Libav 12
12 is an even number so we're going to break things.
Anton: Yes I think it's about time.
Luca: Let's make something disappear.
12-Remove:
* get_buffer
For everything that gets removed, we need a migration path and example
code for that.
more things to remove?
* request channels?
*
On Tue, Sep 16, 2014 at 12:30 AM, Diego Biurrun di...@biurrun.de wrote:
apetag: Improve APE tag size check
On Tue, Sep 16, 2014 at 01:40:24AM +0200, Katerina Barone-Adesi wrote:
The size variable is (correctly) unsigned, but is passed to several
functions
which take signed parameters,
On Tue, Sep 16, 2014 at 3:14 PM, Luca Barbato lu_z...@gentoo.org wrote:
Diego probably didn't read the code below. One of the wishlist items for
plaid is to provide git integration to the point you get:
1- If the patch applies correctly to master.
2- Expand the patchview so it shows you the
On Mon, Aug 18, 2014 at 11:04 AM, Anton Khirnov an...@khirnov.net wrote:
+One specific API issue in libavformat deserves mentioning here. When the
calling
+code uses libavcodec for decoding or encoding and libavformat for
+demuxing/muxing, the standard practice was to use the stream codec
On Fri, Aug 15, 2014 at 1:12 PM, Diego Biurrun di...@biurrun.de wrote:
The same is done for GCC and clang already.
https://lists.libav.org/mailman/listinfo/libav-devel
I agree with making this consistent with GCC and clang.
However now that we're on the subject - it's a pretty useful
On Tue, Aug 12, 2014 at 12:26 PM, Luca Barbato lu_z...@gentoo.org wrote:
And convert it to Markdown.
What's this?! Next we'll be accepting GitHub pull requests :-)
___
libav-devel mailing list
libav-devel@libav.org
On Tue, Aug 12, 2014 at 1:02 PM, Luca Barbato lu_z...@gentoo.org wrote:
On 12/08/14 21:30, Andrew Kelley wrote:
What's this?! Next we'll be accepting GitHub pull requests :-)
I'm thinking about figuring out how to do that actually.
I use it quite a bit and I find it to be convenient. Libav
On Sun, Jul 20, 2014 at 2:02 PM, Reinhard Tartler siret...@gmail.com
wrote:
Another limitation I found https://travis-ci.com/plans:
Your first 100 builds are free!
What happens on the 101st build? Would you be willing to sponsor a travis
plan?
I believe Travis has unlimited free
On Sun, Jul 20, 2014 at 4:52 PM, Reinhard Tartler siret...@gmail.com
wrote:
To answer my own question: I have setup travis to build our release/10
branch:
https://travis-ci.org/siretart/libav
As you can see, access to the internet seems to be available. The
build machine is an amd64 bit
On Sat, Jul 19, 2014 at 7:36 AM, Reinhard Tartler siret...@gmail.com
wrote:
Depending how how smooth and
fast that transition works, it *might* even make it for Ubuntu 14.10
(aka utopic), but that's going to be close, so no promises.
This would make me happy. Please tell me if there is
Looks good to me.
On Fri, Jul 11, 2014 at 5:24 AM, Luca Barbato lu_z...@gentoo.org wrote:
---
libavformat/format.c | 139
++-
libavformat/utils.c | 135
-
2 files changed, 138 insertions(+),
Looks good.
On Fri, Jul 11, 2014 at 5:24 AM, Luca Barbato lu_z...@gentoo.org wrote:
It is generic enough to be reused for other similar comma-separated
fields.
---
libavformat/format.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/libavformat/format.c
What about these mime types?
audio/3gpp, audio/3gpp2, audio/mp4, audio/MP4A-LATM, audio/mpeg4-generic
From https://en.wikipedia.org/wiki/Advanced_Audio_Coding
On Fri, Jul 11, 2014 at 5:24 AM, Luca Barbato lu_z...@gentoo.org wrote:
Speed up probing ADTS live streams that are not frame-aligned
What about audio/x-matroska?
