RaspBMC Kodi/XBMC test builds - Printable Version +- Kodi Community Forum (https://forum.kodi.tv) +-- Forum: Support (https://forum.kodi.tv/forumdisplay.php?fid=33) +--- Forum: General Support (https://forum.kodi.tv/forumdisplay.php?fid=111) +---- Forum: Raspberry Pi (https://forum.kodi.tv/forumdisplay.php?fid=166) +---- Thread: RaspBMC Kodi/XBMC test builds (/showthread.php?tid=176043) Pages:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
|
RE: RaspBMC XBMC test builds - popcornmix - 2014-03-29 (2014-03-29, 18:25)Mafarricos Wrote: You are speaking of hddvd sample, correct? As Matrix sample played OK, but only with some artefacts in some parts. I'm referring to the linked matrix sample. It will still play, as most frames don't use the warppoints and will appear correct, but you will get artefacts when they are used. The VC1 sample you reported obviously needs the VC1 codec. It will play if you buy that. RE: RaspBMC XBMC test builds - DBMandrake - 2014-03-29 As promised, here are the specifics of the stuttering problem I mentioned yesterday. I've seen this weird stuttering sporadically on streamed video ever since I started running the Gotham beta's on the Pi (I checked back and reproduced it on March 23 but didn't verify any earlier) - and I suspect it occurs with videos with "glitches" in them that are perhaps poorly edited together or have some corrupted/missing frames, but this is the first time I've found one of my own local files to have the problem so that I can easily reproduce it. First, here's two examples of what it looks like: (I didn't upload them to youtube as they'll probably get blocked for copyright reasons even though short clips should be "fair use"...also I didn't want youtube doing any processing or frame rate conversion on them. Download for best video quality) https://www.dropbox.com/s/ixnz1tsyjrqycay/IMG_0665.MOV https://www.dropbox.com/s/13563by7df84lh3/IMG_0666.MOV On this particular video the symptoms are triggered by large forward jumps (10 minutes) and the first and second large jump from the beginning of the file will reproduce it. The above videos are shortly after two jumps forwards. If I use many small (30 second) jumps from the beginning of the video to same point in the video, it plays fine. Here is an example of the second section of video reached via lots of small jumps instead of two large ones, and it can be seen playback is perfect: https://www.dropbox.com/s/pq3gnchz3h4r06h/IMG_0667.MOV Also if I skip further into the video playback synchronises properly and I can then skip back to the same place and playback is perfect. Here is a debug log taken during playback. It's a fairly clean .xbmc folder to minimise any outside influences. In the debug log I started playback of the video then skipped +10 minutes twice then let it play for a little while while it was stuttering: http://pastebin.com/0Lc3kVe7 Here's a mediainfo on the file: Code: * * * MediaInfo Mac 0.7.36.0 file analysis report. Some further observations - it's not related to display refresh rate, I've tried 24, 50, and 60Hz and with adjust refresh rate to match on and off. It doesn't happen if I switch back to the stable version of Frodo that comes with Raspbmc. It doesn't happen on Gotham Beta 2 on my Mac. The same problem happens on both an HDMI TV and a DVI PC monitor. It doesn't matter where the file is played from - in the above example it was playing directly from SD card. A random sampling of my movie collection didn't show any other files exhibiting the same problem. RE: RaspBMC XBMC test builds - popcornmix - 2014-03-29 (2014-03-29, 19:33)DBMandrake Wrote: First, here's two examples of what it looks like: (I didn't upload them to youtube as they'll probably get blocked for copyright reasons even though short clips should be "fair use"...also I didn't want youtube doing any processing or frame rate conversion on them. Download for best video quality) What I'd really like is the source file you are playing on the Pi (ideally cut it down to a smaller size, assuming that fails) so I can investigate the stutter on my Pi. Explain exactly what you do to get the stutter (e.g. play for 10 seconds, then press big step forward twice rapidly). RE: RaspBMC XBMC test builds - DBMandrake - 2014-03-29 (2014-03-29, 19:58)popcornmix Wrote:What's the best tool to cut it down to a shorter sample ? (Mac or Windows, Mac preferred, as that's where the files are hosted and is the faster machine)(2014-03-29, 19:33)DBMandrake Wrote: First, here's two examples of what it looks like: (I didn't upload them to youtube as they'll probably get blocked for copyright reasons even though short clips should be "fair use"...also I didn't want youtube doing any processing or frame rate conversion on them. Download for best video quality) I'm not sure that an edited version of the file will show the same symptoms given the nature of the problem but I will give it a try. Quote:Explain exactly what you do to get the stutter (e.g. play for 10 seconds, then press big step forward twice rapidly).In the example I showed I just start playback at the beginning of the file then jump forward twice. Doesn't matter if I wait a few seconds or not, and the problem is there after the first jump too, its just there isn't much movement in that part of the movie. (Just scenery and tigers lying in the sun ) If I start playback further in then use reverse jumps to get back to (almost) the beginning then jump forward it doesn't seem to happen. It seems to be loosing synchronisation when performing certain size jumps across certain regions. If I use large jumps further into the video it rights itself again. RE: RaspBMC XBMC test builds - popcornmix - 2014-03-29 (2014-03-29, 20:02)DBMandrake Wrote: What's the best tool to cut it down to a shorter sample ? (Mac or Windows, Mac preferred, as that's where the files are hosted and is