| Thursday, May 17th, 2012, 00:03 UTC | ||
| [00:03:14] | sphery: | UForgotten: /topic (you want #mythtv-users ) |
| [00:05:10] | UForgotten: | meh |
| [00:05:12] | UForgotten (UForgotten!~u4go10@unaffiliated/uforgotten) has left #mythtv () | |
| [00:05:37] | SmallR2002: | so |
| [00:07:33] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [00:17:30] | dmfrey (dmfrey!~dmfrey@64-121-93-243.c3-0.smt-ubr1.atw-smt.pa.cable.rcn.com) has joined #mythtv | |
| [00:19:23] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [00:32:59] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [00:35:54] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Ping timeout: 244 seconds) | |
| [00:50:51] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [01:21:50] | kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has quit (Quit: kormoc) | |
| [01:26:50] | obo (obo!~obo@70-140-190-120.lightspeed.hstntx.sbcglobal.net) has joined #mythtv | |
| [01:26:51] | obo (obo!~obo@70-140-190-120.lightspeed.hstntx.sbcglobal.net) has quit (Changing host) | |
| [01:26:51] | obo (obo!~obo@rockbox/developer/obo) has joined #mythtv | |
| [01:49:59] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Remote host closed the connection) | |
| [01:54:28] | ThisNewGuy (ThisNewGuy!~doug@pool-74-102-15-78.nwrknj.fios.verizon.net) has quit (Read error: Connection reset by peer) | |
| [02:05:57] | ThisNewGuy (ThisNewGuy!~doug@pool-74-102-15-78.nwrknj.fios.verizon.net) has joined #mythtv | |
| [02:18:53] | danielk22 (danielk22!~danielk@96.57.9.142) has joined #mythtv | |
| [02:18:55] | dmfrey (dmfrey!~dmfrey@64-121-93-243.c3-0.smt-ubr1.atw-smt.pa.cable.rcn.com) has quit (Quit: Ex-Chat) | |
| [02:21:20] | danielk22 (danielk22!~danielk@96.57.9.142) has quit (Remote host closed the connection) | |
| [02:28:46] | ben1066 (ben1066!~quassel@unaffiliated/ben1066) has quit (Ping timeout: 272 seconds) | |
| [02:29:31] | ben1066 (ben1066!~quassel@host86-160-28-148.range86-160.btcentralplus.com) has joined #mythtv | |
| [02:29:31] | ben1066 (ben1066!~quassel@host86-160-28-148.range86-160.btcentralplus.com) has quit (Changing host) | |
| [02:29:31] | ben1066 (ben1066!~quassel@unaffiliated/ben1066) has joined #mythtv | |
| [03:03:48] | jya_ (jya_!~jyavenard@mythtv/developer/jya) has quit (Quit: jya_) | |
| [03:10:21] | CaCtus492 (CaCtus492!~Kent@123-243-197-152.static.tpgi.com.au) has quit (Remote host closed the connection) | |
| [03:10:31] | CaCtus491 (CaCtus491!~Kent@123-243-197-152.static.tpgi.com.au) has joined #mythtv | |
| [03:59:17] | stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has joined #mythtv | |
| [04:03:03] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [04:31:11] | spamjunkie (spamjunkie!~spamjunki@c-68-46-29-108.hsd1.pa.comcast.net) has joined #mythtv | |
| [04:52:13] | spamjunkie (spamjunkie!~spamjunki@c-68-46-29-108.hsd1.pa.comcast.net) has quit (Quit: Leaving) | |
| [05:05:23] | jya: | in a mpeg-ts, the resolution of the video stream, is that in the container, or in the stream itself ? |
| [05:05:42] | knightr_ (knightr_!~knightr@69-165-170-178.dsl.teksavvy.com) has quit (Read error: Connection reset by peer) | |
| [05:05:45] | knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv | |
| [05:06:54] | jya: | Well, the drop down of video resolution according to the bandwidth available, is really a lost cause using ffmpeg ; decoding will abort when the resolution change (AVD unknown decoding error)… so I guess myth player doesn't handle the change of resolution |
| [05:07:35] | knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Read error: Connection reset by peer) | |
| [05:07:55] | knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv | |
| [05:16:46] | Captain_Murdoch: | myth player should handle that I thought. I thought we had users with video samples where commercials were different sizes than the main feature. I know we have users where commercials are different aspect ratio but that may just be stretching. |
| [05:16:59] | Captain_Murdoch: | I believe it's only in the stream, not the container. |
| [05:17:08] | jya: | i see in the code that we handle change of refresh rate. |
| [05:17:18] | Captain_Murdoch: | I didn't think that the spec supported change of resolution, just change of bitrate. |
| [05:17:48] | jya: | but here the issue starts with ffmpeg not liking the change of resolution. I see plenty of h264 error about changes not handled in the thread version of h264 |
| [05:18:31] | jya: | Captain_Murdoch: the bipbop all from Apple site, has 4 streams, 2 of the same res different bitrate, and the other 2 different bitrate and different resolution |
| [05:19:25] | jya: | trying to play the Nasa live stream here… It fails after about 30s when the bandwidth makes it go to another stream… that shortly gives AVFD Unkown decoding error, and playback exit |
