rendering does not complete

[James Van Damme]

Binary package hint: pitivi

PiTiVi 0.13.5, Ubuntu 11.04 (wait, what?? I thought this was 10.10, and it's January...well, anyway)

I assembled a 15 minute video, and put a couple JPEG titles at the end, fading the last one out. I want to show the movie on a laptop and projector, so I tried AVI out. I hit render, the movie plays fine and the progress bar tells me how much time I have left, but I get to "About one second left" ....................then it hangs until I cancel. I haven't found any encoders that actually work, but I'm going down the list.

ProblemType: Bug
DistroRelease: Ubuntu 10.10
Package: pitivi 0.13.5-1ubuntu3
ProcVersionSignature: Ubuntu 2.6.35-24.42-generic-pae 2.6.35.8
Uname: Linux 2.6.35-24-generic-pae i686
NonfreeKernelModules: h826d
Architecture: i386
CheckboxSubmission: 582b8142224c5914effc01ba22566b10
CheckboxSystem: 6ce041aeed0a2c17b3343b66d157175d
Date: Mon Jan 24 21:23:28 2011
InstallationMedia: Ubuntu 9.10 "Karmic Koala" - Release i386 (20091028.5)
PackageArchitecture: all
ProcEnviron:
 LANG=en_US.UTF-8
 SHELL=/bin/bash
SourcePackage: pitivi

[Jeff Fortin Tam (kiddo)]

This is a bit confusing. You say you're using Ubuntu 11.04, yet your bug report indicates that it is Ubuntu 10.10, and that the installation media was Ubuntu 9.10...

I'll assume this is ubuntu 10.10 and you're not actually running a dev version of ubuntu.

Could you try without fading out at the end?
Also check if there are errors when running pitivi from a terminal.

If you're able to test pitivi git too, that would be interesting.

[James Van Damme (vandamme)]

>This is a bit confusing. You say you're using Ubuntu 11.04, yet your bug
>report indicates that it is Ubuntu 10.10, and that the installation
>media was Ubuntu 9.10...

The Help menu says 11.04, "released April 2011." Either it was teleported from the future, or somebody released it by mistake ahead of time!
And they forgot to update the release media. Well, anyway, back to PiTiVi...

>Could you try without fading out at the end?

I have been using a 10 second avi clip from my home movies, with no fading:

MPEG4 renders OK, but it reports "1 hour 59 minutes remaining" and I have to terminate the render box; I can't "cancel".

AVIMUX and Ogg encoder work fine, but stop with "about 1 second remaining"; I have to hit cancel. Another strange thing is I have to push the "render" button twice to get it to render.

I will try some more things after I get some sleep...:-)

jim

____________________________________________________________
Travelocity Vacation Package Deals
Book A Vacation Now & Save Up To 30% on Hotels & Vacation Packages.
http://thirdpartyoffers.juno.com/TGL3141/4d3e4f061abead59621st03vuc

[Jeff Fortin Tam (kiddo)]

These kinds of things shouldn't happen. Please at least provide the
output that you get when running pitivi from a terminal.

Without bug reproduction instructions (and ideally, debug logs) it's
very hard to determine what caused it.
Please provide precise steps to trigger the bug. You can also create
debug logs by running pitivi like this:

PITIVI_DEBUG=*:5 GST_DEBUG=2 /usr/bin/pitivi > debug.log 2>&1

Then compress and attach the log file to this bug report.

[James Van Damme (vandammes)]

Hi. Atch are three debug files. The first one I was attempting to render
QVGA using avidemux. The render button had to be hit twice to start the
rendering, it then did the rendering but there was no output. The close
button doesn't close the window, have to X it out.
Debug2 file was a 480p, avidemux, same thing, no output.

My third try was a 480p, mpegpsmux. The rendering ended with "about 1 second
left" and I canceled the window. The output file was fine, however
(debug.mpeg.log file). I tried it again, and it rendered and terminated
normally, with all the buttons behaving (debug3.log).

I used a simple ten second mpeg video capture file with no fades or titles.

