Re: [FFmpeg-devel] [PATCH] doc/filters: Add ascii graphics to clarify what the currently implemented tinterlace modes do
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 02/12/14 17:13, Michael Niedermayer wrote: On Tue, Dec 02, 2014 at 05:02:49PM +, tim nicholson wrote: On 02/12/14 15:33, Michael Niedermayer wrote: Signed-off-by: Michael Niedermayer michae...@gmx.at --- doc/filters.texi | 131 ++ 1 file changed, 131 insertions(+) diff --git a/doc/filters.texi b/doc/filters.texi index 8c16c7a..0ea3955 100644 --- a/doc/filters.texi +++ b/doc/filters.texi @@ -8537,33 +8537,164 @@ Available values are: @item merge, 0 Move odd frames into the upper field, even into the lower field, generating a double height frame at half frame rate. +@example + -- time +Input: +Frame 1 Frame 2 Frame 3 Frame 4 + +1 2 3 4 +1 2 3 4 +1 2 3 4 +1 2 3 4 + +Output: +1 3 +2 4 +1 3 +2 4 +1 3 +2 4 +1 3 +2 4 [..] I find the ascii graphic style of the interlace filter more understandable how do you suggest above (and the other modes) would be documented in that style ? Sorry for the delayed reply ISP issue... I think the main difference is going left to right, not top to bottom for input to output, this then matches not only interlace but the general filtergraph descriptions. I will try and see if I can give an example if you like. [...] - -- Tim. Key Fingerprint 38CF DB09 3ED0 F607 8B67 6CED 0C0B FC44 8B0B FC83 -BEGIN PGP SIGNATURE- Version: GnuPG v2 iQEcBAEBAgAGBQJUgBpMAAoJEAwL/ESLC/yDxkEH/1hWM5A1P3V3cz2XH3pF01/0 64e4A9RRx0BpX0pDHqik9aVNQtL+qSBFYkkQGPn8smTdqWEQXfIe/lWQSXODj95T +/Un1V33cxnPznJnTCfxrZEKuYbEQYehyuasGcgzq/L0CIYamGJmuWZ4feZKQ0Vk ZdCAs/ygX9igc7V+41UG5dQu6n+nhIwAkOAJBiGkeVMKu76YH19fF/G4TO7tYczn GbVgKfu78LK8cZZOWl3w42EO7aF98TeG6q29jQsiDABLNa8JocIVmxaijbaIEJ3R 2Eb5yzCeyZadgOrenRzF9iR0W4JvlXK2Expqr15SISRjVJ4L8AjZqs6nPOsraHY= =KKl4 -END PGP SIGNATURE- ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] avfilter/tinterlace: merge code with interlace
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 03/12/14 13:08, Michael Niedermayer wrote: On Tue, Dec 02, 2014 at 07:43:47PM +0100, Clément Bœsch wrote: --- After this commit, interlace doesn't behave the same if the input frames are already interlaced. Does anyone wants me to keep this behaviour? probably best if its kept but can be forced enabled/disabled with an option. these interlacing flags on the input can be wrong +1 otherwise LGTM [...] - -- Tim. Key Fingerprint 38CF DB09 3ED0 F607 8B67 6CED 0C0B FC44 8B0B FC83 -BEGIN PGP SIGNATURE- Version: GnuPG v2 iQEcBAEBAgAGBQJUgBvNAAoJEAwL/ESLC/yDOmMH/3U73HxpvOsW4QwvfBDQFeD0 dMqyo9NgZFUzsNLptORc9ifRXQrDpxvhkmhPtqpCgpgLBV+E88VyXCEo+QPrq8AY l617h6Wb/FbRHT4opYTfU2DsSgdNKalAZacrckY1dGFNEACQRg04n8jGkbyK5gu/ 1YXFknMic92Vq29uwPrymrmvcGyWtjj2goTL/6ikVbAdFGQ6dks21PR412YUqk6z BcVuWXxXv9HZft2pfE1edcT3FEqiJL/iKZmzq91i6vzkyS5xeHXkLgCNc+P3NfOx pPvKl4lFhRFMh4OTxVT82EqUvrFovgH3bRXXocQeIxuzf+BXzrkVZOSMPjtoG1o= =bJWr -END PGP SIGNATURE- ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] avfilter/tinterlace: merge code with interlace
On 02/12/14 18:43, Clément Bœsch wrote: --- After this commit, interlace doesn't behave the same if the input frames are already interlaced. Does anyone wants me to keep this behaviour? Also, the PTS adjustments might be different, but I didn't follow the current discussions about it so I can't tell what people want. --- libavfilter/Makefile| 2 +- libavfilter/interlace.h | 58 - libavfilter/tinterlace.h| 5 +- libavfilter/vf_interlace.c | 250 libavfilter/vf_tinterlace.c | 44 +++ libavfilter/x86/Makefile| 2 +- libavfilter/x86/vf_interlace_init.c | 47 --- 7 files changed, 49 insertions(+), 359 deletions(-) delete mode 100644 libavfilter/interlace.h delete mode 100644 libavfilter/vf_interlace.c delete mode 100644 libavfilter/x86/vf_interlace_init.c [..] The docs will need updating to remove the vf_interlace entry and add the lp filter options. Maybe that is the time to merge the ascii art graphics and add Michael's additional mode ones as well? -- Tim. Key Fingerprint 38CF DB09 3ED0 F607 8B67 6CED 0C0B FC44 8B0B FC83 ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] avfilter/tinterlace: merge code with interlace
On Thu, Dec 04, 2014 at 09:31:47AM +, tim nicholson wrote: On 02/12/14 18:43, Clément Bœsch wrote: --- After this commit, interlace doesn't behave the same if the input frames are already interlaced. Does anyone wants me to keep this behaviour? Also, the PTS adjustments might be different, but I didn't follow the current discussions about it so I can't tell what people want. --- libavfilter/Makefile| 2 +- libavfilter/interlace.h | 58 - libavfilter/tinterlace.h| 5 +- libavfilter/vf_interlace.c | 250 libavfilter/vf_tinterlace.c | 44 +++ libavfilter/x86/Makefile| 2 +- libavfilter/x86/vf_interlace_init.c | 47 --- 7 files changed, 49 insertions(+), 359 deletions(-) delete mode 100644 libavfilter/interlace.h delete mode 100644 libavfilter/vf_interlace.c delete mode 100644 libavfilter/x86/vf_interlace_init.c [..] The docs will need updating to remove the vf_interlace entry and add the lp filter options. Maybe that is the time to merge the ascii art graphics and add Michael's additional mode ones as well? None of the filter disappear and their options don't change. It's just code refactoring. -- Clément B. pgpqNCweOylFp.pgp Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 2/2] RELEASE: Bump
