GStreamer issueshttps://gitlab.freedesktop.org/groups/gstreamer/-/issues2022-11-17T14:54:23Zhttps://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/128intervideosrc/intervideosink: cannot change channel while in playing state2022-11-17T14:54:23ZBugzilla Migration Userintervideosrc/intervideosink: cannot change channel while in playing state## Submitted by lou..@..il.com
**[Link to original bug (#722877)](https://bugzilla.gnome.org/show_bug.cgi?id=722877)**
## Description
Once the intervideosrc and intervideosink elements are in the playing state, changing the channel ...## Submitted by lou..@..il.com
**[Link to original bug (#722877)](https://bugzilla.gnome.org/show_bug.cgi?id=722877)**
## Description
Once the intervideosrc and intervideosink elements are in the playing state, changing the channel property does not have any effect on their behavior. The elements must be moved to null and then back to playing for the channel behavior to change.
Version: 1.2.0https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/127New umxparse element2021-09-24T14:32:21ZBugzilla Migration UserNew umxparse element## Submitted by Carlos Rafael Giani
**[Link to original bug (#722618)](https://bugzilla.gnome.org/show_bug.cgi?id=722618)**
## Description
Created attachment 266761
Adds umxparse element
Unreal Engine 1 based games such as Un...## Submitted by Carlos Rafael Giani
**[Link to original bug (#722618)](https://bugzilla.gnome.org/show_bug.cgi?id=722618)**
## Description
Created attachment 266761
Adds umxparse element
Unreal Engine 1 based games such as Unreal 1, Unreal Tournament 1, and Deus Ex 1 use module music (like MOD,S3M,XM,IT..) contained within Unreal packages. Such files are called UMX files. These Unreal packages contain only one module song, and nothing more.
This new element parses the UMX data contents, extracts the module, and sends it downstream. Module players such as modplug can then play the data as if it came from an ordinary .s3m or .xm file.
**Patch 266761**, "Adds umxparse element":
[0001-umxparse-add-new-parser-element-for-module-music-in-.patch](/uploads/5b9d01e0c02b0ffbe7917b64626f6625/0001-umxparse-add-new-parser-element-for-module-music-in-.patch)https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/126tq: tee element with embedded queue elements on srcpads2021-09-24T14:32:20ZBugzilla Migration Usertq: tee element with embedded queue elements on srcpads## Submitted by Andrey Utkin
**[Link to original bug (#722511)](https://bugzilla.gnome.org/show_bug.cgi?id=722511)**
## Description
Created attachment 266633
patch
This was discussed in gstreamer-devel maillist a bit, and i w...## Submitted by Andrey Utkin
**[Link to original bug (#722511)](https://bugzilla.gnome.org/show_bug.cgi?id=722511)**
## Description
Created attachment 266633
patch
This was discussed in gstreamer-devel maillist a bit, and i was encouraged to post this as a patch to -plugins-bad.
Besides inclusion into upstream, what interests me is code review, so i figure out if used things are operated correctly. Any comments appreciated.
The code currently lacks management of inside tee and queues properties. I haven't decided yet how exactly they would be passed through, possibly as two properties "tee-props" and "queue-props" as strings representing a set of key-value pairs.
P. S. How xml doc is generated?
**Patch 266633**, "patch":
[0001-Introduce-tq-tee-element-with-embedded-queue-element.patch](/uploads/97957fa1cc2a4633096230a9d03b18c5/0001-Introduce-tq-tee-element-with-embedded-queue-element.patch)https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/125waylandsink: fix surface width/height when caps have changed2021-09-24T14:32:19ZBugzilla Migration Userwaylandsink: fix surface width/height when caps have changed## Submitted by Benjamin Gaignard
**[Link to original bug (#722343)](https://bugzilla.gnome.org/show_bug.cgi?id=722343)**
## Description
Created attachment 266474
waylandsink: fix surface width/height when caps have changed
I...## Submitted by Benjamin Gaignard
**[Link to original bug (#722343)](https://bugzilla.gnome.org/show_bug.cgi?id=722343)**
## Description
Created attachment 266474
waylandsink: fix surface width/height when caps have changed
If a player (like gst-play) only change the state between two files
waylandsink need to be able to resize the surface output according
to the new width/height values.
