| Monday, January 7th, 2013, 00:24 UTC | ||
| [00:24:16] | MavT (MavT!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv | |
| [00:25:39] | MaverickTech (MaverickTech!~MaverickT@111.86.233.220.static.exetel.com.au) has quit (Ping timeout: 260 seconds) | |
| [00:37:27] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Quit: Body blow! Body blow!) | |
| [00:50:49] | joki- (joki-!~joki@p54865667.dip.t-dialin.net) has joined #mythtv | |
| [00:51:18] | joki (joki!~joki@p548629BC.dip.t-dialin.net) has quit (Ping timeout: 276 seconds) | |
| [00:51:19] | joki- is now known as joki | |
| [00:59:31] | dblain (dblain!~dblain@mythtv/developer/dblain) has joined #mythtv | |
| [01:53:48] | jpabq: | neufeld_AFK: danielk22: Actually, with the HD-PVR, to get ffprobe happy, we need to turn OFF the option of waiting for a keyframe before starting to write to the file. |
| [01:54:57] | jpabq: | With it ON, we end up skipping over to preamble stuff that ffprobe seems to need. |
| [01:55:25] | jheizer_laptop: | Woot! Next MobileMyth update will not require the client to have any direct connection to the MBE. Everything is URL rewritten and proxied. |
| [02:08:06] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [02:08:24] | lentferj (lentferj!~lentferj@p579B71FE.dip.t-dialin.net) has quit (Ping timeout: 252 seconds) | |
| [02:09:31] | lentferj (lentferj!~lentferj@p579B7FC7.dip.t-dialin.net) has joined #mythtv | |
| [02:28:46] | jpabq: | neufeld_AFK, neufeld, danielk22: The "start of the frame" is detected at the beginning of the second "packet". However, we don't figure out that it is a keyframe, until we have processed 1000 packets. Assuming a packet is 188 bytes. |
| [02:34:03] | Guest71366 (Guest71366!~quassel@75-161-183-113.mpls.qwest.net) has quit (Remote host closed the connection) | |
| [02:36:39] | jheizer_laptop (jheizer_laptop!~jon@c-98-226-220-178.hsd1.il.comcast.net) has quit (Ping timeout: 260 seconds) | |
| [02:38:15] | jheizer_laptop (jheizer_laptop!~jon@c-98-226-220-178.hsd1.il.comcast.net) has joined #mythtv | |
| [02:45:16] | Guest32100 (Guest32100!~quassel@75-161-183-113.mpls.qwest.net) has joined #mythtv | |
| [02:48:40] | neufeld_AFK is now known as neufeld | |
| [02:53:32] | neufeld: | jpabq: I'm not sure I understand. If I generate a transport stream with ffmpeg, it produces a file whose first frame is a keyframe, 564 bytes into the recording. I assumed that the 564 bytes is an encapsulation header for the container, so the first actual frame data sits that far into the data stream. ffprobe is happy to see this, and doesn't generate any noise. So, when the HD-PVR generates its version of a |
| [02:53:33] | neufeld: | transport stream, is it that the code is dropping this (what I have been assuming is) encapsulation header, so that the first keyframe starts at byte pos=0? So, the keyframe is losing its container header? Is that why ffmpeg seems to be unable to sync up, and runs through up to 400+ kB of data before it finds its place again? |
| [02:56:48] | jpabq: | neufeld: yes I believe that is what is happening. With the data from the HD-PVR, we end up skipping packets until we find the packet that actually tells us that the frame we are in is a keyframe. By then, we have thrown away so much data, that ffmpeg is confused. |
| [02:57:54] | jpabq: | Looking at the code, I see comments that indicate that Myth was setup to handle this situation — at least once-upon-a-time. But, I don't think it works anymore. |
| [03:00:33] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 252 seconds) | |
| [03:01:00] | neufeld: | jpabq: I see. OK, that's interesting. I'll have to take a look at that some time, as I have the hardware and repeatable test cases. Thanks. |
| [03:01:02] | stichnot (stichnot!~stichnot@adsl-69-110-235-166.dsl.pltn13.pacbell.net) has joined #mythtv | |
| [03:01:02] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv | |
| [03:01:02] | stichnot (stichnot!~stichnot@adsl-69-110-235-166.dsl.pltn13.pacbell.net) has quit (Changing host) | |
| [03:02:15] | jpabq: | neufeld: I will take a stab at fixing it, but I want danielk22's input first. It is possible that I am missing something. |
| [03:04:47] | neufeld: | jpabq: thanks. This issue causes pain in both the HD-PVR captioning code and the precise-cutting script on the wiki. |
| [03:04:50] | jheizer_laptop (jheizer_laptop!~jon@c-98-226-220-178.hsd1.il.comcast.net) has quit (Ping timeout: 255 seconds) | |
| [03:05:18] | danielk22: | jpabq: I believe that data should all be queued up in the _payload_buffer, and then spit out as soon as we're sure it is a keyframe. |
| [03:08:23] | jpabq: | danielk22: yes, but because of statements like "if (_wait_for_keyframe_option && _first_keyframe < 0)" it never makes it into the payload_buffer. At least for H.264 streams. |