On Wed, Dec 03, 2014 at 08:15:20PM -0800, Timothy Gu wrote: Signed-off-by: Timothy Gu timothyg...@gmail.com --- RELEASE | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) applied thanks [...] -- 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
[FFmpeg-devel] Fix bug for POWER LE: libavcodec/ppc/vp8dsp_altivec.c
Hi, We propose this patch to fix vp8 bugs for POWER8 little endian. Functions put_vp8_epel_h_altivec_core(), puta_vp8_epel_v_altivec_core(), and put_vp8_pixels16_altivec() are fixed. The passed test cases change from 2296/2328 to 2328/2328. Thanks for your review. Rong Yan -- The world has enough for everyone's need, but not enough for everyone's greed. 0001-avcodec-ppc-vp8dsp_altivec.c-POWER-LE-support-put_vp.patch Description: Binary data ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] Fix bug for POWER LE: avcodec/ppc/vc1dsp_altivec.c
On Thu, Dec 04, 2014 at 02:06:34PM +0800, rongyan wrote: Hi, We propose this patch to fix vc1 bugs for POWER8 little endian. Function vc1_inv_trans_8x4_altivec() is fixed. The passed test cases change from 2296/2328 to 2303/2328. Thanks for your review. Rong Yan -- The world has enough for everyone's need, but not enough for everyone's greed. vc1dsp_altivec.c | 13 ++--- 1 file changed, 10 insertions(+), 3 deletions(-) ea32c44c9e3b4d879c2e4bd854c42755fe0b131b 0001-avcodec-ppc-vc1dsp_altivec-POWER-LE-support-vc1_inv_.patch From 97126a766f2825bfec198dc9ffe0d98d453f6689 Mon Sep 17 00:00:00 2001 From: Rong Yan rongyan...@gmail.com Date: Thu, 4 Dec 2014 05:51:30 + Subject: [PATCH] avcodec/ppc/vc1dsp_altivec: POWER LE support vc1_inv_trans_8x4_altivec() add marcos GET_TMP2() applied thanks [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB Old school: Use the lowest level language in which you can solve the problem conveniently. New school: Use the highest level language in which the latest supercomputer can solve the problem without the user falling asleep waiting. 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] Adding closed caption decoder
Le quartidi 14 frimaire, an CCXXIII, Anshul a écrit : It would be valid, c708 is backward compatible as far as I have seen while working with ccextractor. C708 is made in such a way that people dont have to throw there older TV sets or device. if any decoder is not able to decode c708 part, even then they can get correct subtitle. So if I understand you correctly, the format could have kept the same name, just as when Matroska adds a new codec name, and the name is only relevant as an indication of what a decoder is capable of handling. Yes while doing that only I realized, if we use i then cc must be in first stream. When I had multiple stream(with audio), then I was not able to map subcc to my video stream. using stream_idx , we are saying that if we can use out0+subcc and out1+subcc. i logic fails when there is audio in 1st stream, there was no way i could get the subtitle if my first stream is not the one where closed caption are present. I will look more closely into it. In the meantime, you may want to have a look at this one: http://samples.ffmpeg.org/MPEG2/subcc/ClosedCaptions/subccItalic.mpg I see quite a few subcc packets, but the decoder does not output anything. Furthermore, I do not see any ASS styling in the decoder, that would be a good test case. More comments to come. Regards, -- Nicolas George 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] Adding closed caption decoder
Furthermore, I do not see any ASS styling in the decoder, that would be a good test case. Captions shouldn't be ASS styled - they should be styled as per the specification (see VLC). Kieran ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 2/2] Adding closed caption decoder
On Thu, Dec 04, 2014 at 11:32:40AM +, Kieran Kunhya wrote: Furthermore, I do not see any ASS styling in the decoder, that would be a good test case. Captions shouldn't be ASS styled - they should be styled as per the specification (see VLC). I'd be surprised if you can't get close to the intended rendering with ASS markup. What style do you think wouldn't be supported? -- Clément B. pgpeFkQLdx28C.pgp Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 2/2] Adding closed caption decoder
Le quartidi 14 frimaire, an CCXXIII, Kieran Kunhya a écrit :
Captions shouldn't be ASS styled - they should be styled as per the
specification (see VLC).