Jim

Reply-To: Bug 707239
>
>
> These kinds of things shouldn't happen. Please at least provide the
> output that you get when running pitivi from a terminal.
>
> Without bug reproduction instructions (and ideally, debug logs) it's
> very hard to determine what caused it.
> Please provide precise steps to trigger the bug. You can also create
> debug logs by running pitivi like this:
>
> PITIVI_DEBUG=*:5 GST_DEBUG=2 /usr/bin/pitivi > debug.log 2>&1
>
> Then compress and attach the log file to this bug report.
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/707239
>
> Title:
> rendering does not complete
>
> Status in “pitivi” package in Ubuntu:
> Incomplete
>
> Bug description:
> Binary package hint: pitivi
>
> PiTiVi 0.13.5, Ubuntu 11.04 (wait, what?? I thought this was 10.10,
> and it's January...well, anyway)
>
> I assembled a 15 minute video, and put a couple JPEG titles at the
> end, fading the last one out. I want to show the movie on a laptop and
> projector, so I tried AVI out. I hit render, the movie plays fine and
> the progress bar tells me how much time I have left, but I get to
> "About one second left" ....................then it hangs until I
> cancel. I haven't found any encoders that actually work, but I'm going
> down the list.
>
> ProblemType: Bug
> DistroRelease: Ubuntu 10.10
> Package: pitivi 0.13.5-1ubuntu3
> ProcVersionSignature: Ubuntu 2.6.35-24.42-generic-pae 2.6.35.8
> Uname: Linux 2.6.35-24-generic-pae i686
> NonfreeKernelModules: h826d
> Architecture: i386
> CheckboxSubmission: 582b8142224c5914effc01ba22566b10
> CheckboxSystem: 6ce041aeed0a2c17b3343b66d157175d
> Date: Mon Jan 24 21:23:28 2011
> InstallationMedia: Ubuntu 9.10 "Karmic Koala" - Release i386
> (20091028.5)
> PackageArchitecture: all
> ProcEnviron:
> LANG=en_US.UTF-8
> SHELL=/bin/bash
> SourcePackage: pitivi
>
>
>

[Jeff Fortin Tam (kiddo)]

I'm seeing a lot of these in the first debug log:

De-Activating... (/usr/lib/pitivi/python/pitivi/action.py:138)
Traceback (most recent call last):
  File "/usr/lib/pitivi/python/pitivi/ui/encodingdialog.py", line 179,
in _cancelButtonClickedCb
    self._shutDown()
  File "/usr/lib/pitivi/python/pitivi/ui/encodingdialog.py", line 87, in
_shutDown
    self.removeRecordAction()
  File "/usr/lib/pitivi/python/pitivi/ui/encodingdialog.py", line 234,
in removeRecordAction
    self.renderaction.deactivate()
  File "/usr/lib/pitivi/python/pitivi/action.py", line 140, in
deactivate
    raise ActionError()
pitivi.action.ActionError

I'm pretty sure these issues are fixed with the development version,
which had its rendering UI completely rewritten...
Would you be able to test the git version of pitivi and see if the
problem still happens there (see the pitivi website for instructions)?

[James Van Damme (vandammes)]

>I'm pretty sure these issues are fixed with the development version,
>which had its rendering UI completely rewritten...
>Would you be able to test the git version of pitivi and see if the
>problem still happens there (see the pitivi website for instructions)?

I checked out the instructions and installed some of the dependencies, but
.......it would probably take me as much time to figure this out & find time
to do it as waiting for the next version to come along. I have very limited
Linux experience, and no programming ability.

Experimenting with different encoders at my work machine, I found some
combinations that worked (like the "not recommended" ones). So possibly I'm
just doing it wrong. I'll RTFM, try more encoders, and see what happens for
now.

Thanks for your help!

jim

--
Jim & Mary Margaret Van Damme
Couple to Couple League-Central New York
http://nfp.freehostia.com/ <=Natural Family Planning info
Our real address is <email address hidden>!
-------------------------------------------------------------

[Jeff Fortin Tam (kiddo)]

You don't need programming knowledge to run the git version, but you do need to be a little bit comfortable to the command line to issue the following commands (taken from the website's download page):

sudo apt-get build-dep pitivi # gets all the dependencies, but also make sure you have intltool installed
git clone git://git.pitivi.org/git/pitivi.git # download the dev version
cd pitivi
./autogen.sh && ./configure && make # only need to do this once (if it works)
bin/pitivi

[James Van Damme (vandamme)]

Seems to be the same as bug 637841 (rendering impossible).

I discovered that if I move the first clip at least a fraction of a second from zero time, the video renders normally. This happens with many codecs, resolutions, fading in and out, still clips mixed in, anything. That is, it needs a tiny gap at the beginning. Without that, nothing works.

[Dave Johnson (dave-scomple)]

In my case, it hangs at the end saying "about one second left," but if I go have a cup of joe and build a cake or something it eventually completes. For now, I either take a weekend road trip while it renders the video or use cinelerra.