| [05:19:27] | Captain_Murdoch: | guess not, just reread that section of the spec... :| looks like audio must be identical between them but video is ok to change. |
| [05:20:28] | jya: | in the bipbop, I can handle bipbop 4 to bipbop 3 just fine, it's very smooth transition.. but transition to any other stream makes the player die |
| [05:21:08] | Captain_Murdoch: | may need to treat like a LiveTV transition I guess since that transition handles size changes. |
| [05:21:25] | jya: | I've lodged a bug at ffmpeg https://ffmpeg.org/trac/ffmpeg/ticket/1312 |
| [05:21:48] | jya: | but I'm fairly certain they will come back stating the issue is a user error as context needs to be reset or something |
| [05:22:13] | jya: | Captain_Murdoch: thanks for the hint about livetv transition.. will have a look at those |
| [05:22:19] | Captain_Murdoch: | I had my Roku streaming video from my laptop via HLS using libfaac. need to look at float conversion before we could use internal AAC encoder. |
| [05:22:52] | jya: | Captain_Murdoch: if you decode audio, you can request float output |
| [05:23:15] | jya: | most API will work with float for both decoding and encoding.. |
| [05:23:24] | jya: | even the AC3 bit which is very bizarre. |
| [05:23:33] | Captain_Murdoch: | any examples in our code that I could look at? |
| [05:23:47] | jya: | to decode audio or to convert to float ? |
| [05:24:30] | jya: | to convert to float, the function you want to look at is libs/libmyth/audio/audiooutpututil.cpp |
| [05:24:47] | Captain_Murdoch: | probably would have to convert. transcoder and HLS use the regular audio decoding code, so I think that's still int right? |
| [05:24:48] | jya: | toFloat(AudioFormat format, void *out, void *in, |
| [05:24:48] | jya: | int bytes) |
| [05:25:38] | jya: | AudioFormat is either FORMAT_U8, FORMAT_S16, etc… (handle 8, 16, 24 bits LSB, 24 bits and S32, and float) |
| [05:25:47] | jya: | and same for fromFloat |
| [05:26:38] | Captain_Murdoch: | so if input is 2048 S16, output will be 2048 floats? |
| [05:26:44] | jya: | yes |
| [05:26:54] | jya: | mind you, the size returned and used are all in bytes |
| [05:26:59] | jya: | not in the number of samples |
| [05:27:04] | Captain_Murdoch: | ok, thanks. sounds easy enough. I may take a look at that tomorrow. time for bed now. |
| [05:27:29] | jya: | so if converting 2 samples of 16 bits audio, size is 4 bytes |
| [05:27:49] | lincoln (lincoln!~quassel@203-114-182-185.dsl.sta.inspire.net.nz) has joined #mythtv | |
| [05:27:59] | jya: | but using floats all the way, may be faster … just saying :) |
| [05:28:14] | jya: | though, I'm pretty convinced our to and from float routine are faster than ffmpeg ones |
| [05:28:31] | jya: | courtesy of foobum |
| [05:29:18] | Captain_Murdoch: | I'll take a look. if I can get that to work and use the native AAC encoder then we could default to that for now and optionally provide a way to select mp3 audio instead. |
| [05:29:50] | lincoln (lincoln!~quassel@203-114-182-185.dsl.sta.inspire.net.nz) has quit (Client Quit) | |
| [05:30:47] | jya: | Captain_Murdoch: then why did it work when we just changed the codec name from libfaac to aac then ? |
| [05:31:00] | jya: | that would have make it use the native aac encoder i thought |
| [05:31:17] | lincoln (lincoln!~quassel@203-114-182-185.dsl.sta.inspire.net.nz) has joined #mythtv | |
| [05:31:30] | lincoln (lincoln!~quassel@203-114-182-185.dsl.sta.inspire.net.nz) has quit (Client Quit) | |
| [05:32:32] | Captain_Murdoch: | did you try that? with master? before or after ffmpeg resync? |
| [05:32:51] | lincoln (lincoln!~lincoln@203-114-182-185.dsl.sta.inspire.net.nz) has joined #mythtv | |
| [05:33:28] | jya: | with master, after ffmpeg resync |
| [05:33:48] | jya: | it was a one liner |
| [05:34:26] | Captain_Murdoch: | well, on my system, HLS wouldn't even work after the ffmpeg resync. not even for mp3. are you sure it was after the sync? |
| [05:34:49] | lincoln (lincoln!~lincoln@203-114-182-185.dsl.sta.inspire.net.nz) has quit (Client Quit) | |
| [05:34:50] | Captain_Murdoch: | I was getting segfaults because we weren't initializing the packet in the audio encoder. |
| [05:34:55] | Captain_Murdoch: | mp3 didn't even work |
| [05:35:00] | jya: | I didn't try it myself, I gave the patch to someone to try here and he did |
| [05:35:09] | jya: | and it worked with his sony phone |
| [05:35:57] | Captain_Murdoch: | I think he was either on 0.25 or fixes before the sync. the native aac encoder in the newer ffmpeg kicks an error out if you try to use it since it doesn't support S16. |
| [05:36:02] | jya: | yeah, with the new ffmpeg is very easy to get leaks too, you have to free yourself the packet now (but only if there are no errors, as if there are errors ffmpeg frees them itself) |