You are certainly right, but we are talking about a decoder, and FFmpeg's
internal format for decoded text subtitles is still ASS. If something in the
specification can say italic (seeing the file name, I guess it is
possible), then the decoder must output proper {\i1} and {\i0} tags, and
same for bold, colors, whatever exists in the spec.
Or am I missing something?
Regards,
--
Nicolas George
signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] lavfi: add tdiff filter
On date Sunday 2014-11-23 14:57:41 +0100, Clément Bœsch encoded:
On Sun, Nov 23, 2014 at 02:50:51PM +0100, Stefano Sabatini wrote:
On date Sunday 2014-11-23 02:01:16 +0100, Clément Bœsch encoded:
On Sat, Nov 22, 2014 at 11:44:29PM +0100, Stefano Sabatini wrote:
TODO: bump minor
---
Changelog| 1 +
doc/filters.texi | 4 ++
libavfilter/Makefile | 1 +
libavfilter/allfilters.c | 1 +
libavfilter/vf_tdiff.c | 131
+++
5 files changed, 138 insertions(+)
create mode 100644 libavfilter/vf_tdiff.c
What about adding a tblend in vf_blend.c instead?
What do you exactly mean? tdiff works with a single input, blend with
two inputs.
That's the point: make another filter with temporal dimension in blend.
Blend filter has many features such as doing something more than diff
and having threading. The main function takes 2 frames so just add the
glue code around to take the previous and the current frame instead.
That's just a suggestion, maybe the other functions don't make much sense.
Feel free to ignore.
Uh, what about this?
--
FFmpeg = Frightening Faithless Martial Powerful Evil Goblin
From c805460ceb04ec2da3f607e69067f4cc710c0613 Mon Sep 17 00:00:00 2001
From: Stefano Sabatini stefa...@gmail.com
Date: Thu, 4 Dec 2014 12:27:53 +0100
Subject: [PATCH] lavfi: add tblend filter
---
libavfilter/Makefile | 1 +
libavfilter/allfilters.c | 1 +
libavfilter/vf_blend.c | 73 +++-
3 files changed, 74 insertions(+), 1 deletion(-)
diff --git a/libavfilter/Makefile b/libavfilter/Makefile
index d41a52e..d1be7e3 100644
--- a/libavfilter/Makefile
+++ b/libavfilter/Makefile
@@ -187,6 +187,7 @@ OBJS-$(CONFIG_STEREO3D_FILTER) += vf_stereo3d.o
OBJS-$(CONFIG_SUBTITLES_FILTER) += vf_subtitles.o
OBJS-$(CONFIG_SUPER2XSAI_FILTER) += vf_super2xsai.o
OBJS-$(CONFIG_SWAPUV_FILTER) += vf_swapuv.o
+OBJS-$(CONFIG_TBLEND_FILTER) += vf_blend.o
OBJS-$(CONFIG_TDIFF_FILTER) += vf_tdiff.o
OBJS-$(CONFIG_TELECINE_FILTER) += vf_telecine.o
OBJS-$(CONFIG_THUMBNAIL_FILTER) += vf_thumbnail.o
diff --git a/libavfilter/allfilters.c b/libavfilter/allfilters.c
index f13f82d..e8914cc 100644
--- a/libavfilter/allfilters.c
+++ b/libavfilter/allfilters.c
@@ -202,6 +202,7 @@ void avfilter_register_all(void)
REGISTER_FILTER(SUBTITLES, subtitles, vf);
REGISTER_FILTER(SUPER2XSAI, super2xsai, vf);
REGISTER_FILTER(SWAPUV, swapuv, vf);
+REGISTER_FILTER(TBLEND, tblend, vf);
REGISTER_FILTER(TDIFF, tdiff, vf);
REGISTER_FILTER(TELECINE, telecine, vf);
REGISTER_FILTER(THUMBNAIL, thumbnail, vf);
diff --git a/libavfilter/vf_blend.c b/libavfilter/vf_blend.c
index 8bf19ff..3c1572c 100644
--- a/libavfilter/vf_blend.c
+++ b/libavfilter/vf_blend.c
@@ -95,6 +95,8 @@ typedef struct {
double all_opacity;
FilterParams params[4];
+int tblend;
+AVFrame *prev_frame;/* only used with tblend */
} BlendContext;
#define OFFSET(x) offsetof(BlendContext, x)
@@ -285,7 +287,8 @@ static AVFrame *blend_frame(AVFilterContext *ctx, AVFrame *top_buf,
ctx-internal-execute(ctx, filter_slice, td, NULL, FFMIN(outh, ctx-graph-nb_threads));
}
-av_frame_free(top_buf);
+if (!b-tblend)
+av_frame_free(top_buf);
return dst_buf;
}
@@ -295,6 +298,8 @@ static av_cold int init(AVFilterContext *ctx)
BlendContext *b = ctx-priv;
int ret, plane;
+b-tblend = !strcmp(ctx-filter-name, tblend);
+
for (plane = 0; plane FF_ARRAY_ELEMS(b-params); plane++) {
FilterParams *param = b-params[plane];
@@ -412,6 +417,8 @@ static av_cold void uninit(AVFilterContext *ctx)
int i;
ff_dualinput_uninit(b-dinput);
+av_freep(b-prev_frame);
+
for (i = 0; i FF_ARRAY_ELEMS(b-params); i++)
av_expr_free(b-params[i].e);
}
@@ -463,3 +470,67 @@ AVFilter ff_vf_blend = {
.priv_class= blend_class,
.flags = AVFILTER_FLAG_SUPPORT_TIMELINE_INTERNAL | AVFILTER_FLAG_SLICE_THREADS,
};
+
+static int tblend_config_output(AVFilterLink *outlink)
+{
+AVFilterContext *ctx = outlink-src;
+AVFilterLink *inlink = ctx-inputs[0];
+BlendContext *b = ctx-priv;
+const AVPixFmtDescriptor *pix_desc = av_pix_fmt_desc_get(inlink-format);
+
+b-hsub = pix_desc-log2_chroma_w;
+b-vsub = pix_desc-log2_chroma_h;
+b-nb_planes = av_pix_fmt_count_planes(inlink-format);
+outlink-flags |= FF_LINK_FLAG_REQUEST_LOOP;
+
+return 0;
+}
+
+static int tblend_filter_frame(AVFilterLink *inlink, AVFrame *frame)
+{
+BlendContext *b = inlink-dst-priv;
+AVFilterLink *outlink = inlink-dst-outputs[0];
+
+if (b-prev_frame) {
+AVFrame *out =
Re: [FFmpeg-devel] [PATCH 3/3] lavc/libxvid: workaround for bug in libxvidcore
On Thu, Dec 04, 2014 at 12:39:29AM +0100, Lukasz Marek wrote:
On 24.11.2014 01:17, Lukasz Marek wrote:
libxvidcore calculate number of threads basing on video height.