I have tested this patch with a command line similar to this one:
gst-play --audiosink fakesink --videosink waylandsink FILE1 FILE2
~~**Patch 266474**~~, "waylandsink: fix surface width/height when caps have changed":
[0001-waylandsink-fix-surface-width-height-when-caps-have-.patch](/uploads/33e502321aa973a55da4db2e0b3a5796/0001-waylandsink-fix-surface-width-height-when-caps-have-.patch)
Version: 1.xhttps://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/124Unable to play animal_planet.mpg video stream2021-09-24T14:32:19ZBugzilla Migration UserUnable to play animal_planet.mpg video stream## Submitted by Jan ONDREJ (SAL)
**[Link to original bug (#721621)](https://bugzilla.gnome.org/show_bug.cgi?id=721621)**
## Description
This stream can't be played with gstreamer:
http://work.salstar.sk/public/gstreamer/animal_...## Submitted by Jan ONDREJ (SAL)
**[Link to original bug (#721621)](https://bugzilla.gnome.org/show_bug.cgi?id=721621)**
## Description
This stream can't be played with gstreamer:
http://work.salstar.sk/public/gstreamer/animal_planet.mpg
Response from Tim@IRC:
<__tim:#gstreamer> the primary bug is that it doesn't play at all, the
secondary issue is the detection of that lpcm or whatever stream
Steps to reproduce:
gst-launch-1.0 playbin uri=http://work.salstar.sk/public/gstreamer/animal_planet.mpg
Version: 1.2.2https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/123game-music-emu files with multiple songs only play first song2021-09-24T14:32:18ZBugzilla Migration Usergame-music-emu files with multiple songs only play first song## Submitted by mih..@..il.com
**[Link to original bug (#720687)](https://bugzilla.gnome.org/show_bug.cgi?id=720687)**
## Description
Hi guys! The GBS format supports multiple songs (from the Game Boy) in one .gbs file. See [1] for ...## Submitted by mih..@..il.com
**[Link to original bug (#720687)](https://bugzilla.gnome.org/show_bug.cgi?id=720687)**
## Description
Hi guys! The GBS format supports multiple songs (from the Game Boy) in one .gbs file. See [1] for an example file. I've checked the git code [2], and even though the game-music-emu library supports playback and querying info about multiple songs in one file (via gme_track_count() and gme_start_track()), track 0 (first track of file) is hardcoded, thus ignoring any other tracks in the same file.
I would write the patch myself, but I don't know the GStreamer API well enough, I'm only familiar with the game-music-emu API. I've seen an article [3] about streamids, so I know this should be possible. If someone can hint to where I can look in the code, maybe a similar plugin that supports multiple tracks, or the API headers, I can make a patch.
[1] http://www.zophar.net/music/gbs/castlevania-2-belmont-s-revenge.html
[2] http://cgit.freedesktop.org/gstreamer/gst-plugins-bad/tree/ext/gme/gstgme.c
[3] http://blogs.gnome.org/uraeus/2012/11/25/improved-handling-of-files-with-multiple-tracks-in-gstreamer/https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/120hlssink: does not always adhere to the segment target duration2021-09-24T14:32:17ZBugzilla Migration Userhlssink: does not always adhere to the segment target duration## Submitted by Baldur Gislason
**[Link to original bug (#719785)](https://bugzilla.gnome.org/show_bug.cgi?id=719785)**
## Description
It seems that the force-key-unit events emitted by hlssink are being ignored by the upstream in s...## Submitted by Baldur Gislason
**[Link to original bug (#719785)](https://bugzilla.gnome.org/show_bug.cgi?id=719785)**
## Description
It seems that the force-key-unit events emitted by hlssink are being ignored by the upstream in some cases. This depends on the source file.