| [05:36:42] | Captain_Murdoch: | "Specified sample_fmt is not supported." is what I got when changed "libmp3lame" to "aac" |
| [05:36:56] | Captain_Murdoch: | that's in the libavcodec/utils.c file |
| [05:36:56] | jya: | ok.. my bad then... |
| [05:37:17] | jya: | but I remember having to change the AC3 encoder to use floats too |
| [05:37:31] | Captain_Murdoch: | no problem, I meant to reply to him to doublecheck the version he was using. |
| [05:38:07] | jya: | Captain_Murdoch: what if you change the sample_fmt used for the decoder, so it outputs floats ? |
| [05:38:32] | Captain_Murdoch: | the decoder is the same decoder used for playback, so I don't want to get that invasive. :) |
| [05:38:44] | jya: | i see... |
| [05:39:01] | jya: | I do get float outputs for some content |
| [05:39:05] | Captain_Murdoch: | and that would affect passthrough mode, etc. in mythtranscode as well. easier to just convert to float on encode. |
| [05:39:15] | jya: | for the audio framework it's all transparent, as it all works in floats anyway |
| [05:41:22] | Captain_Murdoch: | headed to bed. goal of watching a MythTV HLS video on the Roku was accomplished. :) even learned a little about making an app for the Roku looking at the samples and the MythRokuPlayer code. |
| [05:42:31] | jya: | good night |
| [05:43:15] | Captain_Murdoch: | 'night |
| [05:57:06] | pheld (pheld!~heldal@cl-5.osl-01.no.sixxs.net) has joined #mythtv | |
| [06:08:44] | jya: | anyone with some clue on readahead ? |
| [06:24:26] | fff (fff!62edcfc9@gateway/web/freenode/ip.98.237.207.201) has joined #mythtv | |
| [06:43:58] | Goga777 (Goga777!~Goga777@2.95.1.148) has joined #mythtv | |
| [07:00:01] | fff (fff!62edcfc9@gateway/web/freenode/ip.98.237.207.201) has quit (Ping timeout: 245 seconds) | |
| [07:21:52] | rsiebert_ (rsiebert_!~quassel@g226062193.adsl.alicedsl.de) has joined #mythtv | |
| [07:25:10] | rsiebert (rsiebert!~quassel@g226062099.adsl.alicedsl.de) has quit (Ping timeout: 272 seconds) | |
| [07:28:09] | foxbuntu (foxbuntu!~foxbuntu@ubuntu/member/foxbuntu) has quit (Ping timeout: 265 seconds) | |
| [07:29:36] | Goga777 (Goga777!~Goga777@2.95.1.148) has quit (Ping timeout: 265 seconds) | |
| [07:43:45] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has joined #mythtv | |
| [07:57:30] | MavT (MavT!~MaverickT@111.86.233.220.static.exetel.com.au) has quit (Ping timeout: 260 seconds) | |
| [08:05:55] | MaverickTech (MaverickTech!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv | |
| [08:11:34] | ThisNewGuy1 (ThisNewGuy1!~doug@pool-74-102-15-78.nwrknj.fios.verizon.net) has joined #mythtv | |
| [08:13:38] | stoffel (stoffel!~quassel@pD9E43712.dip.t-dialin.net) has joined #mythtv | |
| [08:14:12] | ThisNewGuy (ThisNewGuy!~doug@pool-74-102-15-78.nwrknj.fios.verizon.net) has quit (Ping timeout: 255 seconds) | |
| [08:38:00] | stoffel (stoffel!~quassel@pD9E43712.dip.t-dialin.net) has quit (Read error: Operation timed out) | |
| [08:49:30] | jya: | Captain_Murdoch: I had more thoughts about how to play the HLS generated by myth. Until the transcode is complete the stream is marked as live (it doesn't have the #EXT-X-ENDLIST. |
| [08:50:29] | jya: | Right now, for a live stream, I position the playback position to be at the end of the live stream, with just enough buffer to guarantee that playback won't stall… |
| [08:51:16] | jya: | So if I started to play the HLS stream with myth now, it would start at about where the transcode is up to.. |
| [08:52:35] | jya: | You mentioned that you preferred to start at the beginning of the list.. Unfortunately, doing so would break playback of some live streams. The first segment on a live stream (like NASA TV) is quite short lived. The playlist only have about 80s worth of video. |
| [08:53:21] | jya: | The HLSBuffer class is for playback only, it doesn't have the ability to buffer much, and if I did buffer the whole lot, it could virtually use GB of disk or RAM space. |
| [08:54:16] | jya: | So the choice is about playing the HLS stream generated by myth, which until they are fully transcoded are actually broken or properly play live stream. |
| [08:55:00] | jya: | Just like we can pause a live stream, but when you resume, once the buffer is exhausted it will resume where the live stream is up to |
| [09:44:56] | yb0t (yb0t!quassel@nat/nokia/x-mwlpvwwdmrdgcdsd) has joined #mythtv | |
| [09:46:43] | ybot_ (ybot_!quassel@nat/nokia/x-bkiwktammagunxik) has joined #mythtv | |
| [09:47:38] | ybot (ybot!quassel@nat/nokia/x-gndggdwnryixipxt) has quit (Ping timeout: 250 seconds) | |
| [09:49:41] | yb0t (yb0t!quassel@nat/nokia/x-mwlpvwwdmrdgcdsd) has quit (Ping timeout: 248 seconds) | |