| [03:09:22] | neufeld is now known as neufeld_AFK | |
| [03:13:06] | danielk22: | jpabq: I think the determining if it is a keyframe vs just a frame probably happens in the same packet with MPEG-2, but not so for H.264, so we need more smarts for H.264. Maybe a separate buffer that just keeps the current video pes packet's ts packets and pushes it to _payload_buffer if it turns out this pes is the first keyframe. |
| [03:14:39] | jpabq: | Testing H.264 from the HD-PVR, we do find out that it is a frame and a keyframe in the same pkt, but we don't find either out until 1000 packets into the "frame". |
| [03:15:18] | danielk22: | Yikes, I'm sure that is the issue then. We need the beginning of that PES packet.. |
| [03:15:45] | jpabq: | Yeah. |
| [03:16:14] | danielk22: | How has this not caused issues before? Did an older ffmpeg just ignore the data until the first keyframe? |
| [03:16:53] | jpabq: | I was going to change the logic such that we do buffer it all up in payload_buffer, and then if it turns out not to be a keyframe, throw all that data away. Other option is, like you said, to have a second buffer going. |
| [03:17:33] | danielk22: | Either one is fine. Just implement the one that is simpler to reason about. |
| [03:17:39] | jpabq: | This is not actually causing us a problem, because ffmpeg will ignore the data until it finds something it is happy with. However, it is just a little ugly. |
| [03:18:07] | jpabq: | If you run ffprobe on the file generated by Myth, is complains a lot, until it finds good data. |
| [03:18:17] | jpabq: | but we can play the file just fine. |
| [03:18:30] | danielk22: | yep, I agree. I also wonder if it means some frames are not shown with back-to-back recordings. |
| [03:18:54] | jpabq: | That is possible. |
| [03:18:54] | danielk22: | A whole 128 frames might be lost, that is a few seconds at 30 fps. |
| [03:19:27] | neufeld_AFK: | Just dropping in: it causes some problems in ffmpeg, but not in playback. First, the ffprobe might give completely wrong bitrates and durations. Second, if ffmpeg can't figure out the keyframes early enough in the file, it makes up its own pts values, but starting from t=1s instead of t=0s. |
| [03:19:40] | jpabq: | I have not analyzed any H.264 from any source other than the HD-PVR. |
| [03:20:19] | danielk22: | It is common code for all H.264. It may very well cause more problems for non-HD-PVR sources. |
| [03:21:41] | stichnot (stichnot!~stichnot@adsl-69-110-235-166.dsl.pltn13.pacbell.net) has joined #mythtv | |
| [03:21:42] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv | |
| [03:21:42] | stichnot (stichnot!~stichnot@adsl-69-110-235-166.dsl.pltn13.pacbell.net) has quit (Changing host) | |
| [03:33:59] | stichnot (stichnot!~stichnot@adsl-69-110-235-166.dsl.pltn13.pacbell.net) has joined #mythtv | |
| [03:34:00] | stichnot (stichnot!~stichnot@adsl-69-110-235-166.dsl.pltn13.pacbell.net) has quit (Changing host) | |
| [03:34:00] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv | |
| [03:34:58] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [03:39:48] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [03:50:15] | terron (terron!~t@unaffiliated/terron) has joined #mythtv | |
| [03:50:21] | terron (terron!~t@unaffiliated/terron) has left #mythtv () | |
| [04:40:41] | sraue (sraue!~stephan@xbmc/staff/sraue) has quit (Ping timeout: 255 seconds) | |
| [04:42:41] | fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has quit (Ping timeout: 248 seconds) | |
| [04:43:51] | fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has joined #mythtv | |
| [04:53:21] | sraue (sraue!~stephan@xbmc/staff/sraue) has joined #mythtv | |
| [05:39:09] | amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has quit (Quit: Konversation terminated!) | |
| [06:28:50] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has joined #mythtv | |
| [06:34:01] | stoffel (stoffel!~quassel@pD9E4272F.dip.t-dialin.net) has joined #mythtv | |
| [06:39:11] | rsiebert (rsiebert!~quassel@g224249100.adsl.alicedsl.de) has joined #mythtv | |
| [06:42:14] | rsiebert_ (rsiebert_!~quassel@g224248073.adsl.alicedsl.de) has quit (Ping timeout: 260 seconds) | |
| [07:01:25] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has joined #mythtv | |
| [07:11:05] | rasusto (rasusto!~devnull@70.44.136.21.res-cmts.sm.ptd.net) has joined #mythtv | |
| [07:11:16] | rasusto (rasusto!~devnull@70.44.136.21.res-cmts.sm.ptd.net) has left #mythtv () | |
| [07:28:10] | ggervasio (ggervasio!~gervasio@2601:9:4b00:5d:d00f:c4d0:66d4:7e61) has left #mythtv () | |
| [07:35:40] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has quit (Quit: Leaving) | |
| [07:46:49] | toeb (toeb!~tob@HSI-KBW-37-49-118-160.hsi14.kabel-badenwuerttemberg.de) has joined #mythtv | |
