Re: [FFmpeg-devel] [PATCH] define AVPixelFormat aliases as enumerators instead of macros
On Fri, 15 Jan 2016 10:39:59 -0800
Richard Smith wrote:
> On Fri Jan 15 08:51:07 CET 2016 wm4 wrote;
> > On Thu, 14 Jan 2016 13:58:14 -0800 Richard Smith
> > wrote:
> > > libavutil/pixfmt.h defines a collection of endian-specific pixel formats
> > > as
> > > macros. These macro names can cause conflicts with external projects that
> > > use those identifiers for their own purposes. Here's a patch to define
> > > these aliases as enumerators instead of macros, please consider merging:
> > >
> > >
> > > https://github.com/zygoloid/FFmpeg/commit/c20a0e2e66e52c45b9193bc750165b7ecf7f3ca4
> > >
> > > (Note that AV_PIX_FMT_Y400A was already defined as an enumerator in the
> > > PixelFormat enumeration, so I deleted its (no-op) macro entirely.)
> >
> > API users might check for the existence of such pixfmts with #ifdef,
>
> That would be a very odd thing for them to do, as most pixfmts do not
> have #defines.
>
> > and I don't understand the reasoning behind your patch. Why would
> > external projects redefine these macros?
>
> The project in question has its own enumeration:
>
> namespace MyProject {
> enum PixelFormatToUse {
> // ... some other values ...
> AV_PIX_FMT_RGB32, // use ffmpeg's AV_PIX_FMT_RGB32
> // ...
> };
> }
Then maybe it shouldn't do that, or if it does, this source file
shouldn't include any ffmpeg related headers. This is just asking for
trouble big-time. Use a different prefix if you really want to do this.
FFmpeg uses preprocessor defines for a lot of things, and you just have
to expect that FFmpeg will add macros prefixed with AV_ as it pleases.
> The names are intentionally chosen to be in 1-1 correspondence with
> ffmpeg's names. But ffmpeg's macro sometimes renames this project's
> enumerator, depending on whether its header is included before that
> file.
This is not a question of what should be, it's question of
compatibility. We try not to break downstream projects unnecessarily.
I'd just suggest defining aliases with names that do not violate
ffmpeg's namespace.
If you want to do this change, you IMHO need to keep the defines under a
compatibility ifdef, that will remain active until the next major
library bump. (Major bumps are when we change the API/ABI in
incompatible ways. Look e.g. into libavutil/version.h for a bunch of
APIs that will be removed or changed on the next bump.)
I don't know if others shares my opinion whether compatibility is
needed here.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] remove the deprecated avpicture_get_size() function
On Fri, 15 Jan 2016 23:34:01 -0800 Eddie Hao wrote: > --- > libavcodec/avcodec.h | 6 -- > libavcodec/avpicture.c | 5 - > libavcodec/libutvideodec.cpp | 3 ++- > libavcodec/libutvideoenc.cpp | 3 ++- > libavdevice/decklink_dec.cpp | 2 +- > libavdevice/lavfi.c | 3 ++- > libavformat/frmdec.c | 3 ++- > 7 files changed, 9 insertions(+), 16 deletions(-) > > diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h > index f365775..6fab787 100644 > --- a/libavcodec/avcodec.h > +++ b/libavcodec/avcodec.h > @@ -4908,12 +4908,6 @@ int avpicture_layout(const AVPicture *src, enum > AVPixelFormat pix_fmt, > unsigned char *dest, int dest_size); > > /** > - * @deprecated use av_image_get_buffer_size() instead. > - */ > -attribute_deprecated > -int avpicture_get_size(enum AVPixelFormat pix_fmt, int width, int height); > - > -/** Replacing uses of this function is welcome, but it can't be fully removed yet. This would break API users. It has to happen on the next major bump instead. ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] lavc/rawdec: Use AV_PIX_FMT_PAL8 for 1-bit raw QuickTime video
On Wed, Jan 13, 2016 at 01:42:24PM +0100, Mats Peterson wrote: > Match the use of AV_PIX_FMT_PAL8 for 1-bit QuickTime Animation in > lavc/qtrle. To reiterate, 1-bit video is not necessary black & white in > QuickTime, merely bi-level. The two colors can be any color. The palette, > either included in the sample description, or the default Macintosh > palette (black & white for 1-bit video) will be set in lavf/qtpalette. > See the QuickTime File Format Specification for details. > > Mats > raw.c|4 ++-- > rawdec.c | 28 > 2 files changed, 22 insertions(+), 10 deletions(-) > de416185ecb440201cd8235d6e8dbb01b733aec9 > 0001-lavf-rawdec-Use-AV_PIX_FMT_PAL8-for-1-bit-QuickTime-.patch > From d1f62ac4a51329eff0c12fbab19ac8f22329d791 Mon Sep 17 00:00:00 2001 > From: Mats Peterson > Date: Wed, 13 Jan 2016 13:29:13 +0100 > Subject: [PATCH] lavc/rawdec: Use AV_PIX_FMT_PAL8 for 1-bit raw QuickTime > video > > Match the use of AV_PIX_FMT_PAL8 for 1-bit QuickTime Animation in > lavc/qtrle. To reiterate, 1-bit video is not necessary black & white in > QuickTime, merely bi-level. The two colors can be any color. The palette, > either included in the sample description, or the default Macintosh > palette (black & white for 1-bit video) will be set in lavf/qtpalette. > See the QuickTime File Format Specification for details. breaks https://trac.ffmpeg.org/raw-attachment/ticket/823/1bpp.mov and https://trac.ffmpeg.org/raw-attachment/ticket/823/1bpp_129.mov do you have a file that this fixes ? [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB No human being will ever know the Truth, for even if they happen to say it by chance, they would not even known they had done so. -- Xenophanes signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] lavc/rawdec: Use AV_PIX_FMT_PAL8 for 1-bit raw QuickTime video