If height is small enough it allocates 0 bytes long memory and
writes to it.
Setting thread_count to 0 uses 1 thread and skips bugged code.
Signed-off-by: Lukasz Marek lukasz.m.lu...@gmail.com
---
libavcodec/libxvid.c | 11 +++
1 file changed, 11 insertions(+)
diff --git a/libavcodec/libxvid.c b/libavcodec/libxvid.c
index 70d52b9..efe7b8a 100644
--- a/libavcodec/libxvid.c
+++ b/libavcodec/libxvid.c
@@ -471,6 +471,17 @@ static av_cold int xvid_encode_init(AVCodecContext
*avctx)
xvid_enc_create.num_zones = 0;
xvid_enc_create.num_threads = avctx-thread_count;
+/* workaround for a bug in libxvidcore */
+if (avctx-height 18) {
+if (avctx-thread_count 2) {
+xvid_enc_create.num_threads = 0;
+} else {
+av_log(avctx, AV_LOG_ERROR,
+ Too small height for threads 1.);
+ret = AVERROR(EINVAL);
+goto fail;
+}
+}
xvid_enc_create.plugins = plugins;
xvid_enc_create.num_plugins = 0;
Please comment this patch. xvid accepted fix for it.
Can you add a version check, or at least a comment saying what versions
are affected by this?
Personally I don't care, but when configured correctly ffmpeg is using not
allocated memory via external library.
I just don't know if keep it or just forget about it.
--
Clément B.
pgplwv3XvnNEk.pgp
Description: PGP signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH] Incorrect aspect ratio in MPEG2 essence
Hi, Setting setting sample_aspect_ratio in AVContext does not produce the expected aspect index value in the MPEG2 sequence header. Problem seems to be around line 263 in mpeg12enc.c. Setting the sample_aspect_ratio to 16/9 leads the current error minimization routine to determine that error -= (1LL32)*(221/100)*(1080/1920) is smaller than error -= (1LL32)*(16/9)*(1080/1920) thus picking aspect index value 4 rather than 3. Proposed fix in attached patch. -- Erik Johansson mpeg2_aspect.patch Description: Binary data ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 2/2] Adding closed caption decoder
Le quartidi 14 frimaire, an CCXXIII, Anshul a écrit :
Yes while doing that only I realized, if we use i then cc must be in
first stream.
When I had multiple stream(with audio), then I was not able to map subcc
to my video stream.
using stream_idx , we are saying that if we can use out0+subcc and
out1+subcc.
i logic fails when there is audio in 1st stream, there was no way i
could get the subtitle
if my first stream is not the one where closed caption are present.
I tested various combinations, and I find it works with my i version and
not with your stream_idx version. You can try this:
http://samples.ffmpeg.org/MPEG2/subcc/ClosedCaptions/Starship_Troopers.vob
http://samples.ffmpeg.org/ffmpeg-bugs/trac/ticket1482/NCIS.wtv
st=~/tmp/samples/Starship_Troopers_with_subcc.vob
ncis=~/tmp/samples/NCIS_with_subcc_stream.wtv
input=movie=${st}:s=0+1[out0+subcc][out1];movie=${ncis}:s=2+0[out2+subcc][out3]
function prefix { sed s/^/$1 : / }
./ffmpeg_g -loglevel error -f lavfi -i $input -y \
-map 0:s:0 -f srt (prefix 1_) \
-map 0:s:1 -f srt (prefix _2)
(the captions in ALL CAPS are from ST, the other ones from NCIS)
The mapping logic is this:
- ${st}:s=0+1[out0+subcc][out1]: take stream #0 (video) and #1 (audio)
from ${st} and make them stream #0 and #1 for the device.
- ${ncis}:s=2+0[out2+subcc][out3]: take stream #2 (video) and #0 (audio)
from ${ncis} and make them stream #2 and #3 for the device.
- out0+subcc: output stream #0 (video from $st) has CC, make it the first
extra stream, i.e. #4 for the device.
- out2+subcc: output stream #2 (video from $ncis) has CC, make it the next
extra stream, i.e. #5 for the device.
- -map 0:s:0: take the first subtitle stream, i.e. #4, i.e. CC for #2,
i.e. CC for video from $st.
- -map 0:s:1: take the second subtitle stream, i.e. #5.
