| Thursday, August 30th, 2018, 00:04 UTC | ||
| [00:04:13] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has joined #mythtv | |
| [00:08:46] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has quit (Ping timeout: 250 seconds) | |
| [00:30:51] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has joined #mythtv | |
| [00:35:04] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has quit (Ping timeout: 240 seconds) | |
| [00:40:14] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has joined #mythtv | |
| [00:44:44] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has quit (Ping timeout: 250 seconds) | |
| [01:50:56] | pppingme (pppingme!~pppingme@unaffiliated/pppingme) has joined #mythtv | |
| [01:52:56] | gigem: | peterbennett: Are you sure the stuttering with vaapi2 isn't similar to that with mediacode before the framereate was corrected. Even if not, the frame drop prediction should still handle when the frame rate and refresh rates don't exactly match. That's the whole reason it exists. |
| [03:02:33] | peterbennett: | gigem: not sure. I need to look at it a bit more I am wondering how it normally should work. With software ffmpeg decoding speedup seems fine but maybe there is code to change the deinterlace during speedup. |
| [03:05:00] | peterbennett: | gigem: The framerate correction code that I added for mediacodec is still working as far as I can see. |
| [03:05:14] | peterbennett: | with vaapi2 |
| [03:12:33] | peterbennett: | gigem: thanks for the tip- I will look at the frame rate drop prediction. I have seen that in the code and not really known what it is for. Perhaps it is not picking up the doubled frame rate correctly. |
| [03:18:53] | Fear (Fear!~fear@unaffiliated/warlord) has joined #mythtv | |
| [04:22:05] | ShapeShifter499 (ShapeShifter499!~ShapeShif@unaffiliated/shapeshifter499) has joined #mythtv | |
| [06:36:33] | ShapeShifter499 (ShapeShifter499!~ShapeShif@unaffiliated/shapeshifter499) has quit (Quit: ZNC 1.7.1 - https://znc.in) | |
| [06:36:57] | ShapeShifter499 (ShapeShifter499!~ShapeShif@unaffiliated/shapeshifter499) has joined #mythtv | |
| [06:46:04] | Steve-Goodey (Steve-Goodey!~steve@host86-189-217-242.range86-189.btcentralplus.com) has joined #mythtv | |
| [06:46:04] | Mode for #mythtv by ChanServ!ChanServ@services. : +v Steve-Goodey | |
| [07:57:54] | Merlin83b (Merlin83b!~Daniel@2a00:1ee0:3:1337:101e:cad8:a6be:774e) has joined #mythtv | |
| [07:58:06] | willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has joined #mythtv | |
| [08:10:19] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has joined #mythtv | |
| [08:10:19] | markspieth22 (markspieth22!~yaaic@mythtv/developer/markspieth) has joined #mythtv | |
| [08:10:19] | markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has quit (Read error: Connection reset by peer) | |
| [08:10:19] | Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth22 | |
| [08:10:19] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has quit (Ping timeout: 250 seconds) | |
| [08:29:11] | rmeden (rmeden!~Robert@107-131-102-250.lightspeed.rcsntx.sbcglobal.net) has quit (Read error: Connection reset by peer) | |
| [08:29:47] | rmeden (rmeden!~Robert@107-131-102-250.lightspeed.rcsntx.sbcglobal.net) has joined #mythtv | |
| [08:31:35] | Merlin83b2 (Merlin83b2!~Daniel@thirtyfoursp.plus.com) has joined #mythtv | |
| [08:33:34] | Merlin83b (Merlin83b!~Daniel@2a00:1ee0:3:1337:101e:cad8:a6be:774e) has quit (Ping timeout: 240 seconds) | |
| [08:34:39] | dan__ (dan__!~Daniel@2a00:1ee0:3:1337:101e:cad8:a6be:774e) has joined #mythtv | |
| [08:36:04] | Merlin83b2 (Merlin83b2!~Daniel@thirtyfoursp.plus.com) has quit (Ping timeout: 240 seconds) | |
| [08:36:21] | markspieth22 (markspieth22!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds) | |