On 01/16/2016 12:10 PM, Michael Niedermayer wrote: breaks https://trac.ffmpeg.org/raw-attachment/ticket/823/1bpp.mov and https://trac.ffmpeg.org/raw-attachment/ticket/823/1bpp_129.mov do you have a file that this fixes ? Interesting. Well, I don't have a file that gets "fixed" in any way, but I do have a 1 bpp raw file that works OK here: https://drive.google.com/open?id=0B3_pEBoLs0fabm83dG5YLUsybDA ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] lavc/rawdec: Use AV_PIX_FMT_PAL8 for 1-bit raw QuickTime video
On 01/16/2016 12:10 PM, Michael Niedermayer wrote: breaks https://trac.ffmpeg.org/raw-attachment/ticket/823/1bpp.mov and https://trac.ffmpeg.org/raw-attachment/ticket/823/1bpp_129.mov do you have a file that this fixes ? Could it have something to do with the "linesize_align" variable that is 4 by default? I have no idea. Mats ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] lavc/rawdec: Use AV_PIX_FMT_PAL8 for 1-bit raw QuickTime video
On 01/16/2016 12:24 PM, Mats Peterson wrote: Could it have something to do with the "linesize_align" variable that is 4 by default? I have no idea. I suppose since the width of both these files is an odd value, there has to be some alignment to even boundaries. I'll try various values of linesize_align. Mats ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] lavf/matroskadec: Get sample size from private data
On Fri, Jan 15, 2016 at 11:31:49PM +0100, Mats Peterson wrote: > Since track->audio.bitdepth is zero for A_QUICKTIME, the sample size > has to be retrieved from the private data. Also, 'twos' and 'sowt' > audio can be signed 8-bit. > > Mats > matroskadec.c | 10 -- > 1 file changed, 8 insertions(+), 2 deletions(-) > f0b9dab8756fd89584300e3c934c3149e1983ed8 > 0001-lavf-matroskadec-Get-sample-size-from-private-data.patch > From dbddf5d5fba7992e52bb25db214e7b4debfb8023 Mon Sep 17 00:00:00 2001 > From: Mats Peterson > Date: Fri, 15 Jan 2016 23:25:38 +0100 > Subject: [PATCH] lavf/matroskadec: Get sample size from private data applied thx [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB Many things microsoft did are stupid, but not doing something just because microsoft did it is even more stupid. If everything ms did were stupid they would be bankrupt already. signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] lavf/mov: Don't limit fourcc 0 -> raw/twos to version 0 sample descriptions
On Fri, Jan 15, 2016 at 11:35:08PM +0100, Mats Peterson wrote: > I had the notion that the fourcc 0x to raw/twos mapping was > only valid for version 0 sound sample descriptions. However, the > documentation is quite foggy regarding this subject, so it's better > to do this mapping regardless of sample description version. > > Mats > > -- > Mats Peterson > http://matsp888.no-ip.org/~mats/ > mov.c |2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > 8e959656435f6511d5fa0efa4bacf963bda828d2 > 0001-lavf-mov-Don-t-limit-fourcc-0-raw-twos-to-version-0-.patch > From 9e3354877146b29a30810a886342145c74d1a7bd Mon Sep 17 00:00:00 2001 > From: Mats Peterson > Date: Fri, 15 Jan 2016 23:28:06 +0100 > Subject: [PATCH] lavf/mov: Don't limit fourcc 0 -> raw/twos to version 0 > sample descriptions applied thanks [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB Those who would give up essential Liberty, to purchase a little temporary Safety, deserve neither Liberty nor Safety -- Benjamin Franklin signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] mov: Add an option to toggle dref opening
On Fri, Jan 15, 2016 at 05:03:49PM +, Derek Buitenhuis wrote: > This feature is mostly only used by NLE software, and is > both of dubious value being enabled by default, and a > possible security risk. > > Signed-off-by: Derek Buitenhuis > --- > libavformat/isom.h| 1 + > libavformat/mov.c | 22 +- > libavformat/version.h | 4 ++-- > 3 files changed, 20 insertions(+), 7 deletions(-) i wonder if this should not be a generic option for all demuxers or maybe a whitelist of what pathes are allowed to be opened, maybe similar to the existing codec/format whitelists but thats not a objection to the dref option here ... [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB The educated differ from the uneducated as much as the living from the dead. -- Aristotle signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] News entry for FFmpeg 2.8.5, 2.7.5, 2.6.7, 2.5.10
On Fri, Jan 15, 2016 at 09:22:27PM -0500, Ganesh Ajjanagadde wrote: > On Fri, Jan 15, 2016 at 9:15 PM, Michael Niedermayer wrote: > > From: Michael Niedermayer > > > > --- > > src/index | 14 ++ > > 1 file changed, 14 insertions(+) > > > > diff --git a/src/index b/src/index > > index 554ad0e..56a7956 100644 > > --- a/src/index > > +++ b/src/index > > @@ -37,6 +37,20 @@ > > News > > > > > > + January 16, 2015, FFmpeg 2.8.5, 2.7.5, 2.6.7, > > 2.5.10 > > + > > +We have made several new point releases ( > href="download.html#release_2.8">2.8.5, > > + 2.7.5, > > + 2.6.7, > > + 2.5.10). > > +They fix various bugs, as well as CVE-2016-1897 and CVE-2016-1898. > > +Please see the changelog for more details. > > Super minor, but filename is Changelog, and/or it can be ref'ed like > in the 2.8 major entry. the text is copy and pasted from previous announcments where it was lowercase. I think it was intended to refer to the english word "changelog" not the file linking isnt trivial because its not one changelog, each release has its own changelog so it would require 4 links, ive reworded it slightly to be a bit clearer pushed thanks [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB The real ebay dictionary, page 2 "100% positive feedback" - "All either got their money back or didnt complain" "Best seller ever, very honest" - "Seller refunded buyer after failed scam" signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 2/2] fate: add fixed-dsp test