| [08:16:58] | ggervasio (ggervasio!~gervasio@2601:9:4b00:5d:e575:5087:41db:920a) has joined #mythtv | |
| [08:48:22] | bas-t (bas-t!~tycho@52484E89.cm-4-1b.dynamic.ziggo.nl) has joined #mythtv | |
| [09:02:43] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has joined #mythtv | |
| [09:04:52] | Merlin83b (Merlin83b!~Daniel@2a00:1ee0:3:1337:4012:9785:4027:60e0) has joined #mythtv | |
| [09:23:01] | jwhite (jwhite!~jwhite@75-146-153-89-minnesota.hfc.comcastbusiness.net) has quit (Read error: Operation timed out) | |
| [09:28:31] | jwhite (jwhite!~jwhite@75-146-153-89-minnesota.hfc.comcastbusiness.net) has joined #mythtv | |
| [09:40:16] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has quit (Remote host closed the connection) | |
| [10:06:44] | bas-t (bas-t!~tycho@52484E89.cm-4-1b.dynamic.ziggo.nl) has quit (Quit: Quit) | |
| [10:07:08] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has joined #mythtv | |
| [10:14:20] | Guest32100 (Guest32100!~quassel@75-161-183-113.mpls.qwest.net) has quit (Remote host closed the connection) | |
| [10:42:34] | purserj (purserj!~purserj@hosting.collaborynth.com.au) has quit (Ping timeout: 246 seconds) | |
| [10:43:11] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has quit (Remote host closed the connection) | |
| [10:43:32] | purserj (purserj!~purserj@hosting.collaborynth.com.au) has joined #mythtv | |
| [10:46:10] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has joined #mythtv | |
| [10:57:22] | anykey_ (anykey_!~anykey@46-126-243-153.dynamic.hispeed.ch) has quit (Quit: leaving) | |
| [10:59:05] | anykey_ (anykey_!~anykey@46-126-243-153.dynamic.hispeed.ch) has joined #mythtv | |
| [11:02:54] | anykey_ (anykey_!~anykey@46-126-243-153.dynamic.hispeed.ch) has quit (Client Quit) | |
| [11:08:23] | anykey_ (anykey_!~anykey@46-126-243-153.dynamic.hispeed.ch) has joined #mythtv | |
| [11:28:53] | wahrhaft (wahrhaft!~quassel@cpe-24-210-71-0.columbus.res.rr.com) has quit (Quit: No Ping reply in 180 seconds.) | |
| [11:29:17] | wahrhaft (wahrhaft!~quassel@cpe-24-210-71-0.columbus.res.rr.com) has joined #mythtv | |
| [11:33:58] | stoffel (stoffel!~quassel@pD9E4272F.dip.t-dialin.net) has quit (Remote host closed the connection) | |
| [11:35:41] | wahrhaft (wahrhaft!~quassel@cpe-24-210-71-0.columbus.res.rr.com) has quit (Quit: No Ping reply in 180 seconds.) | |
| [11:36:07] | wahrhaft (wahrhaft!~quassel@cpe-24-210-71-0.columbus.res.rr.com) has joined #mythtv | |
| [11:50:01] | zombor (zombor!~zombor__@65.29.231.135) has joined #mythtv | |
| [11:50:02] | zombor (zombor!~zombor__@65.29.231.135) has quit (Changing host) | |
| [11:50:02] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [11:52:48] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [12:04:03] | IReboot (IReboot!~doug@CPE10bf48e67915-CM00252eac6f40.cpe.net.cable.rogers.com) has quit (Remote host closed the connection) | |
| [12:06:10] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has quit (Remote host closed the connection) | |
| [12:07:03] | IReboot (IReboot!~doug@99.234.136.184) has joined #mythtv | |
| [12:08:51] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has joined #mythtv | |
| [12:14:41] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [12:18:45] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Read error: Connection reset by peer) | |
| [12:40:22] | Merlin83b2 (Merlin83b2!~Daniel@2a00:1ee0:3:1337:4012:9785:4027:60e0) has joined #mythtv | |
| [12:40:23] | Merlin83b (Merlin83b!~Daniel@2a00:1ee0:3:1337:4012:9785:4027:60e0) has quit (Read error: Connection reset by peer) | |
| [12:42:25] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [13:03:36] | stoffel (stoffel!~quassel@pD9E4272F.dip.t-dialin.net) has joined #mythtv | |
| [13:22:51] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [13:56:08] | amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has joined #mythtv | |
| [14:14:45] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has quit (Remote host closed the connection) | |
| [14:16:52] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has joined #mythtv | |
| [14:19:03] | tmr_ (tmr_!~quassel@chello084114028118.14.vie.surfer.at) has joined #mythtv | |
| [14:23:20] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [14:23:22] | dmfrey (dmfrey!~dmfrey@webdefence.cluster-h.websense.net) has joined #mythtv | |
| [14:32:53] | stoffel (stoffel!~quassel@pD9E4272F.dip.t-dialin.net) has quit (Ping timeout: 255 seconds) | |
| [15:11:28] | amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has quit (Ping timeout: 260 seconds) | |
| [15:16:13] | tmr_ (tmr_!~quassel@chello084114028118.14.vie.surfer.at) has quit (Remote host closed the connection) | |
| [15:17:30] | IReboot (IReboot!~doug@99.234.136.184) has quit (Ping timeout: 272 seconds) | |