| [08:41:03] | markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has joined #mythtv | |
| [08:41:03] | Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth2 | |
| [08:48:07] | Steve-Goodey (Steve-Goodey!~steve@host86-189-217-242.range86-189.btcentralplus.com) has quit (Quit: Konversation terminated!) | |
| [09:09:34] | markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has quit (Read error: Connection reset by peer) | |
| [09:09:37] | markspieth22 (markspieth22!~yaaic@mythtv/developer/markspieth) has joined #mythtv | |
| [09:09:37] | Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth22 | |
| [09:40:01] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has joined #mythtv | |
| [09:44:04] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has quit (Ping timeout: 240 seconds) | |
| [10:41:10] | Steve-Goodey (Steve-Goodey!~steve@2a00:23c5:7da3:4501:4a51:b7ff:fe21:2979) has joined #mythtv | |
| [10:41:11] | Mode for #mythtv by ChanServ!ChanServ@services. : +v Steve-Goodey | |
| [11:16:31] | markspieth22 (markspieth22!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds) | |
| [11:18:55] | markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has joined #mythtv | |
| [11:18:55] | Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth2 | |
| [11:45:07] | markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has quit (Read error: Connection reset by peer) | |
| [11:45:13] | markspieth22 (markspieth22!~yaaic@mythtv/developer/markspieth) has joined #mythtv | |
| [11:45:13] | Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth22 | |
| [12:04:01] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has joined #mythtv | |
| [12:08:32] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has quit (Ping timeout: 250 seconds) | |
| [12:21:06] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has joined #mythtv | |
| [12:25:34] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has quit (Ping timeout: 240 seconds) | |
| [12:38:29] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has joined #mythtv | |
| [12:42:46] | dragonian (dragonian!~dragonian@2604:6000:100e:8165:5962:fd16:587e:6f61) has quit (Ping timeout: 250 seconds) | |
| [12:59:54] | jheizer (jheizer!~jheizer@2601:246:8280:557:5c84:8ed6:9818:ef68) has quit (Quit: Leaving) | |
| [13:00:16] | jheizer (jheizer!~jheizer@2601:246:8280:557:5c84:8ed6:9818:ef68) has joined #mythtv | |
| [13:45:05] | markspieth22 (markspieth22!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 256 seconds) | |
| [13:46:21] | markspieth (markspieth!~markspiet@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds) | |
| [13:48:42] | markspieth (markspieth!~markspiet@mythtv/developer/markspieth) has joined #mythtv | |
| [13:48:42] | Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth | |
| [13:50:48] | jheizer (jheizer!~jheizer@2601:246:8280:557:5c84:8ed6:9818:ef68) has quit (Ping timeout: 250 seconds) | |
| [13:50:57] | jheizer_ (jheizer_!~jheizer@c-73-51-93-177.hsd1.il.comcast.net) has joined #mythtv | |
| [14:00:14] | markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has joined #mythtv | |
| [14:00:14] | Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth2 | |
| [14:37:30] | gigem: | peterbennett: Also, if both 1x and 2x deinterlacers are supported, 1x is supposed to be used if the sped up frame rate is higher than the refresh rate. Frame drop prediction then kicks in if the new frame rate is still higher then the refresh rate. |
| [14:39:35] | peterbennett: | Yes – switching deinterlacers midstream may be problematic |
| [14:41:11] | peterbennett: | The 2x deinterlacer is extremely smooth, 1x not so much. |