[Andreas (q-qless)]

I may have this issue too.

I created a simple project, consisting of one video file (1920x1080@60fps, mp4 file from a sanyo xacti cam), that I devided into 3 parts.
I put the 3rd part infront of the 1st part and discarded the 2nd part.

if I use the defult output setting: 576p (PAL, 25fps)
oggmux, vorbisenc, theorsenc
it renders fine and plays well.

but if I use different settings, like matroskamux, faac, x264enc (which should yield the same hd video files,
found all over the internet.) the render process never starts.
there is no increase of cpu usage or disk access or any other signs that the process is just slow and there is is big workload.
(my version is 0.13.4)

[Jeff Fortin Tam (kiddo)]

0.13.4 is deprecated and has many bugs that were solved in 0.13.5 or pitivi git. Please test with a newer version.

[James Van Damme (vandamme)]

Well, I was running 0.13.5, and I never got it to work until I moved the first clip a fraction of a second past zero. Then all the codecs work fine. Does everyone else experience this?

[Jeff Fortin Tam (kiddo)]

About 0.13.4: I was talking to Andreas :)

Moving the first clip a fraction of second past zero: that's weird, because I never experienced this issue... I just tested with 0.13.5 in Ubuntu 10.10. Or maybe Ubuntu 11.04 magically breaks everything again (I haven't tested that)...

[Jeff Fortin Tam (kiddo)]

Oh right, your hang happens at the *very end of rendering*, not at the beginning.

I can *maybe* reproduce this. What I can reproduce in pitivi git is your result (the video hanging at the end of rendering),
but I cannot reproduce the workaround (using a blank space) because of another stupid bug (not-negotiated errors).

When the hang occurs, the following is repeatedly printed to the terminal:

(pitivi:18082): GStreamer-CRITICAL **: gst_event_new_new_segment_full: assertion `start <= stop' failed
(pitivi:18082): GStreamer-CRITICAL **: gst_pad_push_event: assertion `event != NULL' failed

But the problem is that I can only reproduce this with very specific files (some vob files I have).
Please double-check (with a blank project) if this happens only with some specific media files...

[Jeff Fortin Tam (kiddo)]

Also, please specify the length of the footage clip you are using.

I'm starting to suspect something: in my observations so far, vob files that have "hung for a long time at 99% with 100% CPU usage" have all been files with long durations/big filesizes...

[James Van Damme (vandamme)]

Last night, I remixed the same video that gave me problems in January, to add in a 3 minute avi clip inside it. The original is a 200 MB avi spliced together from a couple other clips and some stills with overdubbed music. The original was 14 minutes, so I should have ended up with a 17 minute movie. NTSC, avi output.

After much fooling around with PiTiVi, and trying my trick of moving the start of the clip a fraction of a second off zero time (which worked in January), I got the old "1 second left" in rendering time which never ends (well, I gave up waiting). I tried different codecs....nothing worked.

This was on the same 'puter that gave me problems in January. I haven't changed anything but kept up with Ubuntu 10.10's upgrades (at least one kernel upgrade in there).

I switched to OpenShot, which was unuseable before because of a worse bug, and that worked for me just fine.

I just tried a chunk of the same movie with PiTiVi at work (also Ubuntu 10.10), cut out all but the intro and credits (stills with fades), and it rendered flawlessly. (2 MB avi file.)

Let me know if you want me to do some more testing. I don't want to go back to splicing VHS tapes together with tape.

[Jeff Fortin Tam (kiddo)]

Urgh. This is why I hate heisenbugs.

I filed a bug upstream (https://bugzilla.gnome.org/show_bug.cgi?id=644533) with my particular findings for the time being... if separate causes are identified, they should probably be made into separate bug reports.

[Charlie Shaw (n2924-dotcom)]

I have just pressed the 'report a problem' button, so all the detailed system data should be uploaded; even though the original install was from a fairly old Ubuntu distro, it was recently upgraded with the Ubuntu 'auto-update' to 11.10 that was recently pushed out (reading the Wikipedia entry on Pitivi, I am wondering if that is what has broken it.. did not try with the earlier version, and cannot go back..).

I have a ~13MByte .AVI file generated by a Casio camera, and all I am trying to do is cut off the last few seconds of footage.

Rendering to avi after moving the end time only seems to happen 'very quick', system appears to hang with the '1 second left' box, but top does not show any excessive CPU usage anywhere; output file name is created but of 0 size (about 24% of the hdd is used).

Hit cancel.