Executing the pipe:
gst-launch-1.0 -v --gst-debug=mpegtsmux:4,hlssink:4,x264enc:4 filesrc location=klovn_1500.mp4 ! decodebin ! queue ! x264enc ! mpegtsmux ! hlssink target-duration=10
result:
http://pastebin.com/UTae4KXK
The interesting messages are:
0:00:02.083604323 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:07.800000000 wanted 0:00:10.000000000
0:00:02.083781840 21359 0xc0e8f0 INFO x264enc gstx264enc.c:1902:gst_x264_enc_encode_frame:`<x264enc0>` Forcing key frame
0:00:02.093407736 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:08.000000000 wanted 0:00:10.000000000
And a few seconds later:
0:00:02.975715319 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:09.800000000 wanted 0:00:10.000000000
0:00:02.998043430 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:09.880000000 wanted 0:00:10.000000000
0:00:02.998910966 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:806:mpegtsmux_sink_event:`<mpegtsmux0>` have downstream force-key-unit event on pad sink_65, seqnum 263, running-time 0:00:10.000000000 count 1
0:00:02.998932445 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:810:mpegtsmux_sink_event:`<mpegtsmux0>` skipping downstream force key unit event as an upstream force key unit is already queued
0:00:02.998955341 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:09.920000000 wanted 0:00:10.000000000
0:00:03.000188674 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:10.000000000 wanted 0:00:10.000000000
0:00:03.000215625 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:943:check_pending_key_unit_event: pending force key unit, waiting for keyframe
0:00:03.010917966 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:09.960000000 wanted 0:00:10.000000000
0:00:03.046916732 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:10.080000000 wanted 0:00:10.000000000
0:00:03.046973034 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:943:check_pending_key_unit_event: pending force key unit, waiting for keyframe
And more seconds later:
0:00:04.253965353 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:943:check_pending_key_unit_event: pending force key unit, waiting for keyframe
0:00:04.264080154 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:12.800000000 wanted 0:00:10.000000000
0:00:04.264115620 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:943:check_pending_key_unit_event: pending force key unit, waiting for keyframe
0:00:04.264906283 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:12.920000000 wanted 0:00:10.000000000
0:00:04.264959113 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:1092:mpegtsmux_collected_buffer:`<mpegtsmux0>` pushing downstream force-key-unit event 234 0:00:12.920000000 count 1
0:00:04.264977565 21359 0xc0e8f0 INFO hlssink gsthlssink.c:530:gst_hls_sink_ghost_event_probe:`<hlssink0>` event: custom-downstream
0:00:04.264993943 21359 0xc0e8f0 INFO hlssink gsthlssink.c:562:gst_hls_sink_ghost_event_probe:`<hlssink0>` setting index 1
0:00:04.265520688 21359 0xc0e8f0 INFO hlssink gsthlssink.c:358:gst_hls_sink_handle_message:`<hlssink0>` COUNT 1
0:00:04.275804950 21359 0xc0e8f0 INFO hlssink gsthlssink.c:588:schedule_next_key_unit:`<hlssink0>` sending upstream force-key-unit, index 2 now 0:00:12.920000000 target 0:00:22.920000000
0:00:04.275849188 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:862:mpegtsmux_src_event:`<mpegtsmux0>` received upstream force-key-unit event, seqnum 266 running_time 0:00:22.920000000 all_headers 1 count 2
0:00:04.275869467 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:883:mpegtsmux_src_event:`<mpegtsmux0>` forwarding to sink_65
0:00:04.275885087 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:885:mpegtsmux_src_event:`<mpegtsmux0>` result 1
0:00:04.278922057 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:13.080000000 wanted 0:00:22.920000000
0:00:04.279804278 21359 0xc0e8f0 INFO mpegtsmux mpegtsmux.c:937:check_pending_key_unit_event: now 0:00:13.000000000 wanted 0:00:22.920000000
Version: 1.xhttps://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/119Add opencv stitching (panorama) element.2021-09-24T14:32:17ZBugzilla Migration UserAdd opencv stitching (panorama) element.## Submitted by Miguel Casas-Sanchez `@miguelecasassanchez`
**[Link to original bug (#719688)](https://bugzilla.gnome.org/show_bug.cgi?id=719688)**
## Description
Created attachment 263303
Full patch containing panography element,...## Submitted by Miguel Casas-Sanchez `@miguelecasassanchez`
**[Link to original bug (#719688)](https://bugzilla.gnome.org/show_bug.cgi?id=719688)**
## Description
Created attachment 263303
Full patch containing panography element, mods to gstopencv.cpp and gsttemplatematch.cpp.