In other words:
lavfi:4 = CC for lavfi:0 = CC for st:0
lavfi:5 = CC for lavfi:2 = CC for ncis:2
CC streams are created in output stream index order, or, in other words, in
order of their X index in outX+subcc.
Regards,
--
Nicolas George
signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] web: Update Bower dependencies
On Wed, Dec 03, 2014 at 08:56:11PM -0800, Timothy Gu wrote:
On Tue, Nov 25, 2014 at 7:50 PM, Timothy Gu timothyg...@gmail.com wrote:
Also explicitly add jQuery as a dependency. It is copied from the Makefile.
Signed-off-by: Timothy Gu timothyg...@gmail.com
---
bower.json| 9 +-
htdocs/css/bootstrap.min.css | 4 +-
htdocs/css/font-awesome.min.css | 4 +-
htdocs/fonts/FontAwesome.otf | Bin 75188 - 85908 bytes
htdocs/fonts/fontawesome-webfont.eot | Bin 72449 - 56006 bytes
htdocs/fonts/fontawesome-webfont.svg | 942
+-
htdocs/fonts/fontawesome-webfont.ttf | Bin 141564 - 112160 bytes
htdocs/fonts/fontawesome-webfont.woff | Bin 83760 - 65452 bytes
htdocs/js/bootstrap.min.js| 5 +-
9 files changed, 491 insertions(+), 473 deletions(-)
Ping.
Applying: web: Update Bower dependencies
fatal: corrupt patch at line 273
patching file bower.json
patching file htdocs/css/bootstrap.min.css
patch: malformed patch at line 46:
tgroup,select,textarea{margin:0;font:inherit;color:inherit}button{overflow:visible}button,select{text-transform:none}button,html
input[type=button],input[type=reset],input[type=submit]{-webkit-appearance:button;cursor:pointer}button[disabled],html
input[disabled]{cursor:default}button::-moz-focus-inner,input::-moz-focus-inner{padding:0;border:0}input{line-height:normal}input[type=checkbox],input[type=radio]{-webkit-box-sizing:border-box;-moz-box-sizing:border-box;box-sizing:border-box;padding:0}input[type=number]::-webkit-inner-spin-button,input[type=number]::-webkit-outer-spin-button{height:auto}input[type=search]{-webkit-box-sizing:content-box;-moz-box-sizing:content-box;box-sizing:content-box;-webkit-appearance:textfield}input[type=search]::-webkit-search-cancel-button,input[type=search]::-webkit-search-decoration{-webkit-appearance:none}fieldset{padding:.35em
.625em .75em;margin:0 2px;border:1px solid
silver}legend{padding:0;border:0}textarea{overflow:auto}optgroup{font-
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
I have often repented speaking, but never of holding my tongue.
-- Xenocrates
signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 4/4] style: Update from ffmpeg-web
On Wed, Dec 03, 2014 at 09:34:11PM -0800, Timothy Gu wrote: Signed-off-by: Timothy Gu timothyg...@gmail.com --- Forgot this one -- doc/style.min.css | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) Applying: style: Update from ffmpeg-web fatal: corrupt patch at line 31 [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB No snowflake in an avalanche ever feels responsible. -- Voltaire signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH] lavd/lavfi: allow to extract subcc.
Signed-off-by: Nicolas George geo...@nsup.org
---
doc/indevs.texi | 14
libavdevice/lavfi.c | 92 +
2 files changed, 99 insertions(+), 7 deletions(-)
Correctly rebased today, with some extra doc.
If the decoder is applied before this patch, then we can change the example
to output SRT or ASS, this would be even better.
diff --git a/doc/indevs.texi b/doc/indevs.texi
index ad823ab..abc4cc3 100644
--- a/doc/indevs.texi
+++ b/doc/indevs.texi
@@ -495,6 +495,14 @@ generated by the device.
The first unlabelled output is automatically assigned to the out0
label, but all the others need to be specified explicitly.
+The suffix +subcc can be appended to the output label to create an extra
+stream with the closed captions packets attached to that output
+(experimental; only for EIA-608 / CEA-708 for now).
+The subcc streams are created after all the normal streams, in the order of
+the corresponding stream.
+For example, if there is out19+subcc, out7+subcc and up to out42, the
+stream #43 is subcc for stream #7 and stream #44 is subcc for stream #19.
+
If not specified defaults to the filename specified for the input
device.