Hi, On Fri, Jan 15, 2016 at 8:10 PM, James Almer wrote: > Adapted from float-dsp So, it seems this just compares C vs. SIMD, right? Shouldn't we use checkasm for that? Ronald ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] lavf/mov: Don't limit fourcc 0 -> raw/twos to version 0 sample descriptions
On 01/16/2016 02:17 PM, Michael Niedermayer wrote: applied thanks Thanks, Michael. ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] lavf/matroskadec: Get sample size from private data
On 01/16/2016 02:16 PM, Michael Niedermayer wrote: applied thx Thanks, Michael. ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] mov: Add an option to toggle dref opening
On Sat, 16 Jan 2016 14:22:21 +0100 Michael Niedermayer wrote: > On Fri, Jan 15, 2016 at 05:03:49PM +, Derek Buitenhuis wrote: > > This feature is mostly only used by NLE software, and is > > both of dubious value being enabled by default, and a > > possible security risk. > > > > Signed-off-by: Derek Buitenhuis > > --- > > libavformat/isom.h| 1 + > > libavformat/mov.c | 22 +- > > libavformat/version.h | 4 ++-- > > 3 files changed, 20 insertions(+), 7 deletions(-) > > i wonder if this should not be a generic option for all demuxers > or maybe a whitelist of what pathes are allowed to be opened, maybe > similar to the existing codec/format whitelists > > but thats not a objection to the dref option here ... > > [...] There's a AVFormatContext.open_cb callback, which an API user can use to decide whether opening a certain URL is fine. (Unfortunately, HLS doesn't use it, but mov does.) It might be fine to make opening as strict as possible (if that callback is not set, which can be used to override it). ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 2/2] fate: add fixed-dsp test
On 1/16/2016 10:43 AM, Ronald S. Bultje wrote: > Hi, > > On Fri, Jan 15, 2016 at 8:10 PM, James Almer wrote: > >> Adapted from float-dsp > > > So, it seems this just compares C vs. SIMD, right? Shouldn't we use > checkasm for that? > > Ronald I guess that'd be best, yeah. I'll port this one and let someone else port the float-dsp one to checkasm. ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] lavc/rawdec: Use AV_PIX_FMT_PAL8 for 1-bit raw QuickTime video
On 01/16/2016 12:10 PM, Michael Niedermayer wrote: do you have a file that this fixes ? I do have a file, *now*. I've added a palette with blue color to a raw 1-bit QuickTime file just to have something to work on. https://drive.google.com/open?id=0B3_pEBoLs0faSDhIMWdmeFJtcDg ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH] avformat/hls: Require the file extension to be m3u / m3u8 for probing to succeed
From: Michael Niedermayer If the filename isnt set by the user application then the code behaves like before Signed-off-by: Michael Niedermayer --- libavformat/hls.c |4 1 file changed, 4 insertions(+) diff --git a/libavformat/hls.c b/libavformat/hls.c index 7a8610c..846d884 100644 --- a/libavformat/hls.c +++ b/libavformat/hls.c @@ -1983,6 +1983,10 @@ static int hls_probe(AVProbeData *p) * somewhere for a proper match. */ if (strncmp(p->buf, "#EXTM3U", 7)) return 0; + +if (p->filename && !av_match_ext(p->filename, "m3u8,m3u")) +return 0; + if (strstr(p->buf, "#EXT-X-STREAM-INF:") || strstr(p->buf, "#EXT-X-TARGETDURATION:") || strstr(p->buf, "#EXT-X-MEDIA-SEQUENCE:")) -- 1.7.9.5 ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH] videodsp: fix 1-byte overread in top/bottom READ_NUM_BYTES iterations.
This can overread (either before start or beyond end) of the buffer in Nx1 (i.e. height=1) images. Fixes mozilla bug 1240080. --- libavcodec/x86/videodsp.asm | 21 ++--- 1 file changed, 6 insertions(+), 15 deletions(-) diff --git a/libavcodec/x86/videodsp.asm b/libavcodec/x86/videodsp.asm index 48f5ac0..a807d3b 100644 --- a/libavcodec/x86/videodsp.asm +++ b/libavcodec/x86/videodsp.asm @@ -193,14 +193,10 @@ hvar_fn movvalb, [srcq+%2-1] %elif (%2-%%off) == 2 movvalw, [srcq+%2-2] -%elifidn %1, body +%else movvalb, [srcq+%2-1] -salvald, 16 +rorvald, 16 movvalw, [srcq+%2-3] -%elifidn %1, bottom -movd mm %+ %%mmx_idx, [srcq+%2-4] -%else ; top -movd mm %+ %%mmx_idx, [srcq+%2-3] %endif %endif ; (%2-%%off) >= 1 %endmacro ; READ_NUM_BYTES @@ -253,18 +249,13 @@ hvar_fn mov [dstq+%2-1], valb %elif (%2-%%off) == 2 mov [dstq+%2-2], valw -%elifidn %1, body -mov [dstq+%2-3], valw -sarvald, 16 -mov [dstq+%2-1], valb %else -movd vald, mm %+ %%mmx_idx -%ifidn %1, bottom -sarvald, 8 -%endif mov [dstq+%2-3], valw -sarvald, 16 +rorvald, 16 mov [dstq+%2-1], valb +%ifnidn %1, body +rorvald, 16 +%endif %endif %endif ; (%2-%%off) >= 1 %endmacro ; WRITE_NUM_BYTES -- 2.1.2 ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] mov: Add an option to toggle dref opening
On Sat, 16 Jan 2016, wm4 wrote: On Sat, 16 Jan 2016 14:22:21 +0100 Michael Niedermayer wrote: On Fri, Jan 15, 2016 at 05:03:49PM +, Derek Buitenhuis wrote: > This feature is mostly only used by NLE software, and is > both of dubious value being enabled by default, and a > possible security risk. > > Signed-off-by: Derek Buitenhuis > --- > libavformat/isom.h| 1 + > libavformat/mov.c | 22 +- > libavformat/version.h | 4 ++-- > 3 files changed, 20 insertions(+), 7 deletions(-) i wonder if this should not be a generic option for all demuxers or maybe a whitelist of what pathes are allowed to be opened, maybe similar to the existing codec/format whitelists but thats not a objection to the dref option here ... [...] There's a AVFormatContext.open_cb callback, which an API user can use to decide whether opening a certain URL is fine. (Unfortunately, HLS doesn't use it, but mov does.) It might be fine to make opening as strict as possible (if that callback is not set, which can be used to override it). CLI users still going to need an option for it. Regards, Marton ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH] avfilter: add afftfilter