- Added a file that stitches together two video feeds - that supposedly overlap, and creates a single larger output panorama. Like panorama picture, for videos :)
Multi pad input to one output synchro and glue code is copy-modified from gstdisparity.cpp :)
From the doco:
* This element stitches two video feeds together, and represents the right
* channel as a reference over which the left one warps - supposed of course
* there is enough overlap for a correct feature matching.
- File gstopencv.c is changed to gstopencv.cpp, allowing for C++ (yei!) usage in gstpanography.{h,cpp}.
- File gsttemplatematch.{h,cpp}: renamed member |template| to |templatename|, clashed with C++ when included in gstopencv.cpp.
- Will upload some pics of how the output looks like.
**Patch 263303**, "Full patch containing panography element, mods to gstopencv.cpp and gsttemplatematch.cpp.":
[gstpanography.0.patch](/uploads/1434c66676c99331412e2aa67bd57629/gstpanography.0.patch)
Version: 1.xhttps://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/118Support the functionality for SCHC broadcasting (3D-TV)2021-09-24T14:32:17ZBugzilla Migration UserSupport the functionality for SCHC broadcasting (3D-TV)## Submitted by ji0..@..ng.com
**[Link to original bug (#719333)](https://bugzilla.gnome.org/show_bug.cgi?id=719333)**
## Description
Support the functionality for SCHC broadcasting
### Depends on
* ~~[Bug 611157](https://bugzill...## Submitted by ji0..@..ng.com
**[Link to original bug (#719333)](https://bugzilla.gnome.org/show_bug.cgi?id=719333)**
## Description
Support the functionality for SCHC broadcasting
### Depends on
* ~~[Bug 611157](https://bugzilla.gnome.org/show_bug.cgi?id=611157)~~ (IMPLEMENTED)https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/117tsdemux: Provide PAT/PMT from upstream2021-09-24T14:32:16ZBugzilla Migration Usertsdemux: Provide PAT/PMT from upstream## Submitted by Eric
**[Link to original bug (#715147)](https://bugzilla.gnome.org/show_bug.cgi?id=715147)**
## Description
Hi,
we can't play a mpegts file that does not have a PMT/PAT table because no sink pad is created for such...## Submitted by Eric
**[Link to original bug (#715147)](https://bugzilla.gnome.org/show_bug.cgi?id=715147)**
## Description
Hi,
we can't play a mpegts file that does not have a PMT/PAT table because no sink pad is created for such a media.
here is a link to a sample file :
https://www.dropbox.com/s/ekk4vrzvfymnajv/mpegts_with_not_PAT.mpeg
this file plays ok with mplayer.
Cheers,
Eric
### Depends on
* [Bug 703884](https://bugzilla.gnome.org/show_bug.cgi?id=703884)https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/115androidmedia: enable hw accelerated video decoding and rendering on Ubuntu Touch2021-09-24T14:32:16ZBugzilla Migration Userandroidmedia: enable hw accelerated video decoding and rendering on Ubuntu Touch## Submitted by jam..@..il.com
**[Link to original bug (#711025)](https://bugzilla.gnome.org/show_bug.cgi?id=711025)**
## Description
Created attachment 258319
Patch set
Attached to this bug is a patch set that enables hardwa...## Submitted by jam..@..il.com
**[Link to original bug (#711025)](https://bugzilla.gnome.org/show_bug.cgi?id=711025)**
## Description
Created attachment 258319
Patch set
Attached to this bug is a patch set that enables hardware accelerated video decoding and rendering on the new Ubuntu Touch platform. It makes use of libhybris, which is an adaptive library to make native calls to the Android bionic libc.