| [15:19:52] | IReboot (IReboot!~doug@CPE10bf48e67915-CM00252eac6f40.cpe.net.cable.rogers.com) has joined #mythtv | |
| [15:20:28] | XChatMav (XChatMav!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv | |
| [15:24:11] | MavT (MavT!~MaverickT@111.86.233.220.static.exetel.com.au) has quit (Ping timeout: 252 seconds) | |
| [15:32:47] | amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has joined #mythtv | |
| [15:52:14] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 240 seconds) | |
| [15:57:42] | foobum (foobum!~foobum@cpc23-acto2-2-0-cust294.4-2.cable.virginmedia.com) has quit (Ping timeout: 276 seconds) | |
| [16:02:04] | foobum (foobum!~foobum@cpc23-acto2-2-0-cust294.4-2.cable.virginmedia.com) has joined #mythtv | |
| [16:11:03] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has joined #mythtv | |
| [16:28:25] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has quit (Read error: Operation timed out) | |
| [16:37:52] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv | |
| [16:49:05] | stichnot: | danielk22: jpabq: neufeld: I strongly suspect that the hdpvr initial keyframe issue is at least partially responsible for the discrepancy in seektable frame numbers between the recorder and mythcommflag --rebuild. |
| [16:53:41] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [17:03:42] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 265 seconds) | |
| [17:14:12] | Korny1 (Korny1!~Korny@c-75-64-127-62.hsd1.tn.comcast.net) has joined #mythtv | |
| [17:23:00] | Merlin83b2 (Merlin83b2!~Daniel@2a00:1ee0:3:1337:4012:9785:4027:60e0) has quit (Read error: Connection reset by peer) | |
| [17:23:25] | Merlin83b2 (Merlin83b2!~Daniel@2a00:1ee0:3:1337:4012:9785:4027:60e0) has joined #mythtv | |
| [17:32:24] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has quit (Quit: FabriceMG) | |
| [17:34:10] | Daniel__ (Daniel__!~Daniel@2a00:1ee0:3:1337:4012:9785:4027:60e0) has joined #mythtv | |
| [17:34:11] | Merlin83b2 (Merlin83b2!~Daniel@2a00:1ee0:3:1337:4012:9785:4027:60e0) has quit (Read error: Connection reset by peer) | |
| [17:43:34] | stichnot (stichnot!stichnot@nat/google/x-mfagnoufewroqbrt) has joined #mythtv | |
| [17:43:34] | stichnot (stichnot!stichnot@nat/google/x-mfagnoufewroqbrt) has quit (Changing host) | |
| [17:43:34] | stichnot (stichnot!stichnot@mythtv/developer/stichnot) has joined #mythtv | |
| [17:44:41] | SteveGoodey (SteveGoodey!~steve@host86-148-172-139.range86-148.btcentralplus.com) has joined #mythtv | |
| [17:46:48] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Read error: Connection reset by peer) | |
| [17:48:10] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has joined #mythtv | |
| [17:48:10] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has quit (Changing host) | |
| [17:48:10] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [17:49:22] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Read error: Connection reset by peer) | |
| [17:49:37] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has joined #mythtv | |
| [17:49:38] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has quit (Changing host) | |
| [17:49:38] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [18:00:05] | Daniel__ (Daniel__!~Daniel@2a00:1ee0:3:1337:4012:9785:4027:60e0) has quit (Read error: Connection reset by peer) | |
| [18:05:19] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Ping timeout: 260 seconds) | |
| [18:14:05] | SteveGoodey (SteveGoodey!~steve@host86-148-172-139.range86-148.btcentralplus.com) has quit (Quit: Konversation terminated!) | |
| [18:16:11] | SteveGoodey (SteveGoodey!~steve@host86-148-172-139.range86-148.btcentralplus.com) has joined #mythtv | |
| [18:16:50] | neufeld_AFK is now known as neufeld | |
| [18:18:00] | neufeld: | stichnot: possibly. My theory was that there was something in the frame reassembly code stitches together read blocks, occasionally it might get confused, miss a start marker, and merge together two consecutive frames believing them to be a single one. |
| [18:18:13] | Steve-Goodey (Steve-Goodey!~steve@host86-148-172-139.range86-148.btcentralplus.com) has joined #mythtv | |
| [18:18:36] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Quit: Body blow! Body blow!) | |
| [18:24:01] | stichnot: | neufeld: are you talking about the issue of the keyframe distance occasionally showing up as 126 or 127 instead of 128? I was actually referring to the discrepancy in initial keyframe number. |
| [18:24:22] | dekarl (dekarl!~dekarl@p4FCEEA0F.dip.t-dialin.net) has quit (Ping timeout: 255 seconds) | |
