:: #mythtv

Daily chat history

Current users (73):

alan`, aloril, Anssi, anykey_, brfransen, cesman, Chutt, clever, coling, Cougar, danielk22, dblain, dekarl1, ElmerFudd, fetzerch, ghoti, Gibby_, gregL, GreyFoxx, IReboot, J-e-f-f-A, jams, jarle, jarryd, jhall_, jheizer, joe___, joki, jpabq, jpabq_, jpharvey, jst, jwhite, kc, Kevin`, knightr, kormoc, kwmonroe, laga, len, mrand, MythBuild, MythLogBot, nephyrin, neufeld, NightMonkey, Peitolm, peper03, Peps, petefunk, poptix, purserj, rsiebert_, Seeker`_, seld, Sharky-Sleep, skd5aner, sl1ce, SmallR2002_, sphery, sraue, stuarta, stuartm, superm1, tgm4883, Tobbe5178, toeb, tonsofpcs, tris, wahrhaft, wolfgang1, XDS2010, _charly_
Friday, March 15th, 2013, 00:14 UTC
[00:14:04] Wolfgang (Wolfgang! has quit (Quit: Wolfgang)
[00:57:06] jheizer_ (jheizer_! has quit (Ping timeout: 264 seconds)
[01:04:16] Technodrome (Technodrome!~asdfasdf@unaffiliated/technodrome) has joined #mythtv
[01:04:35] Technodrome: hey whats the best way to record a channel
[01:04:42] Technodrome: the material is 480i i guess
[01:04:43] Technodrome: or p
[01:05:04] wagnerrp: more than likely i, and this is the mythtv development channel
[01:06:39] Technodrome: ah sorry
[01:06:41] jheizer_ (jheizer_! has joined #mythtv
[01:09:06] wilmoore-misc (wilmoore-misc! has joined #mythtv
[01:36:11] jya (jya!~jyavenard@mythtv/developer/jya) has quit (Quit: jya)
[01:36:26] len (len! has joined #mythtv
[01:41:32] jya (jya! has joined #mythtv
[01:41:32] jya (jya! has quit (Changing host)
[01:41:32] jya (jya!~jyavenard@mythtv/developer/jya) has joined #mythtv
[01:52:28] Beirdo (Beirdo!~gjhurlbu@mythtv/developer/beirdo) has quit (Ping timeout: 240 seconds)
[01:53:09] Beirdo (Beirdo! has joined #mythtv
[01:53:09] Beirdo (Beirdo! has quit (Changing host)
[01:53:09] Beirdo (Beirdo!~gjhurlbu@mythtv/developer/beirdo) has joined #mythtv
[01:57:37] wilmoore-misc (wilmoore-misc! has quit (Remote host closed the connection)
[02:00:48] Technodrome (Technodrome!~asdfasdf@unaffiliated/technodrome) has quit (Quit: Technodrome)
[02:15:47] danielk22 (danielk22!~danielk22@ has left #mythtv ()
[02:20:56] danielk22 (danielk22!~danielk22@ has joined #mythtv
[02:21:52] jheizer_ (jheizer_! has quit (Ping timeout: 252 seconds)
[02:33:44] wilmoore-misc (wilmoore-misc! has joined #mythtv
[03:02:10] wilmoore-misc (wilmoore-misc! has quit (Remote host closed the connection)
[03:02:56] jheizer_ (jheizer_! has joined #mythtv
[03:31:57] Technodrome (Technodrome!~asdfasdf@unaffiliated/technodrome) has joined #mythtv
[03:33:15] jheizer_ (jheizer_! has quit (Ping timeout: 260 seconds)
[03:35:59] xris: it amazes me that ffmpeg still doesn't have a way to do cutlists.. just a single cut (start-at + duration)
[04:12:49] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has quit (Ping timeout: 248 seconds)
[04:14:00] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has joined #mythtv
[04:33:47] Technodrome (Technodrome!~asdfasdf@unaffiliated/technodrome) has quit (Quit: Technodrome)
[04:36:15] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Quit: Body blow! Body blow!)