On Fri, Jul 11, 2014 at 5:24 AM, Luca Barbato lu_z...@gentoo.org wrote:
---
libavformat/matroskadec.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/libavformat/matroskadec.c b/libavformat/matroskadec.c
index 22108ad..3e28664 100644
---
On Fri, Jul 11, 2014 at 7:40 AM, Vittorio Giovara
vittorio.giov...@gmail.com wrote:
Also this should be ifdeffed with FF_API_PROBE_MIME because it's a public
type.
Is it possible to bump the SONAME version instead of introducing #ifdefs in
the code? #ifdefs are an endless source of bugs.
On Sat, Jul 19, 2014 at 1:14 PM, Luca Barbato lu_z...@gentoo.org wrote:
On 19/07/14 22:11, Andrew Kelley wrote:
What about these mime types?
audio/3gpp, audio/3gpp2,
Those do not map to other codecs as well?
audio/MP4A-LATM, audio/mpeg4-generic
those might refer to another container
This fixes a segmentation fault because request_frame in fifo.c assumes
that the call to ff_request_frame will populate fifo-root.next.
Before, it was possible for request_frame in af_compand to not do this,
resulting in a null pointer access. Now, request_frame in af_compand
always will return at
Michael found and fixed a bug in this patch:
http://git.videolan.org/?p=ffmpeg.git;a=commit;h=919c320f7226bf873a9148e1db8994745f9d425d
On Mon, May 26, 2014 at 12:04 AM, Luca Barbato lu_z...@gentoo.org wrote:
On 26/05/14 06:04, Andrew Kelley wrote:
Before, header information for ogg format
From: Michael Niedermayer michae...@gmx.at
This corrects the bug that caused the checksums to change in
9767d7c092c890ecc5953452e8a951fd902dd67b
It caused the EOS flag to be set incorrectly; the ogg spec does not
allow it to be set in the middle of a logical bitstream.
Signed-off-by: Andrew
On Wed, May 28, 2014 at 11:56 AM, Martin Storsjö mar...@martin.st wrote:
Do note that your original patch was incomplete, it broke make fate
(which I corrected by updating the checksum in that patch), which I
apparently shouldn't have done. The same goes for this patch, this changes
the
On Wed, May 28, 2014 at 1:04 PM, Luca Barbato lu_z...@gentoo.org wrote:
Mind if I rework that patch? The code is enough confusing w/out adding
more entropy.
Not sure who you're asking. *I* certainly don't mind.
___
libav-devel mailing list
Before, header information for ogg format files was sent with the first
encoded packet.
This patch makes it so that it is possible for API users to differentiate
between headers and encoded audio. This is useful, for example, when creating
an audio stream where you want to send one set of headers
+ * Copyright (c) 2007 Rob Sykes r...@users.sourceforge.net
+ * Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
+ *
+ * This file is part of libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License
...@users.sourceforge.net
+ * Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
+ *
+ * This file is part of libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software
Ping.
On Thu, Feb 20, 2014 at 8:30 PM, Andrew Kelley superjo...@gmail.com wrote:
This patch adds the `compand` audio filter from ffmpeg master branch
(currently at 9026c49c82) adapted to work with libav.
---
Here is the diff from the previous patch I just sent to the mailing list:
http
Thanks for the review. Comments below:
On Sun, Feb 23, 2014 at 4:03 PM, Justin Ruggles justin.rugg...@gmail.comwrote:
+frame = ff_get_audio_buffer(outlink, FFMIN(2048, s-delay_count));
Why 2048?
Honestly I don't know. I'll spend some time trying to understand.
You can use doubles
+ * Copyright (c) 2007 Rob Sykes r...@users.sourceforge.net
+ * Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
+ *
+ * This file is part of libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License
+ * Copyright (c) 1999 Nick Bailey
+ * Copyright (c) 2007 Rob Sykes r...@users.sourceforge.net
+ * Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
+ *
+ * This file is part of libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under the terms
Bagwell
+ * Copyright (c) 1999 Nick Bailey
+ * Copyright (c) 2007 Rob Sykes r...@users.sourceforge.net
+ * Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
+ *
+ * This file is part of libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under the terms
---
Nothing is changed here from the last patch. I am sending it again just so
that it will not be accidentally forgotten.