the faster machine) I'd suggest mp4box gui (http://my-mp4box-gui.zymichost.com/). It runs on windows. It should be quick, as this won't re-encode the file, so should only take a minute. Re: RE: RaspBMC XBMC test builds - Mafarricos - 2014-03-29 (2014-03-29, 18:33)popcornmix Wrote:(2014-03-29, 18:25)Mafarricos Wrote: You are speaking of hddvd sample, correct? As Matrix sample played OK, but only with some artefacts in some parts. Right, sorry for didn't notice that, vc1 and mpeg2 need licence. Thanks for explanation And about this one: http://distribution.bbb3d.renderfarming.net/video/mp4/bbb_sunflower_1080p_60fps_normal.mp4 Is too much for RPI? With 30fps and 24fps versions work OK. 60fps version lags audio and image. I'm trying to test the max limits RE: RaspBMC XBMC test builds - DBMandrake - 2014-03-29 (2014-03-29, 19:58)popcornmix Wrote: What I'd really like is the source file you are playing on the Pi (ideally cut it down to a smaller size, assuming that fails) so I can investigate the stutter on my Pi.Sample video PM'ed. RE: RaspBMC XBMC test builds - popcornmix - 2014-03-29 (2014-03-29, 21:12)DBMandrake Wrote: Sample video PM'ed. Thanks. Found the problem. Will be fixed in next firmware update. RE: RaspBMC XBMC test builds - DBMandrake - 2014-03-29 (2014-03-29, 21:55)popcornmix Wrote:Wow, talk about quick.(2014-03-29, 21:12)DBMandrake Wrote: Sample video PM'ed. I'm curious, was there actual corruption or poor encoding in the file that just wasn't handled as gracefully as it could have been, or was it an actual bug that just isn't triggered by most files ? Also interesting that the problem is in the firmware when it worked ok in Frodo on the same firmware. (I guess later versions of the player utilise the firmware differently ?) RE: RaspBMC XBMC test builds - Arless - 2014-03-29 Hi everyone, I'm trying to get DD and DTS passthrough using an external USB DAC using this: USB DAC PCM2704. My config is Raspberry Pi -> USB DAC -> Logitech Z5500. I already tested it with my desktop pc and it works, it sends DTS and DD to logitech speakers, but when I try to use with the RPi, it appears that the data is not correctly sent to the speakers because the lcd screen of the speakers switch between Detecting... and dts Digital or DD. I had to add this to the advancedsettings.xml to force to send the sound to the USB DAC because it was ignoring the GUI settings Code: <audiooutput> Any idea? Thanks a lot. RE: RaspBMC XBMC test builds - popcornmix - 2014-03-30 (2014-03-29, 22:17)DBMandrake Wrote: I'm curious, was there actual corruption or poor encoding in the file that just wasn't handled as gracefully as it could have been, or was it an actual bug that just isn't triggered by most files ? The file has no B frames which is fairly unusual. It meant the statistics to determine whether time stamps should be reordered or not (code that is not in Frodo) were very close and could end up oscillating after a bump (i.e. a seek). RE: RaspBMC XBMC test builds - miappa - 2014-03-30 Updated Gotham build, XBMC master from Mar 30 + newclock3 commits. Some info: • Master: Several fixes and updates • NC3: Only need a confirmation dialogue when resolution has actually changed • NC3: [paplayer] Use position based seeks for ffmpeg demuxer • NC3: Fix leaks in the failure handling of jpeg decode to texture • NC3: [omxplayer] Add ability to choose different scaling kernels through OSD, see below Firmware from March 30 included: • Fix for video stutter after seek on streams without b frames • Add gencmd for setting the current scaling kernel, see below Additional info and testing (DVDPlayer etc.), see post #1 To install XBMC build, SSH to Pi: Code: wget -O xbmc-13-20140330-nc3.tar.gz http://goo.gl/5CINQa --no-check-certificate Some words from popcornmix: Quote:The Pi has some quite advanced scaling hardware which is able to do higher quality scaling than other ARM (and probably most PC) boxes. RE: RaspBMC XBMC test builds - mlavende - 2014-03-30 (2014-03-28, 00:54)miappa Wrote: Updated Gotham build, XBMC master from Mar 27 + newclock3 commits.Hi, Apologies if I'm being stupid but I can't find the answer by searching. When I try the nightly the update goes perfectly but do I need to redetect all my media again? In the previous release I have a folder of TV and another of movies. I add things daily and click the options/update to rescan the media. This will then check against tvdb or imdb and add to the list. When I update to gotham this doesn't seem to work. Is the database different or in a different place? Do I just need to add the locations again and do a full scan? Thanks for any help. -Mat RE: RaspBMC XBMC test builds - miappa - 2014-03-30 If you mean that you don´t have the option in Context Menu in Library mode, yes it has been moved. Other than that it is the same except for fixes etc. It is actually somewhat skin dependent, but Confluence (and probably most other skins) will have it in the side menu in library. You probably want to use a skin that supports Gotham (like stock Confluence). Or did you mean something else? F.Y.I Gotham use a new updated database but it should update and include everything such as scanned libraries etc. It will however not work the other way around i.e. media you add to Gotham will not be added in Frodo if you switch back. Re: RE: RaspBMC XBMC test builds - mlavende - 2014-03-30 (2014-03-30, 22:10)miappa Wrote: If you mean that you don´t have the option in Context Menu in Library mode, yes it has been moved. Other than that it is the same except for fixes etc. That's perfect thanks. I did mean that the update didn't seem to work since I was using scan for new content. Moving the update library option to the side menu is a good idea. Easy to find. I just don't think to look there. Thanks for the info on the library too. Of this works I hope to not switch back [WHITE SMILING FACE] |