[04:41:43] Technodrome (Technodrome!~asdfasdf@unaffiliated/technodrome) has joined #mythtv
[05:23:02] Technodrome (Technodrome!~asdfasdf@unaffiliated/technodrome) has quit (Quit: Technodrome)
[05:53:50] cesman (cesman!~cesman@pdpc/supporter/professional/cesman) has joined #mythtv
[06:41:26] neufeld (neufeld! has quit (Ping timeout: 272 seconds)
[06:41:28] HeXiLeD (HeXiLeD!~L3D@unaffiliated/hexiled) has quit (Ping timeout: 252 seconds)
[06:41:52] knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Ping timeout: 260 seconds)
[06:43:04] HeXiLeD (HeXiLeD!~L3D@ has joined #mythtv
[06:43:04] HeXiLeD (HeXiLeD!~L3D@unaffiliated/hexiled) has joined #mythtv
[06:43:04] HeXiLeD (HeXiLeD!~L3D@ has quit (Changing host)
[06:43:20] knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv
[06:57:11] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[07:12:27] Tobbe5178 (Tobbe5178! has joined #mythtv
[07:18:23] rsiebert_ (rsiebert_! has joined #mythtv
[07:20:42] rsiebert (rsiebert! has quit (Ping timeout: 272 seconds)
[07:27:30] len (len! has quit (Read error: Connection reset by peer)
[08:31:03] natanojl (natanojl! has joined #mythtv
[08:36:30] natanojl (natanojl! has quit (Ping timeout: 276 seconds)
[08:36:42] SteveGoodey (SteveGoodey! has joined #mythtv
[09:11:18] jya (jya!~jyavenard@mythtv/developer/jya) has quit (Quit: jya)
[09:25:54] Steve-Goodey (Steve-Goodey! has joined #mythtv
[09:49:19] IReboot (IReboot! has quit (Remote host closed the connection)
[09:51:54] 20WAB4HBZ (20WAB4HBZ! has joined #mythtv
[09:52:25] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has quit (Ping timeout: 260 seconds)
[09:59:24] 20WAB4HBZ (20WAB4HBZ! has quit (Quit: Ex-Chat)
[10:00:08] IReboot (IReboot! has joined #mythtv
[10:02:01] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has joined #mythtv
[10:19:51] bas-t (bas-t! has joined #mythtv
[10:27:07] Sharky112065 is now known as Sharky-Sleep
[10:28:08] gary_buhrmaster (gary_buhrmaster!~gtb@2001:470:80e4:3::36) has left #mythtv ()
[10:32:53] joki- (joki-! has joined #mythtv
[10:33:14] joki (joki! has quit (Ping timeout: 272 seconds)
[10:33:15] joki- is now known as joki
[10:44:48] Technodrome (Technodrome!~asdfasdf@unaffiliated/technodrome) has joined #mythtv
[10:55:54] Cougar (Cougar!~cougar@2a03:5880:104:10:b999:e054:7e1b:2de8) has quit (Ping timeout: 264 seconds)
[11:06:08] neufeld (neufeld! has joined #mythtv
[11:36:48] Technodrome (Technodrome!~asdfasdf@unaffiliated/technodrome) has left #mythtv ()
[12:18:53] bas-t (bas-t! has quit (Quit: Ex-Chat)
[12:43:47] danielk22 (danielk22!~danielk22@ has quit (Ping timeout: 256 seconds)
[12:56:36] danielk22 (danielk22! has joined #mythtv
[13:21:03] stichnot: xris: it seems that ffmpeg may be able to cut an initial section and a final section — ffmpeg -acodec copy -vcodec copy -ss <starttime> -t <duration>
[13:21:18] stichnot: but cutting a section out of the middle? maybe not...