| [14:55:47] | gigem: | peterbennett: On what version is your latest vaapi2 patch based? With current master, I got one hunk failure in avformatdecoder.cpp and many, many with line offsets in various other files. It sure looks like it is based on a version I don't have nor recognize. |
| [14:56:10] | peterbennett: | Current master. |
| [14:56:19] | peterbennett: | strange... |
| [14:56:35] | ijc (ijc!~ijc@46.43.0.16) has left #mythtv () | |
| [14:57:33] | gigem: | I have 296fadf6. What do you have? |
| [14:57:39] | peterbennett: | Somehow I am behind master by 49 commits ! |
| [14:57:59] | peterbennett: | has there been a huge commit recently ? |
| [14:58:23] | gigem: | David Hampton was very busy recently fixing cppcheck issues. |
| [14:58:41] | peterbennett: | oh dear |
| [14:59:09] | peterbennett: | I am on e5c0d0ad883865fc5e7c2a6d1fff6d0c0603fcaa |
| [14:59:14] | gigem: | I don't think he changed anything relating to the patch failure, though. Let me re-run it again. |
| [15:00:34] | peterbennett: | I guess I need to fix my patch to work with latest |
| [15:03:49] | gigem: | Yes, it looks like hampton's recent changes overlap with yours. There is one failed hunk, 1 successful hunk with fuzz and many successful hunks with innocous line offsets. |
| [15:04:18] | peterbennett: | ok thanks I will get it fixed |
| [15:05:39] | peterbennett: | I see the avsync_predictor logic – it uses frame_interval which does take into account the frame doubling. |
| [15:18:24] | peterbennett: | I write messy code than hampton cleans it up :) |
| [15:26:50] | qew (qew!~qew@190.74.169.232) has joined #mythtv | |
| [15:27:02] | peterbennett: | s/than/then/ |
| [15:27:54] | qew (qew!~qew@190.74.169.232) has quit (Killed (Sigyn (Spam is off topic on freenode.))) | |
| [15:28:44] | hampton: | :-) |
| [15:28:55] | hampton: | That's me... the code janitor. |
| [15:29:57] | peterbennett: | hampton: Seriously though I should check compile warnings when I make changes. I only look at them when there are errors. |
| [15:30:56] | peterbennett: | Probably a habit because there were traditionally huge numbers of pre existing warnings |
| [15:31:44] | hampton: | I looked at the warning counts for 29 and master yesterday. They are ~950 and ~40. |
| [15:32:02] | peterbennett: | Good job! |
| [15:32:26] | peterbennett: | MythPopupBox deprecated – what is with that one? |
| [15:32:58] | hampton: | Its taken me a year. The remaining warnings are in the deprecated dialogs, ffmpeg, and a couple of random ones. |
| [15:33:23] | hampton: | That's from the UI change a while back from polled to callback dialogs. |
| [15:33:51] | peterbennett: | I did fix almost all the ffmpeg ones but then the next ffmpeg version deprecated a new bunch |
| [15:34:14] | hampton: | I guess it wasn't quite completed, and the items that still need to be changed were marked as deprecated. |
| [15:36:12] | peterbennett: | with ffmpeg it is an ongoing process. |
| [15:36:17] | hampton: | I think they're mostly in little used drivers. See https://code.mythtv.org/buildbot/builders/mas . . . s%20%2849%29 |
| [15:37:09] | hampton: | That's a concise list of the remaining warnings. |
| [15:38:14] | peterbennett: | In order to add vaapi2 support I use some of the ffmpage sample code and then find it is using deprecated functions. |
| [15:38:19] | peterbennett: | ffmpeg |
| [15:38:23] | hampton: | :-( |
| [15:42:57] | peterbennett: | Looking thru all the warnings in avformatdecoder reminds me why I do not check them. |
| [15:44:08] | peterbennett: | I need a way to flag those that are in new code so that at least I am not introducing new ones |
| [15:48:09] | hampton: | BTW, thank you for tackling ffmpeg. I can handle general coding, networking, gui, etc, etc, but when it comes to the realtime decoding path I'd happy to not have to learn the ins and outs of that code. |
| [15:49:50] | TingPing4 (TingPing4!~TingPing@217.217.232.33.dyn.user.ono.com) has joined #mythtv | |