Signed-off-by: Paul B Mahol
---
configure | 3 +
libavfilter/Makefile | 1 +
libavfilter/af_afftfilt.c | 363 ++
libavfilter/allfilters.c | 1 +
4 files changed, 368 insertions(+)
create mode 100644 libavfilter/af_afftfilt.c
diff --git a/configure b/configure
index 7cef6f5..ec0b350 100755
--- a/configure
+++ b/configure
@@ -2838,6 +2838,8 @@ unix_protocol_deps="sys_un_h"
unix_protocol_select="network"
# filters
+afftfilt_filter_deps="avcodec"
+afftfilt_filter_select="fft"
amovie_filter_deps="avcodec avformat"
aresample_filter_deps="swresample"
ass_filter_deps="libass"
@@ -6061,6 +6063,7 @@ done
enabled zlib && add_cppflags -DZLIB_CONST
# conditional library dependencies, in linking order
+enabled afftfilt_filter && prepend avfilter_deps "avcodec"
enabled amovie_filter && prepend avfilter_deps "avformat avcodec"
enabled aresample_filter&& prepend avfilter_deps "swresample"
enabled asyncts_filter && prepend avfilter_deps "avresample"
diff --git a/libavfilter/Makefile b/libavfilter/Makefile
index e3e3561..242f56d 100644
--- a/libavfilter/Makefile
+++ b/libavfilter/Makefile
@@ -29,6 +29,7 @@ OBJS-$(CONFIG_ACROSSFADE_FILTER) += af_afade.o
OBJS-$(CONFIG_ADELAY_FILTER) += af_adelay.o
OBJS-$(CONFIG_AECHO_FILTER) += af_aecho.o
OBJS-$(CONFIG_AEMPHASIS_FILTER) += af_aemphasis.o
+OBJS-$(CONFIG_AFFTFILT_FILTER) += af_afftfilt.o window_func.o
OBJS-$(CONFIG_ANEQUALIZER_FILTER)+= af_anequalizer.o
OBJS-$(CONFIG_AEVAL_FILTER) += aeval.o
OBJS-$(CONFIG_AFADE_FILTER) += af_afade.o
diff --git a/libavfilter/af_afftfilt.c b/libavfilter/af_afftfilt.c
new file mode 100644
index 000..1a8828a
--- /dev/null
+++ b/libavfilter/af_afftfilt.c
@@ -0,0 +1,363 @@
+/*
+ * Copyright (c) 2016 Paul B Mahol
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg 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,
+ * or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include "libavutil/audio_fifo.h"
+#include "libavutil/avstring.h"
+#include "libavfilter/internal.h"
+#include "libavutil/common.h"
+#include "libavutil/opt.h"
+#include "libavcodec/avfft.h"
+#include "libavutil/eval.h"
+#include "audio.h"
+#include "window_func.h"
+
+typedef struct AFFTFiltContext {
+const AVClass *class;
+char *expr_str;
+int fft_bits;
+
+FFTContext *fft, *ifft;
+FFTComplex **fft_data;
+int nb_exprs;
+int window_size;
+AVExpr **expr;
+AVAudioFifo *fifo;
+int64_t pts;
+int hop_size;
+float overlap;
+AVFrame *buffer;
+int start, end;
+int win_func;
+float *window_func_lut;
+} AFFTFiltContext;
+
+static const char *const var_names[] = {"sr", "b",
"ch","chs", "pts",NULL };
+enum { VAR_SAMPLE_RATE, VAR_BIN,
VAR_CHANNEL, VAR_CHANNELS, VAR_PTS, VAR_VARS_NB };
+
+#define OFFSET(x) offsetof(AFFTFiltContext, x)
+#define A AV_OPT_FLAG_AUDIO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
+
+static const AVOption afftfilt_options[] = {
+{ "exprs", "set channels expressions", OFFSET(expr_str),
AV_OPT_TYPE_STRING, {.str = "1" }, 0, 0, A },
+{ "win_size", "set window size", OFFSET(fft_bits), AV_OPT_TYPE_INT,
{.i64=12}, 4, 16, A, "fft" },
+{ "w16",0, 0, AV_OPT_TYPE_CONST, {.i64=4}, 0, 0, A, "fft" },
+{ "w32",0, 0, AV_OPT_TYPE_CONST, {.i64=5}, 0, 0, A, "fft" },
+{ "w64",0, 0, AV_OPT_TYPE_CONST, {.i64=6}, 0, 0, A, "fft" },
+{ "w128", 0, 0, AV_OPT_TYPE_CONST, {.i64=7}, 0, 0, A, "fft" },
+{ "w256", 0, 0, AV_OPT_TYPE_CONST, {.i64=8}, 0, 0, A, "fft" },
+{ "w512", 0, 0, AV_OPT_TYPE_CONST, {.i64=9}, 0, 0, A, "fft" },
+{ "w1024", 0, 0, AV_OPT_TYPE_CONST, {.i64=10}, 0, 0, A, "fft" },
+{ "w2048", 0, 0, AV_OPT_TYPE_CONST, {.i64=11}, 0, 0, A, "fft" },
+{ "w4096", 0, 0, AV_OPT_TYPE_CONST, {.i64=12}, 0, 0, A, "fft" },
+{ "w8192", 0, 0, AV_OPT_TYPE_CONST, {.i64=13}, 0, 0, A, "fft" },
+{ "w16384", 0, 0, AV_OPT_TYPE_CONST, {.i64=14}, 0, 0, A, "fft" },
+{ "w32768", 0, 0, AV_OPT_TYPE_CONST, {.i64=15}, 0, 0, A, "fft" },
+{ "w65536", 0, 0, AV_OPT_TYPE_CONST, {.i64=16}, 0, 0, A, "fft" },
+{ "wi
Re: [FFmpeg-devel] [PATCH 1/2] ffplay: Seek only when pressing the right mouse button on the video window.
On Fri, 15 Jan 2016, Vittorio Gambaletta (VittGam) wrote: Seeking by clicking on the video window can be annoying, because the user might click on it accidentally while eg. trying to get focus on it, and ffplay seeks instead. This commit changes that behaviour to seek only when the right mouse button is used to click and drag on the window. Signed-off-by: Vittorio Gambaletta Looks good, I will apply in a few days if no one steps up against it. However your patch seem to have some whitespace errors probably caused by your mailer, and because of that it does not apply cleanly so you might consider sending patches next time as an attachment if your mailer messes something up inline. Regards, Marton ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 2/2] ffplay: Toggle full screen when double-clicking the video window with the left mouse button.
On Fri, 15 Jan 2016, Vittorio Gambaletta (VittGam) wrote:
Now that the seek only happens with the right mouse button, it makes
sense to toggle full screen when double-clicking with the left mouse
button, like other video players do.
I am not against this.
Signed-off-by: Vittorio Gambaletta
---
ffplay.c | 18 ++
1 file changed, 18 insertions(+)
diff --git a/ffplay.c b/ffplay.c
index 2fa7165..582ca39 100644
--- a/ffplay.c
+++ b/ffplay.c
@@ -3473,6 +3473,24 @@ static void event_loop(VideoState *cur_stream)
do_exit(cur_stream);
break;
}
+{
Whitespace and indentation seems messed up in this patch as well.