| [18:27:09] | neufeld: | stichnot: I'm referring to the fact that when I use ffmpeg, I can ask for keyframes, and get frame number and file offset. When mythcommflag --rebuild runs on the same file (with your patch), I get exactly the same file offsets, but the frame number falls progressively further and further behind, with no visible periodicity. That's what makes me think there's a subtle bug in the frame count-up sequence. |
| [18:28:11] | neufeld: | mythcommflag --rebuild never includes the n=0 frame, it only logs the keyframes after the n=0 frame. |
| [18:29:11] | dekarl (dekarl!~dekarl@p4FE854BE.dip.t-dialin.net) has joined #mythtv | |
| [18:35:49] | neufeld: | stichnot: BTW, thanks for looking over #11326 and explaining to me what was happening, I agree it's a WONTFIX issue. |
| [18:35:49] | ** MythLogBot http://code.mythtv.org/trac/ticket/11326 ** | |
| [18:55:25] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has joined #mythtv | |
| [18:55:26] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has quit (Changing host) | |
| [18:55:26] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [18:57:29] | stichnot: | neufeld: how do you run ffmpeg to get frames/offsets for keyframes? |
| [18:57:31] | stoffel (stoffel!~quassel@pD9E4272F.dip.t-dialin.net) has joined #mythtv | |
| [18:57:48] | stoffel (stoffel!~quassel@pD9E4272F.dip.t-dialin.net) has quit (Remote host closed the connection) | |
| [18:59:14] | stichnot: | because if mythffmpeg and mythcommflag --rebuild produce different results, then there is probably a bug in the mythcommflag player |
| [18:59:24] | jheizer: | stichnot: you see my msg about the avi file over the weekend? |
| [19:00:48] | stichnot: | jheizer: didn't you report that your avi files are actually mkvs? |
| [19:01:37] | jheizer: | yeah, or xvid. just wasn't sure if you saw it as I was in and out alot. |
| [19:01:53] | jheizer: | Just didn't want you waiting on me |
| [19:02:12] | neufeld: | stichnot: ffmpeg -nostats -i <H264_TRANSPORT_STREAM> -vf showinfo -f null – . Then, my perl script reads the output and uses if ( /.* n:(\d+) .* pos:(\d+) .* iskey:1 .*/ ) { ... $1 = frame number, $2 = file offset ... } |
| [19:02:48] | neufeld: | stichnot: output is on stderr, so I typically stick 2>&1 on the end of the ffmpeg invocation |
| [19:03:35] | stichnot: | jheizer: yeah, thanks. Based on that, I decided to call it fixed in Master, since I suspect there's little chance of getting a (legal) sample to test with. |
| [19:03:42] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [19:03:59] | stichnot: | neufeld: right, sorry, I think you told me that a few days ago. |
| [19:04:19] | neufeld: | stichnot: no problem! |
| [19:04:26] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [19:10:35] | stichnot: | neufeld: IReboot: fyi, somewhere on my to-do list, I have plans to add OSD controls for caption/subtitle sync and drift, in the same spirit as "adjust audio sync". I assume this might be relevant to your subtitle efforts. |
| [19:12:00] | IReboot: | stichnot: The only issue I had left with subtitles was the duration issues (recently fixed) with mythccextractor. |
| [19:12:41] | IReboot: | stichnot: Maybe I am missing the significance of your comment. |
| [19:16:16] | stichnot: | IReboot: I misremembered the issue – looks like there are sometimes audio sync problems, not subtitle sync |
| [19:17:42] | neufeld: | IReboot: it's a kind of edge case. The HD-PVR can reset. For me, it seems to happen when there's a bad signal at the cable provider, the STB's TOSLink output seems to go to an error state. The HD-PVR resets, and when MythTV starts it up again, it starts spitting out packets with PTS reset to zero, but part of the same file. The subtitle code gets the PTS time and uses it to index subtitles, so even though we're |
| [19:17:42] | neufeld: | successfully playing 85 minutes into a movie (barring a 2–3 second glitch), suddenly the subtitles think we're back at the opening credits. |
| [19:18:12] | neufeld: | IReboot: this is when the subtitles are provided in a separate disk file, and are time-stamped, rather than frame-number-stamped. |
| [19:21:14] | IReboot: | stichnot: Infrequently lossless cut videos over ~2hrs will have very slight lip sync drift. I cannot tell if is is due to the keyframe seek values or within mkvmerge or the original recording. At any length it is hardly noticeable but have been identified by more than one person. |
| [19:22:57] | IReboot: | neufeld: I was reading that thread on the mailing lists. |
| [19:25:28] | stichnot: | IReboot: do you know if the audio drift is related to the recorder or video type? |
| [19:26:21] | IReboot: | stichnot: For me it is an HDPVR but I do not remember what the other users who observed the problem was using. |
| [19:26:29] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has joined #mythtv | |
| [19:26:29] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has quit (Changing host) | |