| [10:05:03] | Guest19984 (Guest19984!~mike@c-98-232-220-158.hsd1.or.comcast.net) has quit (Remote host closed the connection) | |
| [10:06:16] | mike|2 (mike|2!~mike@c-98-232-220-158.hsd1.or.comcast.net) has joined #mythtv | |
| [10:09:42] | stoffel (stoffel!~quassel@pD9E43712.dip.t-dialin.net) has joined #mythtv | |
| [11:31:37] | foxbuntu (foxbuntu!~foxbuntu@174-30-104-65.desm.qwest.net) has joined #mythtv | |
| [11:31:44] | foxbuntu (foxbuntu!~foxbuntu@174-30-104-65.desm.qwest.net) has quit (Changing host) | |
| [11:31:44] | foxbuntu (foxbuntu!~foxbuntu@ubuntu/member/foxbuntu) has joined #mythtv | |
| [12:27:46] | stoffel (stoffel!~quassel@pD9E43712.dip.t-dialin.net) has quit (Ping timeout: 244 seconds) | |
| [12:32:15] | dmfrey (dmfrey!~dmfrey@webdefence.cluster-h.websense.net) has joined #mythtv | |
| [12:40:09] | Jordack (Jordack!~jordack@h69-131-44-221.plmomi.dedicated.static.tds.net) has joined #mythtv | |
| [13:21:44] | stoffel (stoffel!~quassel@pD9E43712.dip.t-dialin.net) has joined #mythtv | |
| [13:27:43] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [13:37:51] | jya: | committed HLS... |
| [13:41:27] | MythBuild: | build #3511 of master-linux-32bit is complete: Failure [4failed compile core] Build details are at http://code.mythtv.org/buildbot/builders/mast . . . /builds/3511 blamelist: Jim Stichnoth <jstichnoth@mythtv.org >, Jean-Yves Avenard <jyavenard@mythtv.org > |
| [13:44:36] | stuarta: | mostly works :) |
| [13:47:08] | stoffel (stoffel!~quassel@pD9E43712.dip.t-dialin.net) has quit (Ping timeout: 245 seconds) | |
| [13:48:33] | stuarta: | jya: can you drop me an email with the outline of what needs doing to swing the osx buildslave over to the new mechanism? |
| [13:49:10] | jya: | stuarta: it's all in the README file in the folder… |
| [13:49:51] | stuarta: | ah right. i'll have a bash when i get an evening |
| [13:49:55] | jya: | there's no openssl ?? |
| [13:50:02] | jya: | damn.. need to add a compilation option |
| [13:50:10] | stuarta: | it'll just need adding to the buildslave |
| [13:50:13] | danielk22 (danielk22!~danielk@96.57.9.142) has joined #mythtv | |
| [13:50:25] | jya: | i need to cater for that and disable AES-128 support |
| [13:50:35] | jya: | shouldn't be too complicated |
| [13:50:53] | jya: | pretty cool those buildbots... |
| [13:51:05] | stuarta: | yeah, they are nice |
| [13:51:17] | jya: | i don't think our AES-128 support for HLS makes any sense any way.. |
| [13:51:54] | jya: | it needs to download the decoding key, which is most likely on a protected server. and there's no way to prompt the user to enter credentials |
| [13:52:21] | jya: | i guess this could be added to MythDownloadManager.. automatically prompt for protected content |
| [13:52:25] | jya: | Captain_Murdoch: ^^ |
| [13:56:12] | kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has joined #mythtv | |
| [13:59:18] | jya: | stuarta: basically, it should be exactly how you are now doing, except that you set a Qt method: either -qtsdk or -qtbin & -qtplugins |
| [13:59:27] | blue (blue!~blue@irc.mmh.dk) has joined #mythtv | |
| [13:59:41] | jya: | I would delete the whole .osx-projects you currently have |
| [13:59:55] | stuarta: | oh i will :) |
| [14:01:32] | jya: | otherwise, all the options should remain the same, that is your stuff that stops after the dependencies has been built |
| [14:01:49] | jya: | if you manually run it with -verbose |
| [14:02:01] | jya: | it will print the environment variable it is using |
| [14:02:46] | jya: | as you're actually compiling the core mythtv without the build script, you'll like going to have to simulate the arguments I provide to configure and ENV |
| [14:03:17] | stuarta: | i'll see how i go when i get a chance to look at it |
| [14:03:45] | stuarta: | i'll find you if i need some assistance on it |
| [14:03:54] | jya: | sure |
| [14:04:08] | jya: | i may head to bed soon however.. some sleeps to catch on |
| [14:04:29] | stuarta: | oh, i can't even start looking at it for a minimum of 6hrs |
| [14:04:47] | stuarta: | may not even be tonight |
| [14:05:41] | stuarta: | go forth and sleep |
| [14:06:00] | jya: | ok |
| [14:25:06] | Jordack (Jordack!~jordack@h69-131-44-221.plmomi.dedicated.static.tds.net) has quit (Quit: ..must focus.. on work...) | |
| [14:27:40] | danielk22: | wagnerrp: The DishNet EIT lets you collect EIT on dishnet. It's available unencrypted to any DishNet customer. But they then need to use a HD-PVR or similar if they want to record anything other than the NASA channel or the fish bowl cam. |
| [14:28:50] | danielk22: | stichnot: If the preview image isn't too big in compressed size it is sent in the notification message. Otherwise it's transferred over the myth:// protocol like any other media. |