+static int mouse_left_click_status = 0;
+static double mouse_left_click_last_x, mouse_left_click_last_y;
+if (event.button.button == SDL_BUTTON_LEFT) {
+if (mouse_left_click_status == 1 && av_gettime_relative() -
cursor_last_shown <= 50
+&& fabs(event.button.x - mouse_left_click_last_x) <= 1 &&
fabs(event.button.y - mouse_left_click_last_y) <= 1) {
+toggle_full_screen(cur_stream);
+cur_stream->force_refresh = 1;
+mouse_left_click_status = 0;
+} else {
+mouse_left_click_status = 1;
+mouse_left_click_last_x = event.button.x;
+mouse_left_click_last_y = event.button.y;
+}
+} else {
+mouse_left_click_status = 0;
+}
+}
I don't think we need the complex stuff about not clicking too far out of
the first click, what are the chances of a fast moving mouse with double
clicking? Slim to none I guess. Are you aware of a case when this matters?
I think it is enough if you just store the timestamp of the last left
click in an int64_t and check the time difference next time the user do a
left click, this way you can detect double clicking with a single variable
and you don't have to use cursor_last_shown which is updated on every
motion as well.
Regards,
Marton
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH] lavf/img2dec: change the default pattern to none
The current default pattern type, glob_sequence is deprecated for almost 3.5
years, before removing it, we need a new, sensible pattern default. I suggest
"none" which turns off pattern matching, because using the filename as a
pattern can be a security risk exposing other files in the source directory,
and can also cause problems for file names which are patterns by themselves.
Signed-off-by: Marton Balint
---
doc/demuxers.texi | 2 +-
libavformat/img2dec.c | 2 +-
libavformat/seek-test.c | 1 +
tests/fate-run.sh | 2 +-
tests/fate/filter-video.mak | 196 ++--
tests/fate/utvideo.mak | 2 +-
tests/fate/video.mak| 4 +-
tests/ffserver.conf | 2 +-
tests/lavf-regression.sh| 18 ++--
9 files changed, 115 insertions(+), 114 deletions(-)
diff --git a/doc/demuxers.texi b/doc/demuxers.texi
index fb1e4fb..d544436 100644
--- a/doc/demuxers.texi
+++ b/doc/demuxers.texi
@@ -377,7 +377,7 @@ This pattern type is deprecated in favor of @var{glob} and
@var{sequence}.
@end table
-Default value is @var{glob_sequence}.
+Default value is @var{none}.
@item pixel_format
Set the pixel format of the images to read. If not specified the pixel
format is guessed from the first image file in the sequence.
diff --git a/libavformat/img2dec.c b/libavformat/img2dec.c
index db4b4b7..258f741 100644
--- a/libavformat/img2dec.c
+++ b/libavformat/img2dec.c
@@ -557,7 +557,7 @@ const AVOption ff_img_options[] = {
{ "framerate","set the video framerate",
OFFSET(framerate),AV_OPT_TYPE_VIDEO_RATE, {.str = "25"}, 0, 0, DEC },
{ "loop", "force loop over input file sequence", OFFSET(loop),
AV_OPT_TYPE_BOOL, {.i64 = 0 }, 0, 1, DEC },
-{ "pattern_type", "set pattern type",
OFFSET(pattern_type), AV_OPT_TYPE_INT,{.i64=PT_GLOB_SEQUENCE}, 0,
INT_MAX, DEC, "pattern_type"},
+{ "pattern_type", "set pattern type",
OFFSET(pattern_type), AV_OPT_TYPE_INT,{.i64=PT_NONE}, 0, INT_MAX,
DEC, "pattern_type"},
{ "glob_sequence","select glob/sequence pattern type", 0,
AV_OPT_TYPE_CONST, {.i64=PT_GLOB_SEQUENCE}, INT_MIN, INT_MAX, DEC,
"pattern_type" },
{ "glob", "select glob pattern type",0,
AV_OPT_TYPE_CONST, {.i64=PT_GLOB }, INT_MIN, INT_MAX, DEC,
"pattern_type" },
{ "sequence", "select sequence pattern type",0,
AV_OPT_TYPE_CONST, {.i64=PT_SEQUENCE }, INT_MIN, INT_MAX, DEC,
"pattern_type" },
diff --git a/libavformat/seek-test.c b/libavformat/seek-test.c
index bfd06db..f9de804 100644
--- a/libavformat/seek-test.c
+++ b/libavformat/seek-test.c
@@ -87,6 +87,7 @@ int main(int argc, char **argv)
av_dict_set(&format_opts, "channels", "1", 0);
av_dict_set(&format_opts, "sample_rate", "22050", 0);
+av_dict_set(&format_opts, "pattern_type", "sequence", 0);
/* initialize libavcodec, and register all codecs and formats */
av_register_all();
diff --git a/tests/fate-run.sh b/tests/fate-run.sh
index 16087cb..a353440 100755
--- a/tests/fate-run.sh
+++ b/tests/fate-run.sh
@@ -191,7 +191,7 @@ video_filter(){
label=${test#filter-}
raw_src="${target_path}/tests/vsynth1/%02d.pgm"
printf '%-20s' $label
-ffmpeg $DEC_OPTS -f image2 -vcodec pgmyuv -i $raw_src \
+ffmpeg $DEC_OPTS -f image2 -vcodec pgmyuv -pattern_type sequence -i
$raw_src \
$FLAGS $ENC_OPTS -vf "$filters" -vcodec rawvideo -frames:v 5 $* -f nut
md5:
}
diff --git a/tests/fate/filter-video.mak b/tests/fate/filter-video.mak
index 4186996..ae0d228 100644
--- a/tests/fate/filter-video.mak
+++ b/tests/fate/filter-video.mak
@@ -81,23 +81,23 @@ fate-filter-lavd-scalenorm: CMD = framecrc -f lavfi
-graph_file $(TARGET_PATH)/t
FATE_FILTER_VSYNTH-$(CONFIG_BOXBLUR_FILTER) += fate-filter-boxblur
-fate-filter-boxblur: CMD = framecrc -c:v pgmyuv -i $(SRC) -vf boxblur=2:1
+fate-filter-boxblur: CMD = framecrc -c:v pgmyuv -pattern_type sequence -i
$(SRC) -vf boxblur=2:1
FATE_FILTER_VSYNTH-$(call ALLYES, COLORCHANNELMIXER_FILTER FORMAT_FILTER
PERMS_FILTER) += fate-filter-colorchannelmixer
-fate-filter-colorchannelmixer: CMD = framecrc -c:v pgmyuv -i $(SRC) -vf
format=rgb24,perms=random,colorchannelmixer=.31415927:.4:.31415927:0:.27182818:.8:.27182818:0:.2:.6:.2:0
-flags +bitexact -sws_flags +accurate_rnd+bitexact
+fate-filter-colorchannelmixer: CMD = framecrc -c:v pgmyuv -pattern_type
sequence -i $(SRC) -vf
format=rgb24,perms=random,colorchannelmixer=.31415927:.4:.31415927:0:.27182818:.8:.27182818:0:.2:.6:.2:0
-flags +bitexact -sws_flags +accurate_rnd+bitexact
FATE_FILTER_VSYNTH-$(CONFIG_DRAWBOX_FILTER) += fate-filter-drawbox
-fate-filter-drawbox: CMD = framecrc -c:v pgmyuv -i $(SRC) -vf
drawbox=224:24:88:72:red@0.5
+fate-filter-drawbox: CMD = framecrc -c:v pgmyuv -pattern_type sequence -i
$(SRC) -vf drawbox=224:24:88:72:red@0.5
FATE_FILTER_VS
Re: [FFmpeg-devel] [PATCH 1/2] ffplay: Seek only when pressing the right mouse button on the video window.