| [15:50:25] | TingPing4 (TingPing4!~TingPing@217.217.232.33.dyn.user.ono.com) has quit (Killed (Sigyn (Spam is off topic on freenode.))) | |
| [15:52:28] | peterbennett: | hampton: did you see that we created our own ffmpeg repo in mythtv and I merge into there before copying into mythtv repo |
| [15:58:15] | peterbennett: | gigem: I have attached an updated patch for latest master |
| [16:12:21] | brfranse- (brfranse-!~brfransen@47.42.27.251) has joined #mythtv | |
| [16:15:10] | brfransen (brfransen!~brfransen@47.42.27.251) has quit (Ping timeout: 252 seconds) | |
| [16:19:07] | hampton: | I missed that. |
| [16:35:58] | peterbennett: | hampton: So if you update anything in external/FFmpeg you should also update it in the FFmpeg repo, master as well as release/4 branches. |
| [16:37:03] | peterbennett: | I added to the sync instructions that before syncing a new FFmpeg version to check for any updates made in mythtv FFmpeg and make them in the FFmpeg repo if not already done. |
| [16:37:54] | gigem: | peterbennett: Thanks. I'll grab the new patch and build with it shortly. |
| [16:41:24] | Sousapro6 (Sousapro6!~Sousapro@188.246.245.206) has joined #mythtv | |
| [16:41:42] | gigem: | hampton: IMHO, any warning fixes in external/FFmpeg should go to the ffmpeg project. They might or might not be receptive. Some projects care about keeping clean compiles and some don't. I don't know which camp they are in. We shouldn't make any changes that we don't absolutely have to. |
| [16:42:05] | Sousapro6 (Sousapro6!~Sousapro@188.246.245.206) has quit (Remote host closed the connection) | |
| [16:45:28] | em (em!~em@118-104-46-205.aichiwest1.commufa.jp) has joined #mythtv | |
| [16:46:08] | ktechmidas26 (ktechmidas26!~ktechmida@host-89-229-103-107.dynamic.mm.pl) has joined #mythtv | |
| [16:49:39] | ktechmidas26 (ktechmidas26!~ktechmida@host-89-229-103-107.dynamic.mm.pl) has quit (Remote host closed the connection) | |
| [16:50:11] | em (em!~em@118-104-46-205.aichiwest1.commufa.jp) has quit (Remote host closed the connection) | |
| [16:50:46] | peterbennett: | gigem: The predict drop frame was not working because the frame doubling does not start immediately, a few frames are not doubled at the start – fixed that, now it is better but still not good |
| [16:51:58] | hampton: | There are only two warnings in FFmpeg. They're in the same assembly file, and I wouldn't touch them with a 100' pole. I'm not sure of the writer's intent, and there's a weird macro implementation issue such that the redefined macro assembly appears in the object file between instances of the original macro assembly. |
| [16:52:08] | peterbennett: | seems to alternate between too slow and too fast. |
| [16:53:09] | hampton: | My solution was to ask stuarta to add an extra step to the buildbots to compile the external directory separately from the MythTV code. |
| [16:53:10] | peterbennett: | hampton: I disabled warnings in the ffmpeg compile because there were hundreds of the same warning about redefinitions. |
| [16:53:42] | hampton: | ok, so there's more than two warnings. :-) |
| [16:54:05] | peterbennett: | yes it was flooding with warnings. |
| [16:55:03] | hampton: | I figure that splitting out the external directory compiles gives us a warning count that we're not trying to fix, and a warning count that we should try to fix. |
| [16:57:00] | dan__ (dan__!~Daniel@2a00:1ee0:3:1337:101e:cad8:a6be:774e) has quit (Quit: Leaving) | |
| [17:19:40] | willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has quit (Quit: Do your hobbies) | |
| [17:19:58] | gigem: | taylorr: ^^^ I think you were the original author of the smooth timestretch. Can you help peterbennett? |
