth dev box |
| [21:24:09] | TandyUK: | until then im not much help lol |
| [21:40:21] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has joined #mythtv | |
| [21:41:09] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Ping timeout: 248 seconds) | |
| [21:41:21] | jya (jya!~jyavenard@mythtv/developer/jya) has quit (Quit: jya) | |
| [21:58:03] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has quit (Quit: Leaving.) | |
| [22:18:38] | cattelan (cattelan!~cattelan@c-66-41-26-220.hsd1.mn.comcast.net) has quit (Read error: Connection reset by peer) | |
| [22:19:41] | j-rod is now known as j-rod|afk | |
| [22:22:52] | cattelan (cattelan!~cattelan@c-66-41-26-220.hsd1.mn.comcast.net) has joined #mythtv | |
| [22:24:25] | cattelan is now known as cattelan_away | |
| [22:24:56] | cattelan_away is now known as cattelan_away_aw | |
| [22:38:17] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has quit (Ping timeout: 252 seconds) | |
| [22:49:51] | stichnot: | markk: I didn't do profiling, I was just guessing the cause. I could test with two separate subtitles with the same formatting but different text. That should indicate where the caching is happening. |
| [22:51:17] | stichnot: | Question about http live streaming. Does it (or will it) support cut lists and captions/subtitles? |
| [22:52:11] | wagnerrp: | if it doesnt support them yet, cut lists would be fairly simple |
| [22:52:26] | wagnerrp: | captions would be more difficult |
| [22:54:32] | sphery: | pretty sure it supports cut lists, already |
| [23:01:19] | stichnot: | what is the issue with captions? I assumed the system would decode video frame by frame into memory and then reencode into x264, so if the frame just happened to contain OSD stuff, then it would also be encoded. |
| [23:03:26] | cattelan_away_aw is now known as cattelan_away | |
| [23:11:10] | wagnerrp: | since you are transcoding, you could cook the rendered subtitles directly into the video |
| [23:11:38] | wagnerrp: | but softsubs would mean reformulating the existing streams, which would be more or less hard depending on the type |
| [23:17:22] | stichnot: | yes, I assume you would be burning the subtitles into the stream. if you don't want subtitles, you just rejigger your backend/playlist/whatever and get it to transcode without subtitles. |
| [23:48:06] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [23:50:41] | jpabq (jpabq!~jpabq@mythtv/developer/jpabq) has quit (Quit: jpabq) | |
IRC Logs collected by
BeirdoBot.
Please use the above link to report any bugs.