Hi, On 16/01/2016 21:13:14 CET, Marton Balint wrote: On Fri, 15 Jan 2016, Vittorio Gambaletta (VittGam) wrote: Seeking by clicking on the video window can be annoying, because the user might click on it accidentally while eg. trying to get focus on it, and ffplay seeks instead. This commit changes that behaviour to seek only when the right mouse button is used to click and drag on the window. Signed-off-by: Vittorio Gambaletta Looks good, I will apply in a few days if no one steps up against it. However your patch seem to have some whitespace errors probably caused by your mailer, and because of that it does not apply cleanly so you might consider sending patches next time as an attachment if your mailer messes something up inline. Thank you for your suggestion, I thought I finally fixed that problem with my mailer while in fact I failed again... Cheers, Vittorio ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 2/2] ffplay: Toggle full screen when double-clicking the video window with the left mouse button.
Hi,
On 16/01/2016 21:25:00 CET, Marton Balint wrote:
On Fri, 15 Jan 2016, Vittorio Gambaletta (VittGam) wrote:
Now that the seek only happens with the right mouse button, it makes
sense to toggle full screen when double-clicking with the left mouse
button, like other video players do.
I am not against this.
Signed-off-by: Vittorio Gambaletta
---
ffplay.c | 18 ++
1 file changed, 18 insertions(+)
diff --git a/ffplay.c b/ffplay.c
index 2fa7165..582ca39 100644
--- a/ffplay.c
+++ b/ffplay.c
@@ -3473,6 +3473,24 @@ static void event_loop(VideoState *cur_stream)
do_exit(cur_stream);
break;
}
+{
Whitespace and indentation seems messed up in this patch as well.
+static int mouse_left_click_status = 0;
+static double mouse_left_click_last_x, mouse_left_click_last_y;
+if (event.button.button == SDL_BUTTON_LEFT) {
+if (mouse_left_click_status == 1 && av_gettime_relative() -
cursor_last_shown <= 50
+&& fabs(event.button.x - mouse_left_click_last_x) <= 1 &&
fabs(event.button.y - mouse_left_click_last_y) <= 1) {
+toggle_full_screen(cur_stream);
+cur_stream->force_refresh = 1;
+mouse_left_click_status = 0;
+} else {
+mouse_left_click_status = 1;
+mouse_left_click_last_x = event.button.x;
+mouse_left_click_last_y = event.button.y;
+}
+} else {
+mouse_left_click_status = 0;
+}
+}
I don't think we need the complex stuff about not clicking too far out of the
first click, what are the chances of a fast moving mouse with double clicking?
Slim to none I guess. Are you aware of a case when this matters?
I don't know, I just looked at how the click counter is implemented in SDL2 and
emulated that behaviour with SDL1. But it can be overkill, yes; in fact VLC
does not seem to check this.
By the way, VLC doesn't even clear the status/timestamp when clicking another
button, so maybe that is overkill too.
I think it is enough if you just store the timestamp of the last left click in
an int64_t and check the time difference next time the user do a left click,
this way you can detect double clicking with a single variable and you don't
have to use cursor_last_shown which is updated on every motion as well.
Right, also I didn't think twice about the fact that cursor_last_shown is
updated on motion too.
Thanks for the suggestions, I will send an updated patch soon.
Cheers,
Vittorio
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] lavf/img2dec: change the default pattern to none
On Sat, 16 Jan 2016 21:42:48 +0100 Marton Balint wrote: > The current default pattern type, glob_sequence is deprecated for almost 3.5 > years, before removing it, we need a new, sensible pattern default. I suggest > "none" which turns off pattern matching, because using the filename as a > pattern can be a security risk exposing other files in the source directory, > and can also cause problems for file names which are patterns by themselves. > > Signed-off-by: Marton Balint I'm very much in favor of this, but doesn't it break compatibility big time? Also, if we break this, maybe this should be reimplemented as protocol prefix? So you would use e.g. "ffmpeg -i pattern:%02d.jpg". ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH v2 1/2] ffplay: Seek only when pressing the right mouse button on the video window.
Seeking by clicking on the video window can be annoying, because
the user might click on it accidentally while eg. trying to get
focus on it, and ffplay seeks instead.
This commit changes that behaviour to seek only when the right
mouse button is used to click and drag on the window.
Signed-off-by: Vittorio Gambaletta
---
ffplay.c |4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/ffplay.c b/ffplay.c
index d2e3dc6..2fa7165 100644
--- a/ffplay.c
+++ b/ffplay.c
@@ -3480,9 +3480,11 @@ static void event_loop(VideoState *cur_stream)
}
cursor_last_shown = av_gettime_relative();
if (event.type == SDL_MOUSEBUTTONDOWN) {
+if (event.button.button != SDL_BUTTON_RIGHT)
+break;
x = event.button.x;
} else {
-if (event.motion.state != SDL_PRESSED)
+if (!(event.motion.state & SDL_BUTTON_RMASK))
break;
x = event.motion.x;
}
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH v2 2/2] ffplay: Toggle full screen when double-clicking the video window with the left mouse button.
Now that the seek only happens with the right mouse button, it makes
sense to toggle full screen when double-clicking with the left mouse
button, like other video players do.