| [17:24:33] | theaetetus (theaetetus!~theaetetu@95.6.20.16) has joined #mythtv | |
| [17:26:11] | hampton: | someone will need to voice him so he can respond |
| [17:27:46] | gigem: | I believe I have the privileges to do it but don't remember the steps. |
| [17:28:09] | hampton: | I think its just "/voice taylorr" |
| [17:28:49] | theaetetus (theaetetus!~theaetetu@95.6.20.16) has quit (Remote host closed the connection) | |
| [17:28:51] | gigem: | I think there's another step to become a ChanOp or something like that. I'm looking. |
| [17:29:15] | gigem: | Yep, need to be a chanop to /voice someone. |
| [17:30:22] | Mode for #mythtv by ChanServ!ChanServ@services. : +o gigem | |
| [17:30:29] | Mode for #mythtv by gigem!~david@mythtv/developer/gigem : +v taylorr | |
| [17:30:45] | gigem: | I think that did it. |
| [17:33:45] | Mode for #mythtv by ChanServ!ChanServ@services. : -o gigem | |
| [17:37:15] | Xenogenesis6 (Xenogenesis6!~Xenogenes@85.91.205.246) has joined #mythtv | |
| [17:41:30] | Xenogenesis6 (Xenogenesis6!~Xenogenes@85.91.205.246) has quit (Remote host closed the connection) | |
| [17:51:25] | kaniini_12 (kaniini_12!~kaniini_@60-199-18-34.static.tfn.net.tw) has joined #mythtv | |
| [17:53:43] | jpabq (jpabq!~quassel@97-123-177-62.albq.qwest.net) has quit (Changing host) | |
| [17:53:43] | jpabq (jpabq!~quassel@mythtv/developer/jpabq) has joined #mythtv | |
| [17:53:44] | Mode for #mythtv by ChanServ!ChanServ@services. : +v jpabq | |
| [17:54:12] | jpabq: | gigem: Did you see my last message? " it was markspieth that got timestretch smooth." |
| [17:54:48] | jpabq: | Your talk about "voice" made me realize I don't know what is going on, here, regarding permissions. |
| [17:56:21] | hampton: | Lots of irc spam, so most people were unvoiced. I think the regular speakers were given a permanent voice. |
| [17:56:30] | kaniini_12 (kaniini_12!~kaniini_@60-199-18-34.static.tfn.net.tw) has quit (Ping timeout: 252 seconds) | |
| [17:56:35] | peterbennett: | I will first take a closer look and try to understand what is going on. Maybe I have made a stupid mistake. |
| [17:58:38] | gigem: | jpabq: Was that on irc or email? Either way, I easily forget minutia like that when the need for it is fleeting. I check the git annotate output and taylorr looked like the most frequent person to touch that code. |
| [18:00:39] | gigem: | jpabq: As hampton said, spam. stuarta (where is he, btw), turned on moderation a couple of weeks ago when the spam was terrible. He also gave a few of us who were around at the time the priviliegs to /voice users. /voice means their posts don't need to be moderated. |
| [18:01:37] | hampton: | stuarta is on vacation. Back next week. |
| [18:01:52] | hampton: | Err, holiday. :-) |
| [18:03:14] | gigem: | Ah, yes, those Europeans all take August off. :) |
| [18:04:17] | peterbennett: | He is on "leave"? |
| [18:04:25] | hampton: | I wouldn't mind a vacation from the 100+ heat index outside. |
| [18:56:49] | cooldude11 (cooldude11!~cooldude@118.136.149.92) has joined #mythtv | |
| [19:01:32] | cooldude11 (cooldude11!~cooldude@118.136.149.92) has quit (Remote host closed the connection) | |
| [19:10:25] | natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has joined #mythtv | |
| [19:10:25] | Mode for #mythtv by ChanServ!ChanServ@services. : +v natanojl | |
| [20:05:56] | justache7 (justache7!~justache@41.43.156.120) has joined #mythtv | |
| [20:06:01] | justache7 (justache7!~justache@41.43.156.120) has quit (Remote host closed the connection) | |
| [20:11:00] | Hoolootwo (Hoolootwo!~Hoolootwo@92.176.240.181) has joined #mythtv | |
| [20:15:42] | Hoolootwo (Hoolootwo!~Hoolootwo@92.176.240.181) has quit (Remote host closed the connection) | |
| [21:02:41] | Steve-Goodey (Steve-Goodey!~steve@2a00:23c5:7da3:4501:4a51:b7ff:fe21:2979) has quit (Quit: Konversation terminated!) | |