| [19:26:29] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [19:28:58] | IReboot: | stichnot: I have seen on one of my own HDPVR recordings, of close to 4hrs, get lossless cut paired down to 2h10min and not loose lip sync. I was fairly impressed. |
| [19:30:12] | stichnot: | IReboot: if the audio drift is limited to HDPVR recordings, it's likely to be the "occasional contracted frame" problem neufeld was mentioning. |
| [19:30:36] | dijonyummy (dijonyummy!~dijonyumm@64.56.206.254) has joined #mythtv | |
| [19:31:05] | IReboot: | stichnot, neufeld: Thanks for the info I will keep it in mind. |
| [19:33:16] | dijonyummy: | is there an easy script that can query the mythtv db for all tv shows and descriptions? or a way to have mythtv write a file with the description along with the actual mpg or something? |
| [19:33:27] | IReboot: | stichnot: I think a lot of issues users metnion with the HDPVR are due to the cable providers crappy signal. NFL football from both CBS and FOX and the PBS news seem to have endless audio glitches and some video artifacts. |
| [19:55:46] | gregL (gregL!~greg@cpe-74-76-105-205.nycap.res.rr.com) has quit (Ping timeout: 265 seconds) | |
| [20:03:03] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Ping timeout: 260 seconds) | |
| [20:06:47] | jheizer: | My HDPVR works great as long as I use composite audio. Have been afraid to try tos in the last 2+ years. |
| [20:09:19] | bas-t (bas-t!~tycho@52484E89.cm-4-1b.dynamic.ziggo.nl) has joined #mythtv | |
| [20:11:05] | SteveGoodey (SteveGoodey!~steve@host86-148-172-139.range86-148.btcentralplus.com) has quit (Quit: Konversation terminated!) | |
| [20:11:08] | peper03 (peper03!~peper03@port-92-203-45-84.dynamic.qsc.de) has joined #mythtv | |
| [20:15:33] | jheizer_laptop (jheizer_laptop!~jon@c-98-226-220-178.hsd1.il.comcast.net) has joined #mythtv | |
| [20:20:24] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [20:24:05] | SteveGoodey (SteveGoodey!~steve@host86-148-172-139.range86-148.btcentralplus.com) has joined #mythtv | |
| [20:38:40] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [20:43:43] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has joined #mythtv | |
| [20:44:24] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has joined #mythtv | |
| [20:44:25] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has quit (Changing host) | |
| [20:44:25] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [20:47:03] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [20:48:19] | stichnot: | jheizer: I hear you. I've been running my HDPVR with analog audio input and an old firmware version after being burned by too many upgrade attempts. |
| [20:49:28] | jheizer: | stichnot: Yup! Was afraid to upgrade to ubuntu 12.10 so I could get myth .26 from the packages as I didn't want to break the hdpvr blaster |
| [20:49:42] | jheizer: | thankfully the new modules were good to go out of the box |
| [21:07:36] | peper03: | stichnot: Where is the current playing position maintained? I've found some more weirdness with DVDs showing a still frame and playing audio – the audio plays from the beginning but the current position starts at something like 6 seconds. This is on 0.26, so maybe it doesn't occur with your fixes in master but I'd like to check. |
| [21:17:20] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has joined #mythtv | |
| [21:17:21] | zombor (zombor!~zombor__@50-73-122-41-ip-static.hfc.comcastbusiness.net) has quit (Changing host) | |
| [21:17:21] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [21:17:24] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [21:20:00] | peper03: | stichnot: Never mind. Found the source. Just got to work out what the problem is :) |
| [21:37:45] | zombor (zombor!~zombor__@kohana/developer/zombor) has joined #mythtv | |
| [21:37:58] | zombor (zombor!~zombor__@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [21:58:08] | SteveGoodey (SteveGoodey!~steve@host86-148-172-139.range86-148.btcentralplus.com) has quit (Quit: Konversation terminated!) | |
| [21:59:54] | dmfrey (dmfrey!~dmfrey@webdefence.cluster-h.websense.net) has quit (Quit: Ex-Chat) | |
| [22:01:28] | Guest3256 (Guest3256!~quassel@75-161-183-113.mpls.qwest.net) has joined #mythtv | |
| [22:03:19] | stichnot: | peper03: FWIW, the new code in Master tries to use the duration map in recordedseek/filemarkup, but if there is none, it is supposed to fall back to the original method. This would be true for DVD playback. |
| [22:05:22] | stichnot: | peper03: Do you happen to know why the player cannot skip/ff through a still frame? I think I tried to look into that a couple years ago, but eventually gave up. |