| [14:28:58] | stuarta: | i remember finding a fishtank channel on astra, but it's long since disappeared |
| [14:29:14] | stichnot: | jya: I've been unfairly implicated in the build failure. :) |
| [14:29:17] | stuarta: | that kept me busy "testing" for at least 2 mins |
| [14:29:28] | stuarta: | stichnot: so has he :) |
| [14:29:40] | stichnot: | ah, funny. |
| [14:29:41] | stuarta: | well less so than you |
| [14:30:14] | danielk22: | stuarta: I think they do have it so that installers know they are pointing at the right bird more quickly (i.e. before waiting 5 minutes for the proper keys to download.) |
| [14:30:16] | stuarta: | i spose the buildbot did pick up an actually issue, it fails to build if you don't have the openssl libs installed |
| [14:30:34] | stuarta: | danielk22: yeah sky have a welcome channel for just that |
| [14:30:45] | stuarta: | 2m promo loop |
| [14:30:51] | stuarta: | 2min* |
| [14:31:51] | danielk22: | stuarta: The fish are what the engineers come up with.. the welcome channel is what the marketers insist on once they discover the fish channel exists.. |
| [14:32:03] | stuarta: | hahahaha |
| [14:32:24] | ** stuarta demands a fish channel ** | |
| [14:35:29] | danielk22: | Echostar is completely run by engineers, they even spun off the marketing and customer relations into an independent company. |
| [14:35:41] | stichnot: | danielk22: I was thinking about how to request multiple previews at once from the same recording, so the ringbuffer opening, seeking, and decoding can be reused. Though I have no evidence yet whether there would be a performance issue to address there. |
| [14:37:01] | danielk22: | stichnot: If the previews are closely spaced then you can get a performance boost from this. |
| [14:38:24] | danielk22: | stichnot: I don't know if anyone mentioned it but my plan was for the preview generation to be handled by a long running daemon that is auto restarted whenever it crashes (with the smarts to avoid trying to regenerate the preview that caused the last crash). |
| [14:38:42] | stichnot: | danielk22: I would love that. |
| [14:40:41] | stichnot: | danielk22: the biggest performance boost would likely be when asking for multiple frames between the same pair of adjacent keyframes. This is even more common with HDPVR recordings where keyframes are 128 frames apart. |
| [14:42:01] | danielk22: | yep. |
| [14:42:57] | jya: | danielk22: a question for you as you are probably the most familiar |
| [14:43:05] | danielk22: | reading from disk uses some time and so does seeking, but decoding those frames and throwing away most of the results is going to be the biggest time waster. |
| [14:43:55] | jya: | 2 things… First i was testing the Readhead bit in the ring buffer, that thing calls safe_read without ever pausing, so exhaust quickly all my buffers and playback dies… any idea on how I can tell it to only call from time to time ? |
| [14:44:49] | jya: | 2: I've located the problem on seeking interrupting playback with a EOF detected. When I seek a big value, safe_read is called over and over, right after the message is displayed about flushing buffers |
| [14:45:22] | jya: | even if I have the whole stream loaded (over 500MB in RAM), it will call safe_read until all 500MB are gone, and exit with EOF |
| [14:48:40] | danielk22: | jya: The RingBuffer is too complex for me to be thinking about right now; lots of work stuff due in the next couple days.. Maybe this weekend. |
| [14:48:50] | jya: | no probs... |
| [14:49:08] | jya: | it's rather frustrating why it just dies like this when I seek.. I'm sure it's something dead obvious |
| [14:50:40] | danielk22: | I didn't apply that patch from some time ago because I didn't feel I really understood why it was dying.. I just saw where deep in the bowels it where was dying at least some of the time. |
| [14:51:32] | jya: | i think the issue I'm seeing is different. |
| [14:51:47] | jya: | it's not about the timestamp at the beginning being less than the end one |
| [14:51:56] | jya: | i never reach the code your patch touch |
| [14:53:10] | jya: | oh, another question, maybe you have a quick answer… when a stream change occur, and the resolution is different. AVFD just dies with an unknown decoding error… When format change is detected, I would need to restart the player and continue where it was. Similar to what livetv is doing.. could you give me a pointer where to look ? |
| [14:53:37] | danielk22: | jya: such a timecode shouldn't cause a problem, it's normal for the timecode to wrap. |
| [14:56:31] | danielk22: | jya: That was working at least with mpeg-2 files at some point, but i haven't tested in a while. Our mpegts.c calls a callback in our player that handles resolution and aspect ratio changes. We throw away some frames when this happens but the transition should occur without a decoding error or avfd dying. |