Signed-off-by: Vittorio Gambaletta
---
ffplay.c | 10 ++
1 file changed, 10 insertions(+)
diff --git a/ffplay.c b/ffplay.c
index 2fa7165..5b473e9 100644
--- a/ffplay.c
+++ b/ffplay.c
@@ -3473,6 +3473,16 @@ static void event_loop(VideoState *cur_stream)
do_exit(cur_stream);
break;
}
+if (event.button.button == SDL_BUTTON_LEFT) {
+static int64_t last_mouse_left_click = 0;
+if (av_gettime_relative() - last_mouse_left_click <= 50) {
+toggle_full_screen(cur_stream);
+cur_stream->force_refresh = 1;
+last_mouse_left_click = 0;
+} else {
+last_mouse_left_click = av_gettime_relative();
+}
+}
case SDL_MOUSEMOTION:
if (cursor_hidden) {
SDL_ShowCursor(1);
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH v2 1/2] ffplay: Seek only when pressing the right mouse button on the video window.
Hi,
It seems to me that whitespace was right on my end in the other
patches...
This time I've literally piped the mails to my mailserver using
netcat, like I always do with OpenWrt patches anyway, so please
check if I'm still getting something wrong with whitespace now.
Thank you,
Vittorio
On 16/01/2016 22:41:30 CET, Vittorio Gambaletta (VittGam) wrote:
Seeking by clicking on the video window can be annoying, because
the user might click on it accidentally while eg. trying to get
focus on it, and ffplay seeks instead.
This commit changes that behaviour to seek only when the right
mouse button is used to click and drag on the window.
Signed-off-by: Vittorio Gambaletta
---
ffplay.c |4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/ffplay.c b/ffplay.c
index d2e3dc6..2fa7165 100644
--- a/ffplay.c
+++ b/ffplay.c
@@ -3480,9 +3480,11 @@ static void event_loop(VideoState *cur_stream)
}
cursor_last_shown = av_gettime_relative();
if (event.type == SDL_MOUSEBUTTONDOWN) {
+if (event.button.button != SDL_BUTTON_RIGHT)
+break;
x = event.button.x;
} else {
-if (event.motion.state != SDL_PRESSED)
+if (!(event.motion.state & SDL_BUTTON_RMASK))
break;
x = event.motion.x;
}
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 1/2] x86/fixed_dsp: add ff_butterflies_fixed_sse2
On 1/16/16, James Almer wrote: > Signed-off-by: James Almer > --- > libavutil/fixed_dsp.c | 3 +++ > libavutil/fixed_dsp.h | 2 ++ > libavutil/x86/Makefile | 2 ++ > libavutil/x86/fixed_dsp.asm| 48 > ++ > libavutil/x86/fixed_dsp_init.c | 35 ++ > 5 files changed, 90 insertions(+) > create mode 100644 libavutil/x86/fixed_dsp.asm > create mode 100644 libavutil/x86/fixed_dsp_init.c > looks ok, assuming its bitexact with c ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 1/2] x86/fixed_dsp: add ff_butterflies_fixed_sse2
On 1/16/2016 8:55 PM, Paul B Mahol wrote: > On 1/16/16, James Almer wrote: >> Signed-off-by: James Almer >> --- >> libavutil/fixed_dsp.c | 3 +++ >> libavutil/fixed_dsp.h | 2 ++ >> libavutil/x86/Makefile | 2 ++ >> libavutil/x86/fixed_dsp.asm| 48 >> ++ >> libavutil/x86/fixed_dsp_init.c | 35 ++ >> 5 files changed, 90 insertions(+) >> create mode 100644 libavutil/x86/fixed_dsp.asm >> create mode 100644 libavutil/x86/fixed_dsp_init.c >> > > looks ok, assuming its bitexact with c It is. You can check it with the second patch (that i'm not going to commit since i need to make it a checkasm test instead). Pushed, thanks. ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH] libavutil: add mastering display metadata sidedata
Adding mastering display metadata struct to avutil. The mastering display
metadata contains information
about the mastering display color volume (SMPTE 2086:2014).
This info comes from HEVC in the SEI_TYPE_MASTERING_DISPLAY_INFO and is soon to
be included in MKV:
https://mailarchive.ietf.org/arch/search/?email_list=cellar&gbt=1&index=sZyfPTM-QY69P-0omfOIiTN622o
so it is similar to SEI FPA / stereo_mode in MKV and as such this patch follows
how AVStereo3D is implemented.
I'll add support to HEVC in a follow-up (and MKV when spec is approved).
Signed-off-by: Neil Birkbeck
---
libavutil/Makefile | 2 +
libavutil/frame.h | 7 ++-
libavutil/mastering_display_metadata.c | 43 +
libavutil/mastering_display_metadata.h | 87 ++
libavutil/version.h| 2 +-
5 files changed, 139 insertions(+), 2 deletions(-)
create mode 100644 libavutil/mastering_display_metadata.c
create mode 100644 libavutil/mastering_display_metadata.h
diff --git a/libavutil/Makefile b/libavutil/Makefile
index bf8c713..65b2d25 100644
--- a/libavutil/Makefile
+++ b/libavutil/Makefile
@@ -38,6 +38,7 @@ HEADERS = adler32.h
\
log.h \
macros.h \
mathematics.h \
+ mastering_display_metadata.h \
md5.h \
mem.h \
motion_vector.h \
@@ -115,6 +116,7 @@ OBJS = adler32.o
\
log.o\
log2_tab.o \
mathematics.o\
+ mastering_display_metadata.o \
md5.o\
mem.o\
murmur3.o\
diff --git a/libavutil/frame.h b/libavutil/frame.h
index 9c6061a..308355b 100644
--- a/libavutil/frame.h
+++ b/libavutil/frame.h
@@ -106,12 +106,17 @@ enum AVFrameSideDataType {
* @endcode
*/
AV_FRAME_DATA_SKIP_SAMPLES,
-
/**
* This side data must be associated with an audio frame and corresponds to
* enum AVAudioServiceType defined in avcodec.h.
*/
AV_FRAME_DATA_AUDIO_SERVICE_TYPE,
+/**
+ * Mastering display metadata associated with a video frame. The payload is
+ * an AVMasteringDisplayMetadata type and contains information about the
+ * mastering display color volume.