@@ -541,6 +549,12 @@ Read an audio stream and a video stream and play it back
with
ffplay -f lavfi movie=test.avi[out0];amovie=test.wav[out1]
@end example
+@item
+Dump decoded frames to images and closed captions to a file (experimental):
+@example
+ffmpeg -f lavfi -i movie=test.ts[out0+subcc] -map v frame%08d.png -map s -c
copy -f rawvideo subcc.bin
+@end example
+
@end itemize
@section libcdio
diff --git a/libavdevice/lavfi.c b/libavdevice/lavfi.c
index 1398ece..d0e6bf5 100644
--- a/libavdevice/lavfi.c
+++ b/libavdevice/lavfi.c
@@ -51,7 +51,10 @@ typedef struct {
int *sink_stream_map;
int *sink_eof;
int *stream_sink_map;
+int *sink_stream_subcc_map;
AVFrame *decoded_frame;
+int nb_sinks;
+AVPacket subcc_packet;
} LavfiContext;
static int *create_all_formats(int n)
@@ -82,6 +85,7 @@ av_cold static int lavfi_read_close(AVFormatContext *avctx)
av_freep(lavfi-sink_stream_map);
av_freep(lavfi-sink_eof);
av_freep(lavfi-stream_sink_map);
+av_freep(lavfi-sink_stream_subcc_map);
av_freep(lavfi-sinks);
avfilter_graph_free(lavfi-graph);
av_frame_free(lavfi-decoded_frame);
@@ -89,6 +93,27 @@ av_cold static int lavfi_read_close(AVFormatContext *avctx)
return 0;
}
+static int create_subcc_streams(AVFormatContext *avctx)
+{
+LavfiContext *lavfi = avctx-priv_data;
+AVStream *st;
+int stream_idx, sink_idx;
+
+for (stream_idx = 0; stream_idx lavfi-nb_sinks; stream_idx++) {
+sink_idx = lavfi-stream_sink_map[stream_idx];
+if (lavfi-sink_stream_subcc_map[sink_idx]) {
+lavfi-sink_stream_subcc_map[sink_idx] = avctx-nb_streams;
+if (!(st = avformat_new_stream(avctx, NULL)))
+return AVERROR(ENOMEM);
+st-codec-codec_id = AV_CODEC_ID_EIA_608;
+st-codec-codec_type = AVMEDIA_TYPE_SUBTITLE;
+} else {
+lavfi-sink_stream_subcc_map[sink_idx] = -1;
+}
+}
+return 0;
+}
+
av_cold static int lavfi_read_header(AVFormatContext *avctx)
{
LavfiContext *lavfi = avctx-priv_data;
@@ -153,6 +178,7 @@ av_cold static int lavfi_read_header(AVFormatContext *avctx)
/* count the outputs */
for (n = 0, inout = output_links; inout; n++, inout = inout-next);
+lavfi-nb_sinks = n;
if (!(lavfi-sink_stream_map = av_malloc(sizeof(int) * n)))
FAIL(AVERROR(ENOMEM));
@@ -160,6 +186,8 @@ av_cold static int lavfi_read_header(AVFormatContext *avctx)
FAIL(AVERROR(ENOMEM));
if (!(lavfi-stream_sink_map = av_malloc(sizeof(int) * n)))
FAIL(AVERROR(ENOMEM));
+if (!(lavfi-sink_stream_subcc_map = av_malloc(sizeof(int) * n)))
+FAIL(AVERROR(ENOMEM));
for (i = 0; i n; i++)
lavfi-stream_sink_map[i] = -1;
@@ -167,14 +195,22 @@ av_cold static int lavfi_read_header(AVFormatContext
*avctx)
/* parse the output link names - they need to be of the form out0, out1,
...
* create a mapping between them and the streams */
for (i = 0, inout = output_links; inout; i++, inout = inout-next) {
-int stream_idx;
-if (!strcmp(inout-name, out))
-stream_idx = 0;
-else if (sscanf(inout-name, out%d\n, stream_idx) != 1) {
+int stream_idx = 0, suffix = 0, use_subcc = 0;
+sscanf(inout-name, out%n%d%n, suffix, stream_idx, suffix);
+if (!suffix) {
av_log(avctx, AV_LOG_ERROR,
Invalid outpad name '%s'\n, inout-name);
FAIL(AVERROR(EINVAL));
}
+if (inout-name[suffix]) {
+if (!strcmp(inout-name + suffix, +subcc)) {
+use_subcc = 1;
+} else {
+av_log(avctx, AV_LOG_ERROR,
+ Invalid outpad suffix '%s'\n,
Re: [FFmpeg-devel] [PATCH] avfilter/tinterlace: merge code with interlace
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 04/12/14 09:59, Clément Bœsch wrote: On Thu, Dec 04, 2014 at 09:31:47AM +, tim nicholson wrote: On 02/12/14 18:43, Clément Bœsch wrote: --- After this commit, interlace doesn't behave the same if the input frames are already interlaced. Does anyone wants me to keep this behaviour? Also, the PTS adjustments might be different, but I didn't follow the current discussions about it so I can't tell what people want. --- libavfilter/Makefile| 2 +- libavfilter/interlace.h | 58 - libavfilter/tinterlace.h| 5 +- libavfilter/vf_interlace.c | 250 libavfilter/vf_tinterlace.c | 44 +++ libavfilter/x86/Makefile| 2 +- libavfilter/x86/vf_interlace_init.c | 47 --- 7 files changed, 49 insertions(+), 359 deletions(-) delete mode 100644 libavfilter/interlace.h delete mode 100644 libavfilter/vf_interlace.c delete mode 100644 libavfilter/x86/vf_interlace_init.c [..] The docs will need updating to remove the vf_interlace entry and add the lp filter options. Maybe that is the time to merge the ascii art graphics and add Michael's additional mode ones as well? None of the filter disappear and their options don't change. It's just code refactoring. Sorry misread it only looking at the diff. I thought there was a funtionality merger as well as per the irc discussion. [..] - -- Tim. Key Fingerprint 38CF DB09 3ED0 F607 8B67 6CED 0C0B FC44 8B0B FC83 -BEGIN PGP SIGNATURE- Version: GnuPG v2 iQEcBAEBAgAGBQJUgFaLAAoJEAwL/ESLC/yDRGkH/j2ZXp+JcWgfvTIG5+/U5Nrk uDA9Gpn+QQKQHjaF+AHCL8d1KPmXlnKgbha2r9V4EBHiSyZDTNOcio9T0IvY+YDG wbnTuCSV6SaBgm8LvKrvB2B/OC4aMFoP8aYyE10Ghqkd2HHG7jGcctPrlAIdmvpP K6VbHFP+RKuvQLvQPX2SPigcGvUQV566azYguzum82k4CV4+kwugUCDNgGeds8rO ySy/DoeTai0cm5A+PA3ojKMaWkuZt5AN6WFFGcYvamwLDa/WZfK8SlcpvPZoH9y9 9DYizi3PDOfyojpXKoMsi6n3zRpKJ9x7TETA+hUZ8mcJDxnhCvFhYIDeQXrt8LY= =m1ni -END PGP SIGNATURE- ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 2/2] Adding closed caption decoder
On Thu, Dec 4, 2014 at 5:43 PM, Nicolas George geo...@nsup.org wrote:
Le quartidi 14 frimaire, an CCXXIII, Anshul a écrit :
Yes while doing that only I realized, if we use i then cc must be in
first stream.