| [14:57:15] | danielk22: | The last time I'm sure it was working would have been about 4 years ago.. |
| [14:57:46] | jya: | ok… but I'm pretty sure my problem is different.. A seek from bool AvFormatDecoder::DoFastForward, that line: if (av_seek_frame(ic, -1, ts, flags) < 0) |
| [14:57:55] | jya: | make it call safe_read like crazy |
| [14:58:15] | jya: | maybe the callback isn't set |
| [14:58:44] | danielk22: | jya: Stick as much info as you can into the ticket and I'll look at it after I look at the jobs not starting for back to back recordings problem david reported. |
| [14:58:57] | jya: | ok |
| [15:03:37] | danielk22: | Is there a page on building MythTV on windows? http://www.mythtv.org/wiki/MythTV_on_Windows warns that it is out of date.. |
| [15:04:08] | stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 240 seconds) | |
| [15:11:26] | jya: | danielk22: you mean the ticket about duration not being calculated properly? |
| [15:11:41] | jya: | damn, I forgot to mention this in the hls commit log |
| [15:12:31] | danielk22: | jya: There is no ticket yet on the premature exit due to this particular bug? |
| [15:12:59] | jya: | haven't looked… no other ringbuffer seems to experience this when i seek |
| [15:13:22] | jya: | i tried to use the readahead thingy, but i can't get it to work :) |
| [15:15:31] | obo (obo!~obo@rockbox/developer/obo) has quit (Remote host closed the connection) | |
| [15:16:45] | obo (obo!~obo@70-140-190-120.lightspeed.hstntx.sbcglobal.net) has joined #mythtv | |
| [15:16:45] | obo (obo!~obo@70-140-190-120.lightspeed.hstntx.sbcglobal.net) has quit (Changing host) | |
| [15:16:45] | obo (obo!~obo@rockbox/developer/obo) has joined #mythtv | |
| [15:16:57] | jya: | danielk22: i thought the resolution was defined in the video stream and not the container… so how could the mythtv mpegts code handle it ? |
| [15:23:20] | danielk22: | jya: Hmm, it's actually in AvFormatDecoder::MpegPreProcessPkt() We read the sequence header and call SetVideoParams() if there is a change. |
| [15:23:51] | danielk22: | The MPEG-TS callback only gets involved if the PID of the video changes (which it can on resolution changes). |
| [15:24:53] | danielk22: | AvFormatDecoder::H264PreProcessPkt for H.264.. |
| [15:24:53] | jya: | ah, not sure if the pid does change here |
| [15:27:26] | blue (blue!~blue@irc.mmh.dk) has left #mythtv () | |
| [15:30:44] | jya: | it does test there for a change in resolution |
| [15:30:56] | jya: | doesn't seem to work :) |
| [15:31:59] | jya: | danielk22: i've posted a few samples there: https://ffmpeg.org/trac/ffmpeg/ticket/1312 |
| [15:33:46] | danielk22: | jya: Look at the MpegPreProcessPkt, it's actually decoding the WxH, but H264PreProcessPkt is just pulling it from ffmpeg which presumably hasn't seen this packet yet since this is in the preprocessing stage... |
| [15:34:35] | danielk22: | It looks like this just isn't implemented for H.264 yet.. it kinda works with aspect changes, but it will actually detect those a whole keyframe later than it should. |
| [15:35:13] | danielk22: | With WxH changes this will blow up since the video frames themselves will be the wrong size.. possibly too small. |
| [15:37:27] | jya: | cool… i'll have a look tomorrow at this.. right now, Im so tired I can barely keep my eyes opened |
| [15:38:22] | jya: | all of this codec stuff in what it does in MpegPreProcessPkt is way over my head |
| [15:42:25] | MythBuild: | build #3512 of master-linux-32bit is complete: Success [3build successful] Build details are at http://code.mythtv.org/buildbot/builders/mast . . . /builds/3512 |
| [15:42:40] | stuarta: | \o/ |
| [15:42:42] | ** jya is clapping at himself ** | |
| [15:46:44] | MaverickTech (MaverickTech!~MaverickT@111.86.233.220.static.exetel.com.au) has quit (Ping timeout: 252 seconds) | |
| [15:48:48] | jya: | danielk22: i put a breakpoint in H264PreProcessPkt |
| [15:48:59] | jya: | and it is correctly detecting the change of resolution |
| [15:50:13] | jya: | it seems to do just the same as what it does for the mpeg2 code |
| [15:50:31] | jya: | so not sure why it doesn't work |
| [15:53:35] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [16:05:17] | stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has joined #mythtv | |
| [16:09:00] | joki (joki!~joki@p54862B27.dip.t-dialin.net) has quit (Ping timeout: 244 seconds) | |
| [16:09:41] | joki (joki!~joki@p548622AD.dip.t-dialin.net) has joined #mythtv | |
| [16:11:17] | stoffel (stoffel!~quassel@pD9E43712.dip.t-dialin.net) has joined #mythtv | |
| [16:16:27] | danielk22: | jya: It looks to me like it detects the change of resolution at the next sequence header not the one it is supposed to detect it at. This means the video output methods will feed ffmpeg video frames of the wrong size for 12 to 128 frames in practice (possibly many more in theory). |