A repository with a branch that contains the latest code can also be found here:
https://github.com/jhodapp/gst-plugins-bad
**Attachment 258319**, "Patch set":
[gstamc-patchset.tar.bz2](/uploads/2066dc4fe967ae1b77e6313eb4d8270d/gstamc-patchset.tar.bz2)
Version: 1.2.0https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/114smoothstreaming: Support Clip elements in composite manifests2021-09-24T14:32:15ZBugzilla Migration Usersmoothstreaming: Support Clip elements in composite manifests## Submitted by Antonio Ospite `@ao2`
**[Link to original bug (#710221)](https://bugzilla.gnome.org/show_bug.cgi?id=710221)**
## Description
Hi, Rai.tv serves composite manifests[1] as an entrypoints for its broadcasts, like this on...## Submitted by Antonio Ospite `@ao2`
**[Link to original bug (#710221)](https://bugzilla.gnome.org/show_bug.cgi?id=710221)**
## Description
Hi, Rai.tv serves composite manifests[1] as an entrypoints for its broadcasts, like this one:
http://originsmoothreplay.rai.it/csm_nrp/raidue/Made_in_Sud/madeinsud_puntate_2012/1891183.csm
The main difference with plain manifests is that streams are enclosed in Clip elements, I think I can manage to put together a proof of concept patch which just plays the first clip but I may need help for adding proper support.
Anyone interested?
Thanks, Antonio
[1] http://blogs.iis.net/vsood/archive/2010/01/22/composite-manifest-support-for-rough-cut-editing-scenarios-in-ssme.aspxhttps://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/112mpegtsmux: performance issue2021-09-24T14:32:14ZBugzilla Migration Usermpegtsmux: performance issue## Submitted by Edward Hervey `@bilboed`
**[Link to original bug (#709826)](https://bugzilla.gnome.org/show_bug.cgi?id=709826)**
## Description
Created attachment 256917
callgrind graph of mpegtsmux usage
mpegtsmux does a lot...## Submitted by Edward Hervey `@bilboed`
**[Link to original bug (#709826)](https://bugzilla.gnome.org/show_bug.cgi?id=709826)**
## Description
Created attachment 256917
callgrind graph of mpegtsmux usage
mpegtsmux does a lot of allocation, mapping and adapter usage.
Ideally it should introduce very little overhead (it doesn't even need to read the input data, just mux it).
right now it uses` ~80000` cpu instructions per incoming buffer.
**Attachment 256917**, "callgrind graph of mpegtsmux usage":
![mpegtsmux-callgrind](/uploads/805b3534b5f1c919d24350e8f38f03b4/mpegtsmux-callgrind.png)https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/111h264parse: Add a method to blindly timestamp based on framerate2021-09-24T14:32:14ZBugzilla Migration Userh264parse: Add a method to blindly timestamp based on framerate## Submitted by Robert Swain
**[Link to original bug (#709756)](https://bugzilla.gnome.org/show_bug.cgi?id=709756)**
## Description
Sometimes one may encounter a raw h.264 file that doesn't have any timestamp information. It may be ...## Submitted by Robert Swain
**[Link to original bug (#709756)](https://bugzilla.gnome.org/show_bug.cgi?id=709756)**
## Description
Sometimes one may encounter a raw h.264 file that doesn't have any timestamp information. It may be useful, at least for testing purposes, to be able to apply timestamps to the frames using h264parse.