Try again, with the above 'work-around' of moving the start time in a few seconds; this time, the rendering takes a LOT longer (you see the video in slo mo in the right pane); note that it takes some time to 'skip over' the first few seconds that I have tried to delete by moving start time; and completes, writing an output file that is a few MBytes smaller than the original.

BUT .. trying to play the result (using Movieplayer that is I think the default in this distro), I do not get what I expect... (guess I need to file a different bug for that, after I have made sure I am driving pitivi correctly); simply said, you get a few seconds of 'nothing' (black screen, no audio), then some audio & a 'frozen frame'; then a few more 'almost disjoint' frames that do not end where I placed the end point.

I will have a go at collecting the debug logs as described above, and trying the git version of Pitivi; but since nothing has happened on this thread for some time, nor the gnome bug, I am wondering what the best way forward is...

[Jeff Fortin Tam (kiddo)]

Hi Charlie, no use piggybacking onto unclear bug reports. Come ask on IRC or file a bug with the input files you used, the output containers and codecs, etc.

[Jeff Fortin Tam (kiddo)]

Ah and by file a bug, I meant upstream in the gnome bug tracker :)

[Yannick Warnier (ywarnier)]

I've added information into the upstream bug (https://bugzilla.gnome.org/show_bug.cgi?id=644533) but there's not much movement there. What is the status of this and which is the right place to post?

[Charlie Shaw (n2924-dotcom)]

Jean-Francois is the one to answer those questions; but my status has changed:

1. tried to upgrade from 0.15.0, but also ran into problems, reported here: https://bugzilla.gnome.org/show_bug.cgi?id=686343 ; I agree with J-F's answer ( = cannot be done for a Ubuntu 11.10 environment..); IMHO the Pitivi web-site should be updated to reflect what can be used where..

2. Meanwhile, I tried out Openshot (sorry Pitivi guys...but it served my needs for 1 of the clips), & ran into a similar 'seems hung' issue, see here for all the gory details: https://bugs.launchpad.net/openshot/+bug/1067328.

       I have stepped away from trying to debug that for now, but there is 1 error at least in the code (in video.py, there is a test 'if floating_point == 100.0' that will fail), and IMHO the issue is not in the Openshot specific code, but somewhere like the MLT libraries..

3. During all this, some of the packages on this system got upgraded; specifically (I think):

# cat 2012-10-13.231214.log
Commit Log for Sat Oct 13 23:12:14 2012

Installed the following packages:
pitivi (0.15.0-0ubuntu1)
# cat 2012-10-13.235120.log
Commit Log for Sat Oct 13 23:51:20 2012

Completely removed the following packages:
pitivi
# cat 2012-10-17.142653.log
Commit Log for Wed Oct 17 14:26:53 2012

Removed the following packages:
libmlt4
python-mlt3

Upgraded the following packages:
libmlt++3 (0.7.4-3) to 0.8.2-0ubuntu0~sunab~oneiric1
libmlt-data (0.7.4-3) to 0.8.2-0ubuntu0~sunab~oneiric1
melt (0.7.4-3) to 0.8.2-0ubuntu0~sunab~oneiric1

Installed the following packages:
libmlt5 (0.8.2-0ubuntu0~sunab~oneiric1)
python-mlt5 (0.8.2-0ubuntu0~sunab~oneiric1)
# cat 2012-10-18.140406.log
Commit Log for Thu Oct 18 14:04:06 2012

Installed the following packages:
pitivi (0.15.0-0ubuntu1)

4. surprise surprise; pitivi 0.15.0 now seems to work...(ie it looks like I can shorten the original offending clip without causing pitivi to hang...can't be sure I am using exactly the same codecs though..).

5. If I get a chance to try & succeed at cleaning up the package dependencies etc on this machine (aptitude reports that the python-mlt3 & 5 packages are broken..) I will see if I can revert to the state when I first tried pitivi, reproduce the bug, then see exactly which library update (appears to) fix the issue..

[splashis (splashote)]

Bug still persists.