| [16:19:42] | Steve-Goodey (Steve-Goodey!~steve@host86-144-2-182.range86-144.btcentralplus.com) has joined #mythtv | |
| [16:33:18] | stoffel (stoffel!~quassel@pD9E43712.dip.t-dialin.net) has quit (Remote host closed the connection) | |
| [16:36:21] | stoffel (stoffel!~quassel@pD9E43712.dip.t-dialin.net) has joined #mythtv | |
| [16:41:40] | dekarl1 (dekarl1!~dekarl@p4FCEFC5D.dip.t-dialin.net) has joined #mythtv | |
| [16:43:30] | dekarl (dekarl!~dekarl@p4FE85E0C.dip.t-dialin.net) has quit (Ping timeout: 265 seconds) | |
| [17:26:12] | Chutt_ (Chutt_!~ijr@cpe-24-29-225-175.neo.res.rr.com) has joined #mythtv | |
| [17:28:06] | Chutt (Chutt!~ijr@cpe-24-29-225-175.neo.res.rr.com) has quit (Ping timeout: 272 seconds) | |
| [17:33:06] | wagnerrp: | stuartm: you in? |
| [17:35:29] | Chutt_ is now known as Chutt | |
| [17:36:26] | wagnerrp: | stuartm: what are your plans regarding mythwelcome? |
| [17:36:32] | wagnerrp: | interesting issue on the mailing list |
| [17:36:52] | wagnerrp: | since all UI applications run the UDP listener for mythmessage |
| [17:37:08] | wagnerrp: | mythwelcome binds to those ports, rendering them unavailable for mythfrontend |
| [17:37:39] | wagnerrp: | if mythwelcome is here to stay, ill need to put in some mechanism to allow mythwelcome to release those ports to mythfrontend, and acquire them again when it terminates |
| [17:38:08] | wagnerrp: | if you intend to replace all of its functionality with mythfrontend's new idle mode, i wont bother |
| [17:52:00] | stichnot: | I'm starting to get back to #9727 and the "Seek to exact frame" setting. This patch adds a mechanism (but not a policy) for making exact/inexact seeking dependent on how close the target frame is to a keyframe. Some IRC discussion (http://irc.mythtv.org/ircLog/channel/4/2011-04-13:16:04:21) favors removing the ExactSeeking setting in favor of a "do the right thing" approach. Captain_Murdoch sug |
| [17:52:00] | ** MythLogBot http://code.mythtv.org/trac/ticket/9727 ** | |
| [17:52:01] | stichnot: | gests a policy of exact seeking only if the target is not too far past a keyframe, so that seek latencies are more or less bounded. I'm wondering if there are any other opinions on the matter. |
| [17:52:51] | stichnot: | My opinion is that the only place there should be anything close to exact seeking is in the cutlist editor. |
| [17:53:06] | stichnot: | oh, and seeking to a bookmark |
| [17:54:04] | stichnot: | All other seeking is based on relative and absolute times, which aren't really guaranteed to be accurate in the first place. |
| [17:57:24] | stoffel (stoffel!~quassel@pD9E43712.dip.t-dialin.net) has quit (Ping timeout: 244 seconds) | |
| [18:05:54] | stuartm: | wagnerrp: that really depends on other people, I'm all for getting rid of it but there were strong objections the last time I mentioned it |
| [18:10:06] | GreyFoxx (GreyFoxx!~greg@out.of.phaze.org) has quit (*.net *.split) | |
| [18:10:06] | xris (xris!~xris@mythtv/developer/xris) has quit (*.net *.split) | |
| [18:10:06] | jams (jams!~jams@cpe-184-58-217-97.wi.res.rr.com) has quit (*.net *.split) | |
| [18:10:41] | GreyFoxx (GreyFoxx!~greg@out.of.phaze.org) has joined #mythtv | |
| [18:10:41] | xris (xris!~xris@mythtv/developer/xris) has joined #mythtv | |
| [18:10:41] | jams (jams!~jams@cpe-184-58-217-97.wi.res.rr.com) has joined #mythtv | |
| [18:16:10] | stoffel (stoffel!~quassel@pD9E43712.dip.t-dialin.net) has joined #mythtv | |
| [19:19:20] | stichnot: | sphery: right, in the current patch, single-frame and keyframe seeks use a "snap" value of 0 |
| [19:19:53] | stichnot: | other seeks use a snap value of 50% of the seek distance |
| [19:20:09] | stichnot: | which is probably too high |
| [19:22:07] | stichnot: | but since my initial motivation was to speed up seeking in cutlist editing, especially for HD-PVR recordings, 50% was just fine for me |
| [19:24:21] | stichnot: | btw, I observed that HW decoding of HD-PVR recordings on an ION-1 goes at about 2x realtime, e.g. decoding 120 frames of a 720p recording can take 1 second. Does anyone know how fast it is on a Sandybridge CPU using software decoding? |
| [19:32:30] | Beirdo: | wagnerrp: what would you like the repo for trac bits to be called? |
| [19:32:38] | Beirdo: | or was it some other plugins, I forget :) |
| [19:32:57] | wagnerrp: | doesnt matter |
| [19:32:58] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has quit (Remote host closed the connection) | |
| [19:33:02] | wagnerrp: | trac plugins, wiki plugins |
| [19:33:14] | wagnerrp: | basically anything that gets used for spam filtering |