[13:28:48] stichnot: peper03: I do see VDPAU error when displaying those DVD subtitles. I'll have a look, unless you've already solved it :)
[13:34:33] Cougar (Cougar!~cougar@2a03:5880:104:10:81a:80e7:4cff:4cc7) has joined #mythtv
[13:36:59] amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has quit (Quit: Konversation terminated!)
[13:43:46] amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has joined #mythtv
[13:49:35] Jordack (Jordack! has joined #mythtv
[13:56:56] peper03 (peper03! has joined #mythtv
[13:58:15] peper03: stichnot: Not solved it but I have made progress :) It's a bit of a tangled web (in my head at least) so I'll try to explain what I've found as coherently as possible :)
[13:59:33] peper03: I found that the subtitles displayed correctly under VDPAU and OpenGL when I enabled subtitles (even though they are forced subtitles). Subtitles disabled – not centred, Subtitles enabled – they are centred.
[14:01:44] peper03: A bit more debugging and I found that when subtitles are disabled, the subtitle zoom factor changes from 100 to 90. That seems to be due to this line . . . een.cpp#L579
[14:03:07] peper03: When subtitles are disabled, subtitleType gets set to kDisplayNone, which isn't handled in the switch. I have no recollection of ever changing the subtitle zoom before but I guess I must have played with it at some point.
[14:05:44] peper03: At first glance, I would say that bit of code is not ideal but I'm not sure how it should be changed for the best. Perhaps setting the zoom to 100% if subtitles are disabled? I assume that will only affect forced subtitles?
[14:06:28] peper03: Can you have forced subtitles with the other subtitles methods?
[14:07:40] peper03: Next was the question of why it worked ok with Xv. It seems that's down to the aspect ratio.
[14:13:28] stichnot: peper03: I was just coming to similar conclusions. It seems that text subtitles are surprisingly turned on and off during startup.
[14:15:09] stichnot: I think it's possible for mkv files to mark text subtitles (individually and/or entire track) as forced.
[14:15:53] peper03: From what I can tell, an instance of SubtitleScreen is created as normal and EnableSubtitles is called, which sets the zoom factor to 90. Then MythPlayer::DisplayNormalFrame calls CheckAspectRatio, which determines that 1.25 is not the same as 1.3333, and calls ReinitOSD, which deletes the old instance of SubtitlesScreen, creates a new one but *doesn't* cause the zoom to be loaded.
[14:16:34] peper03: Because EnableSubtitles is called with 'forcedOnly = true'.
[14:18:31] peper03: Presumably, there is no aspect ratio issue when using VDPAU or OpenGL so the same instance is used all the time.
[14:19:51] stichnot: I think the aspect ratio depends on the source material and not the display device, but this looks like the right track.
[14:20:10] peper03: There's still the issue of SubtitleScreen::DisplayScaledAVSubtitles causing images with width=0,height=0 to be generated and therefore lots of error messages.
[14:20:22] peper03: But that's not why the subtitles aren't centred.
[14:21:28] peper03: I haven't looked much further into the aspect ratio side of things yet but since I'm using the same source and display resolution, just changing between video renders, it would seem the logicial conclusion.
[14:21:46] stichnot: the height=0 may be a case of < instead of <= or something similar
[14:24:42] peper03: When DisplayScaledAVSubtitles is called with top=true, bbox ends up as (509x0+76+254). But like I said, that bit of code makes my head hurt so I haven't investigated further :)
[14:35:38] peper03: Ok, the aspect ratio trigger is the same but then it hits this line . . . yer.cpp#L549
[14:37:36] peper03: stretch is 100 for OpenGL and 94 for Xv.. osd->GetFontStretch returns 100 in both cases, but that causes the OSD to be reinitialized with Xv.
[14:52:01] danielk22 (danielk22! has quit (Remote host closed the connection)
[14:56:31] stichnot: you think that's what's causing the artifacts?