Just in case this helps:

(pitivi:5193): GStreamer-CRITICAL **: gst_event_new_new_segment_full: assertion `format != GST_FORMAT_UNDEFINED' failed

(pitivi:5193): GStreamer-CRITICAL **: gst_pad_push_event: assertion `event != NULL' failed
ERROR [ 5193] [0x7faa4ae55700] "<Pipeline at 0x2ec9c10>" pipeline Nov 23 17:58:20 _handleErrorMessage: error from /GstPipeline:pipeline0/GstBin:bin0/GnlComposition:gnlcomposition0/GnlSource:gnlsource: VideoTestSourceFactory36/GstBin:bin44/GstVideoTestSrc:videotestsrc19 (__main__.GstVideoTestSrc): GStreamer hat einen allgemeinen Datenstromfehler festgestellt. (gstbasesrc.c(2625): gst_base_src_loop (): /GstPipeline:pipeline0/GstBin:bin0/GnlComposition:gnlcomposition0/GnlSource:gnlsource: VideoTestSourceFactory36/GstBin:bin44/GstVideoTestSrc:videotestsrc19:
streaming task paused, reason not-negotiated (-4)) (/usr/lib/pitivi/python/pitivi/pipeline.py:858)

(pitivi:5193): GStreamer-CRITICAL **: gst_event_new_new_segment_full: assertion `format != GST_FORMAT_UNDEFINED' failed

(pitivi:5193): GStreamer-CRITICAL **: gst_pad_push_event: assertion `event != NULL' failed
ERROR [ 5193] [0x7faa4ae55700] "<Pipeline at 0x2ec9c10>" pipeline Nov 23 17:58:20 _handleErrorMessage: error from /GstPipeline:pipeline0/GstBin:bin0/GnlComposition:gnlcomposition0/GnlSource:gnlsource: VideoTestSourceFactory36/GstBin:bin44/GstVideoTestSrc:videotestsrc19 (__main__.GstVideoTestSrc): GStreamer hat einen allgemeinen Datenstromfehler festgestellt. (gstbasesrc.c(2625): gst_base_src_loop (): /GstPipeline:pipeline0/GstBin:bin0/GnlComposition:gnlcomposition0/GnlSource:gnlsource: VideoTestSourceFactory36/GstBin:bin44/GstVideoTestSrc:videotestsrc19:
streaming task paused, reason not-negotiated (-4)) (/usr/lib/pitivi/python/pitivi/pipeline.py:858)

(pitivi:5193): GStreamer-CRITICAL **: gst_event_new_new_segment_full: assertion `format != GST_FORMAT_UNDEFINED' failed

(pitivi:5193): GStreamer-CRITICAL **: gst_pad_push_event: assertion `event != NULL' failed

(pitivi:5193): GStreamer-CRITICAL **: gst_caps_make_writable: assertion `caps != NULL' failed

(pitivi:5193): GStreamer-CRITICAL **: gst_caps_set_simple: assertion `GST_IS_CAPS (caps)' failed

(pitivi:5193): GStreamer-CRITICAL **: gst_caps_unref: assertion `caps != NULL' failed

** (pitivi:5193): CRITICAL **: gst_video_format_get_size: assertion `format != GST_VIDEO_FORMAT_UNKNOWN' failed
Speicherzugriffsfehler (Speicherabzug geschrieben)

[Zup Arkhen (zup-mondesvirtuels)]

Same problem :

ERROR [30517] [0x -48d37500] "<Pipeline at 0xaf2988c>" pipeline déc. 05 15:59:01 _handleErrorMessage: error from /GstPipeline:pipeline0/GstBin:bin0/GnlComposition:gnlcomposition0/GnlSource:gnlsource: PictureFileSourceFactory60/GstBin:bin61/GstQueue:internal-queue (__main__.GstQueue): GStreamer a rencontré une erreur générale de flux. (gstqueue.c(1291): gst_queue_loop (): /GstPipeline:pipeline0/GstBin:bin0/GnlComposition:gnlcomposition0/GnlSource:gnlsource: PictureFileSourceFactory60/GstBin:bin61/GstQueue:internal-queue:
streaming task paused, reason error (-5)) (/usr/lib/pitivi/python/pitivi/pipeline.py:858)

(pitivi:30517): GStreamer-CRITICAL **: gst_event_new_new_segment_full: assertion `format != GST_FORMAT_UNDEFINED' failed

(pitivi:30517): GStreamer-CRITICAL **: gst_pad_push_event: assertion `event != NULL' failed
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
Cannot connect to server socket err = Aucun fichier ou dossier de ce type
Cannot connect to server socket
jack server is not running or cannot be started
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
AL lib: pulseaudio.c:588: pa_context_new() failed
AL lib: alsa.c:512: Could not open playback device 'default': Périphérique ou ressource occupé
AL lib: oss.c:169: Could not open /dev/dsp: Aucun fichier ou dossier de ce type
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
mmap() failed: Ne peut allouer de la mémoire
Cannot connect to server socket err = Aucun fichier ou dossier de ce type
Cannot connect to server socket
jack server is not running or cannot be started