libavfilter/af_compand.c | 11 ++-
1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/libavfilter/af_compand.c b/libavfilter/af_compand.c
index
Chris Bagwell
+ * Copyright (c) 1999 Nick Bailey
+ * Copyright (c) 2007 Rob Sykes r...@users.sourceforge.net
+ * Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
+ *
+ * This file is part of libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under
Nick Bailey
+ * Copyright (c) 2007 Rob Sykes r...@users.sourceforge.net
+ * Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
+ *
+ * This file is part of libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General
---
libavfilter/af_compand.c | 11 ++-
1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/libavfilter/af_compand.c b/libavfilter/af_compand.c
index 421bd04..9ef658f 100644
--- a/libavfilter/af_compand.c
+++ b/libavfilter/af_compand.c
@@ -53,6 +53,7 @@ typedef struct
(c) 1999 Chris Bagwell
+ * Copyright (c) 1999 Nick Bailey
+ * Copyright (c) 2007 Rob Sykes r...@users.sourceforge.net
+ * Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
+ *
+ * This file is part of libav.
+ *
+ * Libav is free software; you can redistribute
On Thu, Feb 20, 2014 at 8:14 PM, Luca Barbato lu_z...@gentoo.org wrote:
On 21/02/14 01:56, Andrew Kelley wrote:
-cs = s-segments[i];
+cs = s-segments[i - 1];
before cs was all the segments but the first, now it is all the segments
but the last. Is it correct?
Not quite - notice
is welcome.
On Fri, Feb 14, 2014 at 2:56 AM, Andrew Kelley superjo...@gmail.com wrote:
This patch adds the `compand` audio filter from ffmpeg master branch
(currently at 7f0f47b3df) adapted to work with libav.
The following changes are made:
* use float instead of double
* use strtok_r instead
Chris Bagwell
+ * Copyright (c) 1999 Nick Bailey
+ * Copyright (c) 2007 Rob Sykes r...@users.sourceforge.net
+ * Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
+ *
+ * This file is part of libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under
Mahol
+ * Copyright (c) 2014 Andrew Kelley
+ *
+ * This file is part of libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License
On Mon, Feb 10, 2014 at 7:13 AM, Diego Biurrun di...@biurrun.de wrote:
On Sat, Feb 08, 2014 at 08:59:33PM -0500, Andrew Kelley wrote:
--- a/doc/filters.texi
+++ b/doc/filters.texi
@@ -467,6 +467,80 @@ To fix a 5.1 WAV improperly encoded in AAC's native
channel order
+
+@item attacks
On Sat, Feb 8, 2014 at 11:22 PM, Luca Barbato lu_z...@gentoo.org wrote:
In short, do not use av_assert* in new code you are committing to the
mainline.
What should I do about the assert in av_clipd_c?
/**
* Clip a double value into the amin-amax range.
* @param a value to clip
*
Apologies for missing this.
On Mon, Feb 10, 2014 at 1:48 AM, Anton Khirnov an...@khirnov.net wrote:
Since b0c2c09, vorbis would be used if libvorbis is compiled in.
---
tests/lavf-regression.sh |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/lavf-regression.sh
Since 2007, the Xipth.org Foundation recommends that .ogg only be used
for Ogg Vorbis audio files.
Source: http://wiki.xiph.org/index.php/MIME_Types_and_File_Extensions
However we only do it if we have libvorbis available because the
built in vorbis encoder is not very good.
---
argh, my mistake. fixing.
On Sat, Feb 8, 2014 at 12:17 PM, Janne Grunau janne-li...@jannau.netwrote:
On 2014-02-08 18:14:23 +0100, Luca Barbato wrote:
On 08/02/14 17:53, Andrew Kelley wrote:
Since 2007, the Xipth.org Foundation recommends that .ogg only be used
Since 2007, the Xiph.org Foundation recommends that .ogg only be used
for Ogg Vorbis audio files.
Source: http://wiki.xiph.org/index.php/MIME_Types_and_File_Extensions
However we only do it if we have libvorbis available because the
built in vorbis encoder is not very good.