The idea is that if the sink pad caps state the stream-format=byte-stream and contain a framerate, this is can be used to override the buffer timestamps by starting from zero and adding the frame duration for each frame.https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/110rfbsrc dependency on libX11 automagic2021-09-24T14:32:14ZBugzilla Migration Userrfbsrc dependency on libX11 automagic## Submitted by Gilles Dartiguelongue
**[Link to original bug (#709530)](https://bugzilla.gnome.org/show_bug.cgi?id=709530)**
## Description
I haven't noticed this before but it looks like rfbsrc and eglglessink both depend on libX1...## Submitted by Gilles Dartiguelongue
**[Link to original bug (#709530)](https://bugzilla.gnome.org/show_bug.cgi?id=709530)**
## Description
I haven't noticed this before but it looks like rfbsrc and eglglessink both depend on libX11 optionally.
The issue is that the dependency is automagic [1] and it makes a "dependency-less" plugin (rfbsrc) not so dependency-less according to build environment.
I would propose a patch but it looks like something already provides --with-x and I cannot figure out where it comes from yet.
[1] http://www.gentoo.org/proj/en/qa/automagic.xml
Version: 1.2.0https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/109opencv: new plugin to make smooth slowmotion2021-09-24T14:32:13ZBugzilla Migration Useropencv: new plugin to make smooth slowmotion## Submitted by Mathieu Duponchelle `@meh`
**[Link to original bug (#708545)](https://bugzilla.gnome.org/show_bug.cgi?id=708545)**
## Description
The plugin uses code from slowmovideo, porting it from QT and its own internal decodin...## Submitted by Mathieu Duponchelle `@meh`
**[Link to original bug (#708545)](https://bugzilla.gnome.org/show_bug.cgi?id=708545)**
## Description
The plugin uses code from slowmovideo, porting it from QT and its own internal decoding "framework" based on ffmpeg to gstreamer.
the code is available there :
https://github.com/MathieuDuponchelle/gst-plugins-base/commits/slowmotion
and is up for trial / review, one issue I am aware of and will fix is the fact that it should be in bad/ext/opencv.
example launch line:
gst-launch-1.0 -e uridecodebin uri=file:///any/file ! videoconvert ! videorate rate=0.1 ! video/x-raw, format=RGB ! slowmo ! videoconvert ! theoraenc ! oggmux ! filesink location=somewhere.ogv
The videoconversion is necessary, as the code will only accept either RGB or BGR for now.
The videorate is also necessary, with a patch from that comment:
https://bugzilla.gnome.org/show_bug.cgi?id=699077#c25
Please do the review here, as github throws away commit comments when pushing non fast forward.
There's plenty of room for optimization and cleanup, contributions are more than welcome :)
### Depends on
* [Bug 699077](https://bugzilla.gnome.org/show_bug.cgi?id=699077)https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/107vc1parse: misc things that need doing2021-09-24T14:32:12ZBugzilla Migration Uservc1parse: misc things that need doing## Submitted by Tim Müller `@tpm`
**[Link to original bug (#707123)](https://bugzilla.gnome.org/show_bug.cgi?id=707123)**
## Description
Just dropping this somewhere, since it comes up regularly. Might not be entirely up-to-date.
...## Submitted by Tim Müller `@tpm`
**[Link to original bug (#707123)](https://bugzilla.gnome.org/show_bug.cgi?id=707123)**
## Description
Just dropping this somewhere, since it comes up regularly. Might not be entirely up-to-date.
For vc1parse the following has to be done still:
* baseparse: Add a convert vfunc to baseparse that is called after
check_valid_frame and before parse_frame to allow subclasses to convert
one frame to another stream format. This vfunc should return a list of
buffers.
* vc1parse: Implement convert vfunc for all the million stream formats
* vc1parse: check_valid_frame for BDU stream formats always gets
everything up to and including a complete frame, i.e. seqhdr,
entrypoint, frame, field, slice splitted later
* vc1parse: Add conversion from EBDU to RBDU and the other way around.
For ASF in simple/main profile the content has to be RBDU, for the BDU
stream formats with startcode in the beginning it has to be EBDU. See
the VC1 spec about this, EBDU is RBDU with some byte sequences escaped.