When I had multiple stream(with audio), then I was not able to map subcc
to my video stream.
using stream_idx , we are saying that if we can use out0+subcc and
out1+subcc.
i logic fails when there is audio in 1st stream, there was no way i
could get the subtitle
if my first stream is not the one where closed caption are present.
I tested various combinations, and I find it works with my i version and
not with your stream_idx version. You can try this:
http://samples.ffmpeg.org/MPEG2/subcc/ClosedCaptions/Starship_Troopers.vob
http://samples.ffmpeg.org/ffmpeg-bugs/trac/ticket1482/NCIS.wtv
st=~/tmp/samples/Starship_Troopers_with_subcc.vob
ncis=~/tmp/samples/NCIS_with_subcc_stream.wtv
input=movie=${st}:s=0+1[out0+subcc][out1];movie=${ncis}:s=2+0[out2+subcc][out3]
function prefix { sed s/^/$1 : / }
./ffmpeg_g -loglevel error -f lavfi -i $input -y \
-map 0:s:0 -f srt (prefix 1_) \
-map 0:s:1 -f srt (prefix _2)
(the captions in ALL CAPS are from ST, the other ones from NCIS)
The mapping logic is this:
- ${st}:s=0+1[out0+subcc][out1]: take stream #0 (video) and #1 (audio)
from ${st} and make them stream #0 and #1 for the device.
- ${ncis}:s=2+0[out2+subcc][out3]: take stream #2 (video) and #0 (audio)
from ${ncis} and make them stream #2 and #3 for the device.
- out0+subcc: output stream #0 (video from $st) has CC, make it the first
extra stream, i.e. #4 for the device.
- out2+subcc: output stream #2 (video from $ncis) has CC, make it the
next
extra stream, i.e. #5 for the device.
- -map 0:s:0: take the first subtitle stream, i.e. #4, i.e. CC for #2,
i.e. CC for video from $st.
- -map 0:s:1: take the second subtitle stream, i.e. #5.
In other words:
lavfi:4 = CC for lavfi:0 = CC for st:0
lavfi:5 = CC for lavfi:2 = CC for ncis:2
CC streams are created in output stream index order, or, in other words, in
order of their X index in outX+subcc.
Regards,
--
Nicolas George
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Yes your logic is working fine when you know where is video stream and
where is audio stream.
while with stream_idx, even if I you dont know where is video and audio it
works.
but seeing so may of your use case, I feel you are right,
I think there should be some more way of specifying 0+1, we should also
able to specify 0v+0a
means 1st video and 1st audio in stream.
other wise there are lots of chances of wrong mapping. In my particular use
case I take input from rtp.
and the guys does not have any standard for 1st will be video or audio.
As you said before as per stephan's comment you are making stricter
parsing, please consider something like 0v + 0a.
Thanks
-Anshul
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] avfilter/tinterlace: merge code with interlace
On Thu, Dec 04, 2014 at 12:41:47PM +, tim nicholson wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 04/12/14 09:59, Clément Bœsch wrote: On Thu, Dec 04, 2014 at 09:31:47AM +, tim nicholson wrote: On 02/12/14 18:43, Clément Bœsch wrote: --- After this commit, interlace doesn't behave the same if the input frames are already interlaced. Does anyone wants me to keep this behaviour? Also, the PTS adjustments might be different, but I didn't follow the current discussions about it so I can't tell what people want. --- libavfilter/Makefile| 2 +- libavfilter/interlace.h | 58 - libavfilter/tinterlace.h| 5 +- libavfilter/vf_interlace.c | 250 libavfilter/vf_tinterlace.c | 44 +++ libavfilter/x86/Makefile| 2 +- libavfilter/x86/vf_interlace_init.c | 47 --- 7 files changed, 49 insertions(+), 359 deletions(-) delete mode 100644 libavfilter/interlace.h delete mode 100644 libavfilter/vf_interlace.c delete mode 100644 libavfilter/x86/vf_interlace_init.c [..] The docs will need updating to remove the vf_interlace entry and add the lp filter options. Maybe that is the time to merge the ascii art graphics and add Michael's additional mode ones as well? None of the filter disappear and their options don't change. It's just code refactoring. Sorry misread it only looking at the diff. I thought there was a funtionality merger as well as per the irc discussion. There are 2 functional changes, which is why I haven't pushed yet: - the do nothing with already interlaced frame behaviour from vf_interlace is not kept - the PTS adjustments might not be the same so I might need to set the exact_tb flag internally if it's vf_interlace to copy the behaviour (but I need confirmation from the people who are using this filter) [...] -- Clément B. pgp8_s1qotVd_.pgp Description: PGP signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 2/2] Adding closed caption decoder