[14:56:52] stichnot: in any case, I'll be back online in a half hour.
[14:57:52] peper03: No idea. But they've got something to do with whether subtitles are enabled. If they are, there are no artifacts. If they aren't, you get the artifacts along the left and right sides.
[14:58:16] peper03: Yep, I'm off for a bit too :)
[14:59:49] danielk22 (danielk22!~danielk22@ has joined #mythtv
[15:01:29] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 256 seconds)
[15:18:21] wilmoore-misc (wilmoore-misc! has joined #mythtv
[15:20:11] Wolfgang2 (Wolfgang2! has joined #mythtv
[15:25:01] stuartm: Ubuntu continues to dazzle me with it's lousy configuration tools, where's the simple 'Enable Wake-on-Lan' checkbox in the network configuration dialog?
[15:26:15] stuartm: Although, it seems to have disappeared from Mandrake/Mageia at some point too, probably when they made the fateful decision to stop using their own stuff in favour of the network manager abomination
[15:36:42] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[15:54:38] stuarta: i hate network mangler
[15:55:53] stichnot: peper03: See if this gets rid of the artifacts.
[16:00:55] Jordack (Jordack! has quit ()
[16:02:04] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 245 seconds)
[16:24:27] stichnot (stichnot!~stichnot@ has joined #mythtv
[16:24:28] stichnot (stichnot!~stichnot@ has quit (Changing host)
[16:24:28] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[16:27:31] natanojl (natanojl! has joined #mythtv
[16:36:33] peper03: stichnot: Yes, that seems to fix it. Wonder why it's not necessary for VDPAU/OpenGL?
[16:46:15] wilmoore-misc (wilmoore-misc! has quit (Remote host closed the connection)
[16:49:37] anykey_ (anykey_! has quit (Ping timeout: 245 seconds)
[16:49:49] wilmoore-misc (wilmoore-misc! has joined #mythtv
[16:57:08] wilmoore-misc (wilmoore-misc! has quit (Remote host closed the connection)
[16:58:02] jheizer_ (jheizer_! has joined #mythtv
[17:05:40] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv
[17:13:15] wilmoore-misc (wilmoore-misc! has joined #mythtv
[17:14:39] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Ping timeout: 256 seconds)
[17:29:26] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv
[17:36:54] Merlin83b (Merlin83b!~Daniel@2a00:1ee0:3:1337:ecd9:bb72:c99f:d3e2) has quit (Read error: Connection reset by peer)
[17:58:57] stichnot: peper03: MythUI keeps track of the "dirty region" and optimizes updates for just the dirty region. When you add or resize an element, that automatically updates the dirty region, but you have to be careful to trigger it when you delete an element. I think VDPAU and OpenGL ignore the dirty hint and just redraw everything.
[18:01:06] stichnot: I had to do something similar in . . . b21b7d796248 , and I only discovered the problem when I started using a laptop with Xv
[18:05:39] danielk22 (danielk22!~danielk22@ has quit (Ping timeout: 256 seconds)
[18:21:39] danielk22 (danielk22! has joined #mythtv
[18:22:18] stichnot: peper03: btw, with that patch, the artifacts go away, but the subtitles disappear earlier than before. Maybe one of your patches fixes that.
[18:24:54] Wolfgang2 (Wolfgang2! has quit (Quit: Wolfgang2)
[18:28:19] danielk22 (danielk22! has quit (Ping timeout: 245 seconds)
[18:29:09] peper03: stichnot: I hadn't noticed that they go away earlier, but I didn't pay that much attention. To be honest, I thought I'd set the subtitles so that there should be no gap between them but I haven't had chance to check how they're actually encoded on the DVD.
[18:30:38] peper03: I'm not sure whether the patches I submitted would fix anything in this case. #11451 only really applies to menu subpictures as the start and end times were being ignored completely in that case.