| [22:06:31] | peper03: | stichnot: The position problem *might* be due to buffering. In this case, there's really only audio data coming in. We seem to read the first six seconds worth of data very quickly. Thinking about it, changing the volume during playback seems to take about 5 or 6 seconds before it has any effect, just like radio via DVB-S. |
| [22:08:27] | peper03: | stichnot: In general, I think it should be possible to skip through a still frame. There are probably about three scenarios – still frame with no audio, still frame with audio, normal video stream that appears to be a still frame because nothing changes. Do you have a specific example? |
| [22:09:23] | NightMonkey_ (NightMonkey_!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [22:11:51] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Ping timeout: 260 seconds) | |
| [22:12:34] | peper03: | I've just pulled the lastest master and done a distclean. DVD playback is broken. Stills screw everything up. Jumping from an opening still to the root menu gives me lots of 'Video is x.xxxx frames behind audio' errors and horrible playback. |
| [22:13:46] | peper03: | VDPAU/OpenGL doesn't seem to make any difference |
| [22:14:48] | Goga777 (Goga777!~Goga777@128-71-222-125.broadband.corbina.ru) has quit (Remote host closed the connection) | |
| [22:16:20] | gregL (gregL!~greg@cpe-74-76-105-205.nycap.res.rr.com) has joined #mythtv | |
| [22:16:37] | peper03: | Trying to play a DVD that jumps straight to a still menu shows the menu for 5 seconds and then dumps me back in the MythVideo menu with no (as yet) visible reason. |
| [22:16:44] | jheizer_laptop (jheizer_laptop!~jon@c-98-226-220-178.hsd1.il.comcast.net) has quit (Ping timeout: 260 seconds) | |
| [22:20:39] | stichnot: | peper03: I will try to come up with a small test example. For most of my DVDs, I just rip the main title, but there are a few compilations where that's not possible, and this is where I get the "FBI Warning" and such. |
| [22:21:25] | stichnot: | For most of those, I can press EXIT and it takes me to the title menu, but there are a couple where I have to suffer because for whatever reason EXIT instead exits the player. |
| [22:21:41] | peper03: | stichnot: Can you jump to the root or title menu via the OSD menu? |
| [22:22:56] | stichnot: | peper03: IIRC, selecting title or root menu goes to the root menu, which restarts playback of that still frame. |
| [22:24:12] | stichnot: | I will see if I can "dd" a small initial section of the .iso which demonstrates the problem, and if not, I can just upload the whole thing for a short time. |
| [22:24:40] | peper03: | stichnot: Ok, that's possible if that still has been marked as the title menu. |
| [22:25:28] | Korny1: | So the website from Mythtv android Frontend said to come to this channel, but I'm debating asking in here or in myth_users :/ |
| [22:26:40] | peper03: | stichnot: It's possible that still has invisible buttons. That would cause problems trying to skip over the still. That said, I've not looked too deeply into skipping and key handling and whatnot so there may well be some other cases that could be handled. |
| [22:33:20] | Steve-Goodey (Steve-Goodey!~steve@host86-148-172-139.range86-148.btcentralplus.com) has quit (Quit: Konversation terminated!) | |
| [22:37:04] | jheizer_laptop (jheizer_laptop!~jon@c-98-226-220-178.hsd1.il.comcast.net) has joined #mythtv | |
| [22:47:53] | dekarl: | If I understand correctly the default LNB these days is of the Universal/Astra kind so it would be nice to make that the default in MythTV, too. Would this patch achieve that? http://paste.ubuntu.com/1507954/ |
| [23:01:04] | natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has quit (Ping timeout: 246 seconds) | |
| [23:02:17] | peper03: | stichnot: How are you trying to skip through the still frame? JUMPFFWD (PgDown) or SEEKFFWD (Right)? JUMPFFWD certainly seems to work for me. |
| [23:08:38] | stichnot: | peper03: I added some logging, and it turns out this opening is not a still frame, it's actually a 20-second menu with no buttons, where the first 10 seconds just looks like a still frame |
| [23:09:07] | stichnot: | the combination of in_still=false and in_menu=true causes TV::DVDJumpForward() to do nothing |
| [23:10:51] | peper03: | stichnot: Did you try my patch from #11292? That should fix it. |
| [23:11:57] | stichnot: | let me look |
| [23:14:28] | amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has quit (Quit: Konversation terminated!) | |
| [23:17:39] | stichnot: | peper03: that patch does enable skipping to the next title (in this case I was at the last chapter of the title) |
| [23:18:37] | stichnot: | but interestingly, other kinds of seeking, pausing, etc. in a menu are explicitly disabled in the tv_play.cpp code. Do you know of any good reason for those to be disabled? |