| [21:42:16] | markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has quit (Remote host closed the connection) | |
| [22:05:49] | xunplini5 (xunplini5!~xunplini@ppp91-79-218-22.pppoe.mtu-net.ru) has joined #mythtv | |
| [22:06:02] | RokLobsta (RokLobsta!~Dirkka@119-18-15-76.cust.aussiebb.net) has joined #mythtv | |
| [22:08:20] | xunplini5 (xunplini5!~xunplini@ppp91-79-218-22.pppoe.mtu-net.ru) has quit (Remote host closed the connection) | |
| [22:09:35] | peterbennett (peterbennett!~Peter_Ben@mythtv/developer/peterbennett) has quit (Quit: WeeChat 1.6) | |
| [22:10:44] | markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has joined #mythtv | |
| [22:10:44] | Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth2 | |
| [22:11:09] | natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has quit (Ping timeout: 252 seconds) | |
| [22:11:30] | peterbennett (peterbennett!~Peter_Ben@c-75-67-203-208.hsd1.ma.comcast.net) has joined #mythtv | |
| [22:11:31] | peterbennett (peterbennett!~Peter_Ben@c-75-67-203-208.hsd1.ma.comcast.net) has quit (Changing host) | |
| [22:11:31] | peterbennett (peterbennett!~Peter_Ben@mythtv/developer/peterbennett) has joined #mythtv | |
| [22:11:31] | Mode for #mythtv by ChanServ!ChanServ@services. : +v peterbennett | |
| [22:15:23] | markspieth: | peterbennett: I did the smooth avsync a while ago. There was this logic in there (inherited) for handling deinterlaced fields 2 in a pass. I would have to look at it again to advise. All the logic is in MythPlayer::AVSync |
| [22:20:14] | SteveGoodey (SteveGoodey!~steve@2a00:23c5:7da3:4501:7a24:afff:fe9d:c233) has quit (Quit: Konversation terminated!) | |
| [22:20:25] | markspieth: | peterbennett: A quick look shows m_double_framerate is what selects the 1st deinterlaced field to display/process. This should be turned off if something else is handling deinterlacing and providing a progresive frame stream. |
| [22:22:28] | gigem: | peterbennett: I occaionally change deinterlacers via the osd if the current one can't keep up with the desired timestretch rate. It's also extremely useful when comapring deinterlacers. Having to go back to settings to change display profiles would be a major pain. |
| [22:39:25] | peterbennett: | gigem: markspieth: I tested the speedup and there seem to be the same issues with the current master, using standard decode and opengl |
| [22:39:33] | markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has quit (Ping timeout: 244 seconds) | |
| [22:41:04] | peterbennett: | gigem: markspieth: I set my monitor to 30 Hz and speeded up to 2x. The playback is jerky and the log shows it getting alternately behind and ahead |
| [22:41:42] | markspieth2 (markspieth2!~yaaic@mythtv/developer/markspieth) has joined #mythtv | |
| [22:41:42] | Mode for #mythtv by ChanServ!ChanServ@services. : +v markspieth2 | |
| [22:41:45] | peterbennett: | This is the same thing as I observer at 60Hz monitor with frame doubling deinterlace. |
| [22:43:01] | peterbennett: | Something is amiss with the speedup sync when you get to 1.5 or 2x speedup and your monitor does not support double rate. |
| [22:43:43] | peterbennett: | So to get vaapi2 working as well as current, all I really need to do is get the 1x deinterlacer to run when going into speeudp. |
| [22:44:40] | RokLobsta (RokLobsta!~Dirkka@119-18-15-76.cust.aussiebb.net) has left #mythtv () | |
| [22:44:44] | RokLobsta (RokLobsta!~Dirkka@119-18-15-76.cust.aussiebb.net) has joined #mythtv | |
| [22:46:27] | peterbennett: | gigem: I had to add a whole new feature of deinterlacers that work with the decoder instead of with the render, so I will need to hook up features like the menu selection and the switching to 1x when speeding up. |