### See also
* [Bug 743948](https://bugzilla.gnome.org/show_bug.cgi?id=743948)https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/105mpegvideoparse: fix timestamps2021-09-24T14:32:12ZBugzilla Migration Usermpegvideoparse: fix timestamps## Submitted by Matej `@Knopp`
**[Link to original bug (#705066)](https://bugzilla.gnome.org/show_bug.cgi?id=705066)**
## Description
Created attachment 250353
Filter that fixes mpeg timestamps
Mpegvideoparse currently doesn'...## Submitted by Matej `@Knopp`
**[Link to original bug (#705066)](https://bugzilla.gnome.org/show_bug.cgi?id=705066)**
## Description
Created attachment 250353
Filter that fixes mpeg timestamps
Mpegvideoparse currently doesn't do anything about fixing broken / missing timestamps. This affects playback for many files in my case. I'm currently using custom filter after mpegvideoparse to do this, but perhaps the functionality should be added to the parser?
If someone could look at the code and confirm I can provide a patch. Although unlike my filter mpegvideoparser probably can't get away with throwing away D frames so this will need to be addressed. Does anyone have a stream with D frames?
~~**Attachment 250353**~~, "Filter that fixes mpeg timestamps":
[MpegTimestampParser.zip](/uploads/00e3b103649f4ce7d878b95888ef3138/MpegTimestampParser.zip)https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/103Add mpeg2 slice header information to GstMpegVideoMeta2021-09-24T14:32:11ZBugzilla Migration UserAdd mpeg2 slice header information to GstMpegVideoMeta## Submitted by Sreerenj Balachandran `@sree`
**[Link to original bug (#704865)](https://bugzilla.gnome.org/show_bug.cgi?id=704865)**
## Description
Created attachment 250112
mpegvideometa: Add slice information to the GstMpegVide...## Submitted by Sreerenj Balachandran `@sree`
**[Link to original bug (#704865)](https://bugzilla.gnome.org/show_bug.cgi?id=704865)**
## Description
Created attachment 250112
mpegvideometa: Add slice information to the GstMpegVideoMeta.
+++ This bug was initially created as a clone of [Bug 691712](https://bugzilla.gnome.org/show_bug.cgi?id=691712) +++
~~**Patch 250112**~~, "mpegvideometa: Add slice information to the GstMpegVideoMeta.":
[0001-mpegvideometa-Add-slice-information-to-the-GstMpegVi.patch](/uploads/5ea6c06f701ab3a944ed8d9f9d0d071d/0001-mpegvideometa-Add-slice-information-to-the-GstMpegVi.patch)
### Blocking
* [Bug 734547](https://bugzilla.gnome.org/show_bug.cgi?id=734547)https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad/-/issues/102mpegvideoparse: Add GstVideoCropMeta2021-09-24T14:32:10ZBugzilla Migration Usermpegvideoparse: Add GstVideoCropMeta## Submitted by Sreerenj Balachandran `@sree`
**[Link to original bug (#704794)](https://bugzilla.gnome.org/show_bug.cgi?id=704794)**
## Description
Created attachment 250011
mpegvideoparse: Add GstVideoCropMeta
If the encode...## Submitted by Sreerenj Balachandran `@sree`
**[Link to original bug (#704794)](https://bugzilla.gnome.org/show_bug.cgi?id=704794)**
## Description
Created attachment 250011
mpegvideoparse: Add GstVideoCropMeta
If the encoded stream has any associated crop rectangle (SequenceDispalyExtension), then send it as GstVideoCropMeta to the downstream.
Similar bugs: https://bugzilla.gnome.org/show_bug.cgi?id=694068
**Patch 250011**, "mpegvideoparse: Add GstVideoCropMeta":
[0001-mpegvideoparse-Add-GstVideoCropMeta.patch](/uploads/d4bf7be1503f8e36ddec55699a9e35c7/0001-mpegvideoparse-Add-GstVideoCropMeta.patch)