| [23:22:53] | peper03: | stichnot: I think it normally doesn't make much sense to skip or pause a menu. The question is probably "What is a menu?". I think at the moment, it's defined as anything in the video manager domain. That can still be video, though. That's why I added the check for the number of buttons. If there are no buttons, I can't think of any reason to treat is as something special. |
| [23:24:28] | peper03: | I suppose you *could* pause a proper menu but it was probably easier just to disable it than deal with users single stepping frames instead of navigating the menu. |
| [23:31:19] | stichnot: | Most likely the code was there originally to imitate the behavior of standard DVD players. But I wonder if there is a better reason for restricting the behavior. |
| [23:34:05] | peper03: | Possibly but to do that properly, we should be following the "user prohibited operations" flags. That we don't is one of the attractions of Myth for me (among others). It can cause problems though as it makes it possible to skip over code that is required. |
| [23:34:22] | stichnot: | For the usual seek/ffwd/rew keys, DVDRingBuffer::HandleAction() would take precedence and handle those keys, but only if NumMenuButtons()>0. If there are no buttons, it seems reasonable to let seeking, pause, etc. go on as usual. |
| [23:34:54] | stichnot: | sure, the uops are ridiculous and no one in their right mind would implement them if given the option |
| [23:38:08] | peper03: | Yeah, the problem is just that we can never guarantee that you can always ignore them and not have problems. stuartm suggested following them by default to avoid people complaining that DVD 'X' doesn't play. There would be a setting to ignore them though. As long as I can set Myth to ignore them, I can live with that idea. |
| [23:39:58] | stichnot: | hmm, how could skipping around in the same title cause problems? can there be VM commands at chapter boundaries that we might miss? |
| [23:43:21] | peper03: | Yes. I don't think it happens often but you could author a DVD that forced you to watch several minutes of trailers (not unheard of) before getting to the main menu. Set some register at the end of the last trailer to indicate that all the trailers have been watched and then evaluate that register later. If you really wanted to be nasty, you could cause playback to always jump back to the trailers until they've been watched in full ( |
| [23:43:22] | peper03: | or at least, jumped over rather than out of). |
| [23:44:07] | peper03: | Whether it's in a studio's interest to do that is another matter but I wouldn't be *that* surprised. |
| [23:44:11] | bas-t (bas-t!~tycho@52484E89.cm-4-1b.dynamic.ziggo.nl) has quit (Quit: Quit) | |
| [23:44:49] | peper03: | stuartm said he had a DVD that wouldn't play properly if he jumped straight to the menu. |
| [23:48:13] | peper03: | stichnot: Are you having any DVD playback problems on master? It seems to be related to stills. I've not got any local changes at the moment so it appears something is broken. |
| [23:48:20] | jheizer_laptop (jheizer_laptop!~jon@c-98-226-220-178.hsd1.il.comcast.net) has quit (Ping timeout: 248 seconds) | |
| [23:48:38] | stichnot: | I can believe that, but let's say I press "1 >" to jump to one minute before the end of the title, followed by RIGHT RIGHT which skips 60 seconds, and then I drop into the next title. (which of course might be the second of 20 trailers/ads) Is it possible to have lost any of the VM commands? |
| [23:49:02] | stichnot: | peper03: I do seem to be seeing a problem in master |
| [23:49:32] | stichnot: | After playing the 20-second FBI Warning title, it drops out to the mythvideo menu |
| [23:49:47] | stichnot: | but because of my settings, it leaves a bookmark |
| [23:50:11] | stichnot: | and when I start playing again, and select "Play from bookmark", it continues where it should have left off. |
| [23:50:58] | stichnot: | so it looks like something is causing it to do a clean exit of playback |
| [23:51:06] | peper03: | Could you try the 'palettetest.iso' image from stuartm's repository. That jumps straight into a still menu. It shows that for 5 seconds (navigation is not possible) and then drops back out again. |
| [23:54:37] | stichnot: | It displays the still frame for 5 seconds and then exits mythvideo. Is that expected? |
| [23:54:55] | peper03: | stichnot: That's what I'm seeing, but not what's expected. |
| [23:55:39] | stichnot: | ok, I see different behavior in vlc |
| [23:58:02] | peper03: | 0.26 doesn't have that problem. The last time I used master (two or three weeks ago) I didn't notice any problems. |
IRC Logs collected by
BeirdoBot.
Please use the above link to report any bugs.