---
..55d75af
--- /dev/null
+++ b/libavfilter/af_compand.c
@@ -0,0 +1,592 @@
+/*
+ * Copyright (c) 1999 Chris Bagwell
+ * Copyright (c) 1999 Nick Bailey
+ * Copyright (c) 2007 Rob Sykes r...@users.sourceforge.net
+ * Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
+ *
+ * This file
@@
* Copyright (c) 1999 Nick Bailey
* Copyright (c) 2007 Rob Sykes r...@users.sourceforge.net
* Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
*
- * This file is part of FFmpeg.
+ * This file is part of libav.
*
* FFmpeg is free software; you can redistribute
index 000..6308cda
--- /dev/null
+++ b/libavfilter/af_compand.c
@@ -0,0 +1,595 @@
+/*
+ * Copyright (c) 1999 Chris Bagwell
+ * Copyright (c) 1999 Nick Bailey
+ * Copyright (c) 2007 Rob Sykes r...@users.sourceforge.net
+ * Copyright (c) 2013 Paul B Mahol
+ * Copyright (c) 2014 Andrew Kelley
On Sat, Feb 8, 2014 at 10:53 PM, Luca Barbato lu_z...@gentoo.org wrote:
On 09/02/14 02:59, Andrew Kelley wrote:
+av_assert2(channels 0); /* would corrupt delay_count and
delay_index */
Reachable asserts are BUGS and SECURITY ISSUES.
This is not reachable in release code. From
On Sat, Feb 8, 2014 at 11:22 PM, Luca Barbato lu_z...@gentoo.org wrote:
In short, do not use av_assert* in new code you are committing to the
mainline.
OK no problem.
Probably a good chunk of it lives already in avresample or could be part
of avresample, I'd let Justin tell more about
Since 2007, the Xipth.org Foundation recommends that .ogg only be used
for Ogg Vorbis audio files.
Source: http://wiki.xiph.org/index.php/MIME_Types_and_File_Extensions
---
libavformat/oggenc.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/libavformat/oggenc.c
On Fri, Feb 7, 2014 at 1:07 PM, Justin Ruggles justin.rugg...@gmail.comwrote:
As you can see from .extensions, the ogg muxer isn't only used for .ogg
files (although maybe we could add .oga to that list). FLAC was chosen for
the default because it is lossless and we have a good native encoder.
On Fri, Feb 7, 2014 at 2:00 PM, Luca Barbato lu_z...@gentoo.org wrote:
We could enable vorbis output if the libvorbis encoder is available,
otherwise fallback to flac.
Sounds like a good compromise. And since the debian package of libav
depends on libvorbis, this will effectively produce the
Since 2007, the Xipth.org Foundation recommends that .ogg only be used
for Ogg Vorbis audio files.
Source: http://wiki.xiph.org/index.php/MIME_Types_and_File_Extensions
However we only do it if we have libvorbis available because the
built in vorbis encoder is not very good.
---
Vladimir - any comments on this patch?
On Mon, Feb 3, 2014 at 3:21 AM, Luca Barbato lu_z...@gentoo.org wrote:
On 03/02/14 01:05, Andrew Kelley wrote:
This is a workaround for the WMA decoder which incorrectly seeks to
0 for some streams.
Bug-id: 43
---
libavformat/asfdec.c | 6
This is a workaround for the WMA decoder which incorrectly seeks to
0 for some streams.
Bug-id: 43
---
libavformat/asfdec.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/libavformat/asfdec.c b/libavformat/asfdec.c
index 5b4366e..0505269 100644
--- a/libavformat/asfdec.c
+++
is possible.
On Sun, Feb 2, 2014 at 4:25 PM, Andrew Kelley superjo...@gmail.com wrote:
This is a workaround for the WMA decoder which incorrectly seeks to
0 for some streams.
Bug-id: 43
---
libavformat/asfdec.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/libavformat/asfdec.c b
This is a workaround for the WMA decoder which incorrectly seeks to
0 for some streams.
Bug-id: 43
---
libavformat/asfdec.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/libavformat/asfdec.c b/libavformat/asfdec.c
index 5b4366e..87b1f62 100644
--- a/libavformat/asfdec.c
+++
On Sun, Feb 2, 2014 at 6:59 PM, Vittorio Giovara vittorio.giov...@gmail.com
wrote:
!pts
OK done.