[18:30:38] ** MythLogBot **
[18:30:53] Schabo (Schabo!~schabo@2001:470:28:b16:cdd:5e39:98ff:2377) has joined #mythtv
[18:31:23] stichnot: I would guess that they are disappearing as soon as a new forced subtitle is encountered in the stream, without regard to the subtitle's start/end times.
[18:32:22] peper03: stichnot: It's possible. I'll have to check the encoding and then maybe follow the code flow.
[18:32:41] stichnot: In any case, I don't like the way subtitles are heavy-handedly enabled when a forced subtitle is seen, so I'll need to dig into that a bit.
[18:33:28] stichnot: As well as the top/bottom code that may have an off-by-one error leading to an empty box being rendered.
[18:37:28] peper03: I was thinking about the top/bottom code earlier. As I said, I don't really understand in detail what it's doing but I wondered if there isn't a simpler approach. There may well be a flaw in my logic or certain cases where it just won't work, but I would take a different approach.
[18:38:34] peper03: Basically, the zoom should be centred on the middle of the subtitle block. So I don't think it's really that necessary to determine whether a subtitle is towards the top or bottom.
[18:41:14] peper03: Let me see if I can get this written down somewhere first. It's all in my head at the moment and I'm probably not going to do a very good job of explaining it without writing it down and organizing it first :)
[18:42:39] sphery: Anyone else seeing extremely slow response times from Trac?
[18:42:58] jheizer__ (jheizer__! has joined #mythtv
[18:43:22] sphery: Actually, seems like it's getting better, now... Maybe it was just some robot hammering the site for the last 10min I've been trying to use it.
[18:44:30] stoffel (stoffel! has joined #mythtv
[18:45:42] danielk22 (danielk22!~danielk22@ has joined #mythtv
[18:46:18] jheizer_ (jheizer_! has quit (Ping timeout: 264 seconds)
[18:48:29] len (len! has joined #mythtv
[18:48:59] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Remote host closed the connection)
[18:49:53] peper03: Ok, I think this will work. Ignoring the horizontal axis for the moment, lets say we have a display of 480 lines. Line 0 is at the top. We have a subtitle 10 pixels high, where the top is at line 460.
[18:51:17] peper03: If we zoom in so that the subtitle is now 20 pixels high, you take the height delta, divide by two and subtract that from the original top giving a new start line of 455.
[18:51:51] peper03: The centre line remains at line 465.
[18:52:42] peper03: If the subtitle is at the top of the screen, it still works. Top is line 10, so centre line is 15. After zooming, top is line 5 and centre line is still 15 (10–20 -> 5–25)
[18:54:00] peper03: If you then detect that the subtitle is going to go off screen, you'd need to bump it so that the top or bottom line up with line 0 or line 480 (or whatever border you want to define).
[18:54:15] peper03: Horizontally it would work the same way.
[18:55:17] peper03: The bump would be something like: if (top < 0) { top += -top; }
[18:56:00] peper03: Obviously you'd need a bit more logic to detect if you had a very large subtitle that after zooming caused it to go off top *and* bottom, but that shouldn't be too difficult.
[18:56:09] dekarl1 (dekarl1! has joined #mythtv
[18:56:16] peper03: Does that sound reasonable?
[18:58:07] dekarl (dekarl! has quit (Ping timeout: 264 seconds)
[19:11:09] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[19:11:34] Steve-Goodey (Steve-Goodey! has quit (Quit: Konversation terminated!)
[19:22:08] stichnot: peper03: thinking about this.
[19:25:43] stichnot: peper03: The problem you are trying to solve is to keep vertically centered subtitles centered on the screen when a subtitle zoom factor is used (either <100% or >100%), right?
[19:26:23] stichnot: (and of course while keeping high and low subtitles looking nice)
[19:27:50] peper03: stichnot: Basically, yes. Zooming in or out should have the same effect whether a subtitle is high, low or in the middle. If it's in the middle, I'd expect it to stay in the middle. The same top and bottom.