+ */
+AV_FRAME_DATA_MASTERING_DISPLAY_METADATA
};
enum AVActiveFormatDescription {
diff --git a/libavutil/mastering_display_metadata.c
b/libavutil/mastering_display_metadata.c
new file mode 100644
index 000..f7114f6
--- /dev/null
+++ b/libavutil/mastering_display_metadata.c
@@ -0,0 +1,43 @@
+/**
+ * Copyright (c) 2016 Neil Birkbeck
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg 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, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include
+#include
+
+#include "mastering_display_metadata.h"
+#include "mem.h"
+
+AVMasteringDisplayMetadata *av_mastering_display_metadata_alloc(void)
+{
+return av_mallocz(sizeof(AVMasteringDisplayMetadata));
+}
+
+AVMasteringDisplayMetadata
*av_mastering_display_metadata_create_side_data(AVFrame *frame)
+{
+AVFrameSideData *side_data = av_frame_new_side_data(frame,
+
AV_FRAME_DATA_MASTERING_DISPLAY_METADATA,
+
sizeof(AVMasteringDisplayMetadata));
+if (!side_data)
+return NULL;
+
+memset(side_data->data, 0, sizeof(AVMasteringDisplayMetadata));
+
+return (AVMasteringDisplayMetadata *)side_data->data;
+}
diff --git a/libavutil/mastering_display_metadata.
Re: [FFmpeg-devel] [PATCH 1/4] fate: add 10-bit v210 encoder tests
On 2016-01-15 20:07, James Darnley wrote: > ... If nobody has further comments about the patches I will probably push these after I wake up. signature.asc Description: OpenPGP digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH] remove the deprecated avpicture_get_size() function
---
libavcodec/libutvideodec.cpp | 3 ++-
libavcodec/libutvideoenc.cpp | 3 ++-
libavdevice/decklink_dec.cpp | 2 +-
libavdevice/lavfi.c | 3 ++-
libavformat/frmdec.c | 3 ++-
5 files changed, 9 insertions(+), 5 deletions(-)
diff --git a/libavcodec/libutvideodec.cpp b/libavcodec/libutvideodec.cpp
index 47261a6..7c60f4f 100644
--- a/libavcodec/libutvideodec.cpp
+++ b/libavcodec/libutvideodec.cpp
@@ -27,6 +27,7 @@
extern "C" {
#include "avcodec.h"
+#include "libavutil/imgutils.h"
}
#include "libutvideo.h"
@@ -93,7 +94,7 @@ static av_cold int utvideo_decode_init(AVCodecContext *avctx)
}
/* Only allocate the buffer once */
-utv->buf_size = avpicture_get_size(avctx->pix_fmt, avctx->width,
avctx->height);
+utv->buf_size = av_image_get_buffer_size(avctx->pix_fmt, avctx->width,
avctx->height, 1);
#ifdef UTVF_UQY2
if (format == UTVF_v210)
utv->buf_size += avctx->height * ((avctx->width + 47) / 48) * 128; //
the linesize used by the decoder, this does not seem to be exported
diff --git a/libavcodec/libutvideoenc.cpp b/libavcodec/libutvideoenc.cpp
index 8746247..96d8fe1 100644
--- a/libavcodec/libutvideoenc.cpp
+++ b/libavcodec/libutvideoenc.cpp
@@ -27,6 +27,7 @@
extern "C" {
#include "libavutil/avassert.h"
+#include "libavutil/imgutils.h"
#include "avcodec.h"
#include "internal.h"
}
@@ -94,7 +95,7 @@ static av_cold int utvideo_encode_init(AVCodecContext *avctx)
* We use this buffer to hold the data that Ut Video returns,
* since we cannot decode planes separately with it.
*/
-ret = avpicture_get_size(avctx->pix_fmt, avctx->width, avctx->height);
+ret = av_image_get_buffer_size(avctx->pix_fmt, avctx->width,
avctx->height, 1);
if (ret < 0) {
av_free(info);
return ret;
diff --git a/libavdevice/decklink_dec.cpp b/libavdevice/decklink_dec.cpp
index 89f93de..6c5bc5d 100644
--- a/libavdevice/decklink_dec.cpp
+++ b/libavdevice/decklink_dec.cpp
@@ -473,7 +473,7 @@ av_cold int ff_decklink_read_header(AVFormatContext *avctx)
st->codec->time_base.den = ctx->bmd_tb_den;
st->codec->time_base.num = ctx->bmd_tb_num;
-st->codec->bit_rate= avpicture_get_size(st->codec->pix_fmt,
ctx->bmd_width, ctx->bmd_height) * 1/av_q2d(st->codec->time_base) * 8;
+st->codec->bit_rate= av_image_get_buffer_size(st->codec->pix_fmt,
ctx->bmd_width, ctx->bmd_height, 1) * 1/av_q2d(st->codec->time_base) * 8;
if (cctx->v210) {
st->codec->codec_id= AV_CODEC_ID_V210;
diff --git a/libavdevice/lavfi.c b/libavdevice/lavfi.c
index 3453b4d..077879e 100644
--- a/libavdevice/lavfi.c
+++ b/libavdevice/lavfi.c
@@ -30,6 +30,7 @@
#include "libavutil/bprint.h"
#include "libavutil/channel_layout.h"
#include "libavutil/file.h"
+#include "libavutil/imgutils.h"
#include "libavutil/internal.h"
#include "libavutil/log.h"
#include "libavutil/mem.h"
@@ -430,7 +431,7 @@ static int lavfi_read_packet(AVFormatContext *avctx,
AVPacket *pkt)
stream_idx = lavfi->sink_stream_map[min_pts_sink_idx];
if (frame->width /* FIXME best way of testing a video */) {
-size = avpicture_get_size(frame->format, frame->width, frame->height);
+size = av_image_get_buffer_size(frame->format, frame->width,
frame->height, 1);
if ((ret = av_new_packet(pkt, size)) < 0)
return ret;
diff --git a/libavformat/frmdec.c b/libavformat/frmdec.c
index a6f19af..260afbc 100644
--- a/libavformat/frmdec.c
+++ b/libavformat/frmdec.c
@@ -25,6 +25,7 @@
*/
#include "libavcodec/raw.h"
+#include "libavutil/imgutils.h"
#include "libavutil/intreadwrite.h"
#include "avformat.h"
@@ -80,7 +81,7 @@ static int frm_read_packet(AVFormatContext *avctx, AVPacket
*pkt)
if (s->count)
return AVERROR_EOF;
-packet_size = avpicture_get_size(stc->pix_fmt, stc->width, stc->height);
+packet_size = av_image_get_buffer_size(stc->pix_fmt, stc->width,
stc->height, 1);
if (packet_size < 0)
return AVERROR_INVALIDDATA;
--
2.5.0
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