| [22:49:21] | Mode for #mythtv by ChanServ!ChanServ@services. : +o peterbennett | |
| [22:50:10] | Mode for #mythtv by peterbennett!~Peter_Ben@mythtv/developer/peterbennett : +v RokLobsta | |
| [22:50:28] | peterbennett: | RokLobsta: You can post now |
| [22:51:07] | peterbennett: | RokLobsta: #13311 |
| [22:51:07] | ** MythLogBot https://code.mythtv.org/trac/ticket/13311 ** | |
| [22:51:22] | peterbennett: | VAAPI2 is nearly read |
| [22:51:25] | peterbennett: | ready |
| [22:53:07] | RokLobsta: | thanks |
| [22:54:09] | peterbennett: | RokLobsta: You can test the VAAPI2 code, it is still preliminary and has some cleanup to be done, speedup need to be fixed. |
| [22:54:12] | RokLobsta: | vaapi works very well now for mpeg2/h.264/h.265 but need deinterlacing for some broadcast tv. |
| [22:54:47] | peterbennett: | The new one works with hardware deinterlace and is very smooth |
| [22:55:20] | peterbennett: | Also fixes the bug where vaapi playback crashes if the picture resolution changes. |
| [22:55:43] | RokLobsta: | i am using ... mythbuntu 16.04 with i965-va-driver:amd64 1.7.0–1 and myth 29.1 + fixes |
| [22:55:54] | RokLobsta: | what needs updating? |
| [22:56:10] | peterbennett: | I think the patch will only work with master. |
| [22:56:37] | peterbennett: | But you can in fact use a master frontend with a v29 backend |
| [22:56:47] | RokLobsta: | yep |
| [22:57:49] | peterbennett: | You can get the patch and build if you like but be aware it is still an early version and some things may not work |
| [22:57:53] | gigem: | peterbennett: Yes, I know. Deinterlacing in the decoder is a whole new animal requires breaking some very, old assumptions. Mediacodec was the first step with it's single deinterlacer(*) and now vaapi2 is the next step with multiple deinterlacers. |
| [22:58:42] | peterbennett: | gigem: Medicodec does not allow us to turn off or change the deinterlacer so it is simpler. |
| [22:58:48] | gigem: | (*)Maybe we can then add the undocumented method Aman mentioned to change deinterlacers on the Shiled. Maybe that will get me that last 10–20% of timestretch I want on 1080i. :) |
| [22:59:38] | gigem: | BTW, I still have hopes that using surface instead of opengl will do that too. I'm not making any progress on it, though. I keep getting sidetracked. |
| [22:59:57] | peterbennett: | gigem: I suspect that on the shild it does not use the DRM sync method and thus does not encounter the problems I am having with the predict avsync |
| [23:00:42] | gigem: | peterbennett: Could be. |
| [23:00:45] | RokLobsta: | hmmm, no ppa with vaapi 2 for 16.04 |
| [23:00:59] | peterbennett: | I am a bit disillusioned with shiled after the mpeg2 pixellation NBC problem that won't be fixed. |
| [23:01:07] | peterbennett: | *shield |
| [23:01:44] | gigem: | peterbennett: To clarify, I believe Aman said the Shield does have two or 3 deinterlacers, but requires a Shiled specific methog to change it. |
| [23:01:46] | peterbennett: | RokLobsta: It is not committed yet.\ |
| [23:02:07] | peterbennett: | gigem: yes I remember |
| [23:02:55] | peterbennett: | RokLobsta: To get it you have to pull the patch and compile |
| [23:03:00] | RokLobsta: | ohoh ok |
| [23:03:06] | RokLobsta: | i might wait a bit then. |
| [23:03:08] | gigem: | peterbennett: As am I, except its the ABC affiliate for me. I need to follow up with Aman to get the specifics so I can relay to my local station. I don't expect them to change just for me, but they should at least be made aware of the issue. |
| [23:03:24] | RokLobsta: | can't destroy working myth install. i'll be out in the street otherwise. |
| [23:03:47] | ** gigem leaves to go home ** | |
| [23:03:55] | ** peterbennett too ** | |
IRC Logs collected by
BeirdoBot.
Please use the above link to report any bugs.