[19:28:00] stichnot: OK, good.
[19:29:01] peper03: Splitting the screen into top and bottom just seems to be over-complicated. But maybe I just understand too little about how the current code works if that isn't trying to achieve the same thing.
[19:30:27] peper03: The same algorithm should work horizontally too. It's just a case of working with x instead of y.
[19:31:09] stichnot: The split-screen code for DVD subtitles is in the same spirit as the split-screen code for cc608 captions. The problem it is trying to solve (which is detailed in #10402) is that for multi-line cc608 captions with a zoom <100%, the vertical gaps between lines become unpleasantly large. I spent a while pondering this problem and couldn't come up with a better way than Gregorio's.
[19:31:09] ** MythLogBot **
[19:31:50] stichnot: E.g., your approach would still leave unnaturally large gaps between lines.
[19:32:04] stichnot: However, that may not really be applicable for DVDs.
[19:32:28] stichnot: Do you know if multi-line subtitles are rendered as multiple images?
[19:33:15] stichnot: If they tend to be single images in practice, then it seems your approach would be fine.
[19:33:28] peper03: They are just stored as 4-colour images. You can't know whether there's text or pictures in there.
[19:33:42] jheizer__ (jheizer__! has quit (Ping timeout: 264 seconds)
[19:34:03] stichnot: Right, but in practice are 2-line subtitles authored as 1 image or 2 stacked images?
[19:35:31] peper03: Off the top of my head, 1 image. If you wanted to show the first line on its own and then continue to show that whilst showing the second line, you'd have two images – the first with one line, the second with both lines. The second would replace the first.
[19:36:04] peper03: I don't think subpictures are allowed to overlap in time.
[19:36:38] stichnot: So you aren't allowed to have two separate subpictures displayed at the same time?
[19:37:25] peper03: I don't think so, no.
[19:42:01] stichnot: OK. The other concern is that you want to maintain a pleasing transition from say a one-line subtitles to a two-line subtitle. Ideally, however the author chooses to align the successive subtitles (top, bottom, or center), that alignment should still happen after zooming.
[19:42:54] stichnot: I think that the split-screen approach allows that naturally, provided the aligned subtitles are on the same half of the screen.
[19:43:04] stuartm: If you open the EPG while watching a recording, the guide starts from the channel the recording was made on instead of the 'guide start channel', I find that behaviour really annoying, it makes sense for livetv but not so much for recordings
[19:43:16] stuartm: does anyone mind if I change it?
[19:44:07] stichnot: stuartm: you can open the EPG while watching a recording? I don't think I'll be bothered. :)
[19:44:52] stuartm: stichnot: yeah, it's just a quicker way to schedule something that you've just seen the trailer for
[19:45:06] stichnot: ah, nice.
[19:46:39] peper03: stichnot: Ok, I had assumed that all subtitles ended up rendered to single image, and only a single subtitle image was shown at any given time.
[19:47:49] stuartm: which might make it sound like the current behaviour is good since any trailers are going to be for the same channel right ... except when they aren't and you've got to scroll through 200 channels (in the UK 90% of the watchable channels are all at the same end of the guide, numbered 1–40)
[19:56:37] peper03: stuartm: Did you try applying the updated patch from #10451? Just wanted to check it *does* apply cleanly or whether I need to look at it again.
[19:56:56] peper03: Sorry, I meant #11451
[19:56:56] ** MythLogBot **
[19:58:08] stuartm: peper03: haven't found the time yet, I will get to it though
[19:58:19] peper03: Ok, no problem. Thanks!
[20:04:10] gigem: stuartm: I would mind. If I enter the guide from watching a recording, it probably because I just saw a promo for something else I want to record on that channel.
[20:06:12] stuartm: gigem: ok, I'll just patch it locally :)
[20:06:33] gigem: stuartm: Thanks.