I guess this is stable worthy?
I guess so. Unless asf files can contain timestamps, i.e. danger of start
time not being 0. I've never seen such a file.
why do the temporary workaround and not skip straight to the proper
solution?
On Mon, Dec 9, 2013 at 6:11 AM, Anton Khirnov an...@khirnov.net wrote:
This is a temporary workaround to allow deprecating
avcodec_get_frame_defaults(). The proper solution will be using a
properly allocated
Makes sense. Thanks for the explanation.
On Wed, Dec 11, 2013 at 3:33 PM, Anton Khirnov an...@khirnov.net wrote:
On Wed, 11 Dec 2013 15:20:48 -0500, Andrew Kelley superjo...@gmail.com
wrote:
why do the temporary workaround and not skip straight to the proper
solution?
Because it's
Hi, I just ran into this bug today:
https://bugzilla.libav.org/show_bug.cgi?id=596
I'm about to start troubleshooting this bug and contributing a fix but
first I thought I'd see whether anyone has some clues of what I might look
into first.
Thanks.
___
On Sat, Nov 23, 2013 at 12:21 AM, Justin Ruggles justin.rugg...@gmx.comwrote:
My first suspicion is that it is an application issue since libavformat is
not meant to be multi-threaded.
Are you sure about that? This is from libavcodec/avcodec.h:
/**
* Register a user provided lock manager
By the way, are you sure this is enough?
I have tried this exact same configuration and still got some
undefined link errors on my application.
By no means do I consider myself especially knowledgeable on the subject;
this is what worked for me on my Ubuntu system.
What system are you using
seeking to 0 in this .wma
filehttp://superjoe.s3.amazonaws.com/temp/beverly-hills.wmacauses
glitchy playback.
https://bugzilla.libav.org/show_bug.cgi?id=567
___
libav-devel mailing list
libav-devel@libav.org
, Andrew Kelley superjo...@gmail.com wrote:
Note: this example program is affected by this libavfilter bug:
https://bugzilla.libav.org/show_bug.cgi?id=560
On Mon, Sep 16, 2013 at 2:42 PM, Andrew Kelley superjo...@gmail.comwrote:
It sets up a filter chain, decodes audio into the buffersrc
against libav statically and the use of -Bsymbolic
---
doc/platform.texi | 10 ++
1 file changed, 10 insertions(+)
diff --git a/doc/platform.texi b/doc/platform.texi
index 2a7dd45..e4b0c7e 100644
--- a/doc/platform.texi
+++ b/doc/platform.texi
@@ -24,6 +24,16 @@ If not, then you should
Note: this example program is affected by this libavfilter bug:
https://bugzilla.libav.org/show_bug.cgi?id=560
On Mon, Sep 16, 2013 at 2:42 PM, Andrew Kelley superjo...@gmail.com wrote:
It sets up a filter chain, decodes audio into the buffersrc,
and plays audio from the buffersink using
a/libavfilter/api-example.c b/libavfilter/api-example.c
new file mode 100644
index 000..df5073f
--- /dev/null
+++ b/libavfilter/api-example.c
@@ -0,0 +1,321 @@
+/*
+ * copyright (c) 2013 Andrew Kelley
+ *
+ * This file is part of Libav.
+ *
+ * Libav is free software; you can redistribute
I tried using the supposedly anonymous ftp server but it wouldn't let me do
anything without logging in. I also tried filing a bug report with bugzilla
but it's been 10 minutes since it was supposed to send a email to get my
account going.
So I'm mailing libav-devel.
This ogg file:
For what it's worth, it appears that this issue exists for every ogg
downloaded from Bandcamp http://bandcamp.com/.
On Sun, Sep 15, 2013 at 3:41 AM, Andrew Kelley superjo...@gmail.com wrote:
I tried using the supposedly anonymous ftp server but it wouldn't let me
do anything without logging
avprobe can see the metadata too. It seems I made a mistake of assuming
that all the metadata would be in the AVFormatContext. It works if I check
the metadata of all the AVStreams too. Sorry about the false positive.
___
libav-devel mailing list
71 matches
Mail list logo