| [19:33:18] | Beirdo: | hmm. you want one repo for both? |
| [19:33:39] | Beirdo: | antispam sound like a good name? |
| [19:34:03] | wagnerrp: | good as anything else |
| [19:34:36] | Beirdo: | OK. and with no mirroring to github... |
| [19:35:58] | Beirdo: | OK |
| [19:36:12] | Beirdo: | done. New repo for ya ;) |
| [19:39:16] | SteveGoodey (SteveGoodey!~steve@host86-144-2-182.range86-144.btcentralplus.com) has joined #mythtv | |
| [19:51:46] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Remote host closed the connection) | |
| [19:59:13] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [20:05:24] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has joined #mythtv | |
| [20:05:24] | stuartm (stuartm!~stuartm@cpc1-derb9-0-0-cust441.8-3.cable.virginmedia.com) has quit (Changing host) | |
| [20:05:24] | stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has joined #mythtv | |
| [20:32:15] | stichnot_ (stichnot_!~chatzilla@mythtv/developer/stichnot) has joined #mythtv | |
| [20:33:38] | stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 240 seconds) | |
| [20:33:41] | stichnot_ is now known as stichnot | |
| [20:35:51] | SteveGoodey (SteveGoodey!~steve@host86-144-2-182.range86-144.btcentralplus.com) has quit (Remote host closed the connection) | |
| [20:42:32] | jstenback (jstenback!~jstenback@2620:101:8003:200:224:e8ff:fe39:34c2) has quit (Ping timeout: 272 seconds) | |
| [20:46:39] | jstenback (jstenback!~jstenback@2620:101:8003:200:224:e8ff:fe39:34c2) has joined #mythtv | |
| [20:48:37] | stuarta: | the osx slave is offline until further notice. i'm attempting to convert it to the new build scheme |
| [20:54:48] | dmfrey (dmfrey!~dmfrey@webdefence.cluster-h.websense.net) has quit (Quit: Ex-Chat) | |
| [20:55:32] | stuarta (stuarta!~stuarta@callisto.ppp0.squashedfrog.net) has quit (Changing host) | |
| [20:55:32] | stuarta (stuarta!~stuarta@mythtv/developer/stuarta) has joined #mythtv | |
| [21:06:10] | andreax (andreax!~andreaz@p54BF3B74.dip.t-dialin.net) has joined #mythtv | |
| [21:15:52] | Steve-Goodey (Steve-Goodey!~steve@host86-144-2-182.range86-144.btcentralplus.com) has quit (Remote host closed the connection) | |
| [21:17:10] | mako-sb (mako-sb!~makoto@ludgate.privmsg.me.uk) has joined #mythtv | |
| [21:17:24] | mako-sb (mako-sb!~makoto@ludgate.privmsg.me.uk) has left #mythtv ("whoops need the users chan, sorry") | |
| [21:21:31] | mrand (mrand!~mrand@ubuntu/member/mrand) has quit (Ping timeout: 260 seconds) | |
| [21:24:57] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [21:26:08] | Beirdo: | stuarta: cool, good to know |
| [21:26:21] | Beirdo: | that reminds me. gotta finish the restart of the winblows one |
| [21:29:35] | stuarta: | shall we kill off the 0.24 build slaves as well? don't see much point in keeping them now |
| [21:30:19] | Beirdo: | Yeah, I'll go do that if ya want |
| [21:30:49] | stuarta: | i know it was suggested a while ago by someone |
| [21:32:44] | mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv | |
| [21:35:58] | amessina (amessina!~amessina@2001:470:1f11:a4:d6be:d9ff:fe8d:7c1e) has quit (Quit: Konversation terminated!) | |
| [21:36:01] | MythBuild (MythBuild!~MythBuild@alcor.mythtv.org) has quit (Quit: buildmaster reconfigured: bot disconnecting) | |
| [21:36:05] | MythBuild_ (MythBuild_!~MythBuild@alcor.mythtv.org) has joined #mythtv | |
| [21:37:00] | Beirdo: | done |
| [21:39:21] | stuarta: | hey now it fits in the width of one web page :) |
| [21:41:49] | MaverickTech (MaverickTech!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv | |
| [22:03:43] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [22:06:32] | Beirdo: | yeah :) |
| [22:06:41] | Beirdo: | you can put ?category=master |
| [22:06:49] | Beirdo: | and get just master builds (BTW) |
| [22:06:56] | stuarta: | never knew that |
| [22:07:10] | Beirdo: | it is well buried in the docs :) |
| [22:08:01] | Beirdo: | I want my hot-air rework station already! |
| [22:08:09] | Beirdo: | it's supposed to arrive tomorrow |
| [22:15:57] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [22:22:52] | kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has quit (Quit: kormoc) | |
| [22:23:47] | kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has joined #mythtv | |
| [22:37:41] | stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 248 seconds) | |
| [22:42:05] | andreax (andreax!~andreaz@p54BF3B74.dip.t-dialin.net) has quit (Read error: Connection reset by peer) | |
| [23:06:18] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [23:30:34] | knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Read error: Operation timed out) | |
| [23:31:48] | knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv | |
IRC Logs collected by
BeirdoBot.
Please use the above link to report any bugs.