[20:10:30] stichnot: peper03: I'm talking about subtitles that appear sequentially in time, where the height of the first subtitle is different from the height of the second subtitle.
[20:13:21] stichnot: peper03: let me work up an example to illustrate
[20:19:59] stichnot: Let's say the screen height is 480, each line of subtitle text is rendered 20 pixels high, and the myth user has set zoom=50%. The DVD author wants to show a 3-line subtitle from 400..459, followed by a 1-line subtitle from 400..419. Note that the tops are aligned at 400. If you anchor each subtitle at its center, the first subtitle will be at 415..444 and the second subtitle will be at...
[20:20:01] stichnot: ...405..415. They are no longer aligned at the top.
[20:20:33] stichnot: The same misalignment happens if the DVD author makes the 1-line subtitle from 440..459 to be aligned at the bottom.
[20:21:49] stichnot: The idea of splitting the screen into the top and bottom half maintains alignment, as long as the aligned subtitles are on the same half of the screen.
[20:22:13] stoffel (stoffel! has quit (Ping timeout: 248 seconds)
[20:22:44] stichnot: This of course breaks down if you have a subtitle that is supposed to be vertically centered.
[20:24:19] stichnot: But I haven't encountered a subtitle like that yet (apart from your constructed example) – the vast majority are on the bottom half, with occasional subtitles placed at the top so as not to obscure other text like credits.
[20:25:16] stichnot: Anyway, I hope this helps explain the somewhat complicated logic behind the split-screen code.
[20:36:54] stichnot: peper03: stuartm: The patch for #11451 either requires or greatly benefits from one of the patches for #11371. I mentioned this before, but I prefer the first patch ( . . . e-we-.patch) because it is not DVD-specific and potentially fixes problems caused when cable providers naively splice ads...
[20:36:54] ** MythLogBot **
[20:36:54] ** MythLogBot **
[20:36:56] stichnot: ...into the program.
[20:37:23] stichnot: Because it's more general, it's also more risky, clearly.
[20:38:57] stuartm: stichnot: I committed the second patch, but only because I wasn't really prepared to take the risk of breaking anything with the first patch
[20:39:33] stichnot: Ah, I missed that commit. I was going to ask gigem to test the first patch against his recordings with pts discontinuities
[20:40:35] stuartm: if someone wants to revert and run with the first patch that's fine, so long as they accept responsibility for regression testing and troubleshooting any problems that are reported
[20:41:01] stuartm: it wasn't a minefield I felt like navigating :)
[20:41:08] stichnot: someone == me :)
[20:41:46] stuartm: good luck soldier, you're a braver man than I
[20:42:28] stichnot: stuartm: in any case, should 11371 be closed?
[20:44:36] stichnot: The first patch still applies cleanly, I'll run it for a while, and I probably have some of gigem's samples lying around
[20:50:38] Steve-Goodey (Steve-Goodey! has joined #mythtv
[20:51:58] stuartm: stichnot: I left it open just in case you wanted to apply the first patch instead
[20:52:24] stuartm: no harm done in it remaining open if that's what you're planning
[20:53:56] stichnot: ok, I'll grab those two tickets
[20:56:23] stichnot: peper03: in . . . hts-t.patch, did you agree that case 0x00 of DVDRingBuffer::DecodeSubtitles() could set sub->end_display_time=startTime?
[20:56:57] stichnot: and thereby fall into line with the other code that deals with zero-length subtitles?
[20:57:16] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv
[21:05:18] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Ping timeout: 276 seconds)
[21:14:06] stichnot: ok, I'll finish grabbing the tickets when trac comes back...
[21:16:11] Schabo (Schabo!~schabo@2001:470:28:b16:cdd:5e39:98ff:2377) has quit (Quit: Leaving)
[21:19:29] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv
[21:38:09] peper03: stichnot: Regarding the subtitle example: Yes, ok. That makes sense. It doesn't surprise me that my idea was too simple :) I don't think I've ever seen vertically-centred subtitles. As I said, it was only because I stole a menu highlight image that happened to be centred. Highlights are handled separately anyway with no zoom applied (which would be very bad unless the video were zoomed too).
[21:42:36] stichnot: peper03: with that said, one could also imagine splitting the screen into 3 vertical areas – top, center, bottom – and snapping/anchoring to one of them.
[21:45:59] peper03: stichnot: Regarding #11451 yes, I agreed. I fixed that in the second patch to #11455. I hadn't realised that I'd broken it in 11451 but the issue I was trying to fix with 11455 still existed in another aspect with OpenGL (the number of buffered subpictures) so I updated the patch from 11455 rather than 11451. Confused? I am :)
[21:45:59] ** MythLogBot **
[21:45:59] ** MythLogBot **
[21:48:40] peper03: And regarding #11371: Either patch would fix the issue I mentioned in the ticket, but the second patch (which has been committed) fixes other issues too. For example 11451. It's hard to wait for a start PTS when you can't know whether the start PTS is greater or less than the current PTS.
[21:50:44] stichnot: ok, I didn't realize other issues were fixed too.
[21:51:29] peper03: So reverting that second patch from 11371 would cause other issues. The scenario I tackled with the first patch could still theoretically occur (although not with DVDs any more) so I thought it worth pointing out. It should be possible to apply both patches, if necessary.
[21:52:29] peper03: I wasn't aware the other issues would be fixed by that either until I went back to them and realized they didn't exist any more :)
[21:54:26] peper03: I first looked at the start/end times for menu highlights last year, I think, but had at least one sample with a still menu where the PTS of the menu was 0. I left it for a while as I wasn't sure how to get around it. Then when I couldn't reproduce it any more, I realized I'd fixed it with that patch :)
[21:55:55] peper03: It's also possible that there's a better or cleaner solution than I came up with in the first patch. I was winging it a bit with that one :)
[22:02:04] peper03: BTW is there any chance of getting the second patch from 11371 applied to 0.26-fixes? Or the first one for that matter. Anything to get Pixar DVDs to play nicely!
[22:05:55] stichnot: stuartm pushed the second patch 2 days ago. In general for something of that complexity, I like to let it bake in Master for about a week before pushing to -fixes. For something as potentially sensitive as the first patch, I would want to avoid committing to -fixes unless really necessary.
[22:06:53] peper03: Ok, that's fine. Personally, I prefer the 2nd patch anyway :)
[22:08:35] stichnot: yeah, sounds like that's the way to go.
[22:09:47] peper03: Oh, just whilst I think about it. Pushing that patch to 0.26-fixes also requires the patches from #11413, which update libdvdnav and libdvdread.
[22:09:47] ** MythLogBot **
[22:10:15] peper03: I wouldn't rate those as critical (i.e. risky), personally.
[22:12:41] jya (jya!~jyavenard@mythtv/developer/jya) has joined #mythtv
[22:13:04] anykey_ (anykey_! has joined #mythtv
[22:17:05] stichnot: It looks like 11413 has baked in master for almost a month without issues.
[22:17:17] peper03: Yep.
[22:23:08] Steve-Goodey (Steve-Goodey! has quit (Quit: Konversation terminated!)
[23:01:39] amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has quit (Quit: Konversation terminated!)
[23:05:00] natanojl (natanojl! has quit (Ping timeout: 264 seconds)
[23:30:42] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 264 seconds)
[23:32:16] stichnot (stichnot!~stichnot@ has joined #mythtv
[23:32:16] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[23:32:16] stichnot (stichnot!~stichnot@ has quit (Changing host)
[23:35:02] wilmoore-misc (wilmoore-misc! has quit (Remote host closed the connection)

IRC Logs collected by BeirdoBot.
Please use the above link to report any bugs.