Le quartidi 14 frimaire, an CCXXIII, Anshul Maheshwari a écrit : while with stream_idx, even if I you dont know where is video and audio it works. I do not think so, I suspect it only worked by chance. You may as well end up trying to extract subcc from audio, or in a completely random order. I think there should be some more way of specifying 0+1, we should also able to specify 0v+0a In the s= option you mean? This is the realm of the movie filter, this is completely unrelated, and it is already possible as 0:v; unfortunately, due to the comma character, it falls into escaping hell. (Have to go for now.) Regards, -- Nicolas George signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] Fix bug for POWER LE: libavcodec/ppc/vp8dsp_altivec.c
On Thu, Dec 04, 2014 at 06:48:18PM +0800, rongyan wrote: Hi, We propose this patch to fix vp8 bugs for POWER8 little endian. Functions put_vp8_epel_h_altivec_core(), puta_vp8_epel_v_altivec_core(), and put_vp8_pixels16_altivec() are fixed. The passed test cases change from 2296/2328 to 2328/2328. Thanks for your review. Rong Yan -- The world has enough for everyone's need, but not enough for everyone's greed. vp8dsp_altivec.c | 86 +++ 1 file changed, 49 insertions(+), 37 deletions(-) ad171a45e64517a1bfc0fda49ae14d778003c248 0001-avcodec-ppc-vp8dsp_altivec.c-POWER-LE-support-put_vp.patch From b1b4001cfed4d42f9fafb2364a3f935cc9a5b340 Mon Sep 17 00:00:00 2001 From: Rong Yan rongyan...@gmail.com Date: Thu, 4 Dec 2014 10:34:02 + Subject: [PATCH] avcodec/ppc/vp8dsp_altivec.c: POWER LE support put_vp8_epel_h_altivec_core() put_vp8_epel_v_altivec_core() put_vp8_pixels16_altivec() add marcos GET_PIXHL() GET_OUTER() LOAD_HL() applied thanks [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB It is dangerous to be right in matters on which the established authorities are wrong. -- Voltaire signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] revised patches for dv profile
- fixed defect with invalid framerate - included libavformat\dvenc.c in second patch Steve ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH 1/2] add av_dv_codec_profile2 : uses framerate to select best matching profile. default on first matching profile
Signed-off-by: Steve Jiekak devaure...@gmail.com
---
doc/APIchanges |3 +++
libavcodec/dv_profile.c | 24 ++--
libavcodec/dv_profile.h |6 ++
libavcodec/version.h|2 +-
4 files changed, 32 insertions(+), 3 deletions(-)
diff --git a/doc/APIchanges b/doc/APIchanges
index 66b1278..df87465 100644
--- a/doc/APIchanges
+++ b/doc/APIchanges
@@ -15,6 +15,9 @@ libavutil: 2014-08-09
API changes, most recent first:
+2014-12-04 - xxx - lavc 56.14.100 - dv_profile.h
+ Add av_dv_codec_profile2().
+
8 - FFmpeg 2.5 was cut here 8 -
2014-11-21 - ab922f9 - lavu 54.15.100 - dict.h
diff --git a/libavcodec/dv_profile.c b/libavcodec/dv_profile.c
index b301cbf..e336e08 100644
--- a/libavcodec/dv_profile.c
+++ b/libavcodec/dv_profile.c
@@ -315,15 +315,35 @@ const AVDVProfile *av_dv_codec_profile(int width, int
height,
enum AVPixelFormat pix_fmt)
{
#if CONFIG_DVPROFILE
+return av_dv_codec_profile2(width, height, pix_fmt, (AVRational){0, 0});
+#endif
+
+return NULL;
+}
+
+const AVDVProfile *av_dv_codec_profile2(int width, int height,
+ enum AVPixelFormat pix_fmt,
+ AVRational frame_rate)
+{
+const AVDVProfile *p = NULL;
+#if CONFIG_DVPROFILE
int i;
+/* frame rate is necessary to select between 720p50 and 720p60 profiles */
+int invalid_framerate = frame_rate.num == 0 || frame_rate.den == 0;
for (i = 0; i FF_ARRAY_ELEMS(dv_profiles); i++)
if (height == dv_profiles[i].height
pix_fmt == dv_profiles[i].pix_fmt
width == dv_profiles[i].width)
-return dv_profiles[i];
+{
+if( invalid_framerate || av_div_q(dv_profiles[i].time_base,
frame_rate).num == 1 )
+return dv_profiles[i];
+
+if(!p)
+p = dv_profiles[i];
+}
#endif
-return NULL;
+return p;
}
diff --git a/libavcodec/dv_profile.h b/libavcodec/dv_profile.h
index d4437c9..d22ad26 100644
--- a/libavcodec/dv_profile.h
+++ b/libavcodec/dv_profile.h
@@ -83,4 +83,10 @@ const AVDVProfile *av_dv_frame_profile(const AVDVProfile
*sys,
*/
const AVDVProfile *av_dv_codec_profile(int width, int height, enum
AVPixelFormat pix_fmt);
+/**
+ * Get a DV profile for the provided stream parameters.
+ * The frame rate is used as a best-effort parameter.
+ */
+const AVDVProfile *av_dv_codec_profile2(int width, int height, enum
AVPixelFormat pix_fmt, AVRational frame_rate);
+
#endif /* AVCODEC_DV_PROFILE_H */
diff --git a/libavcodec/version.h b/libavcodec/version.h
index 23443ed..ef439d6 100644
--- a/libavcodec/version.h
+++ b/libavcodec/version.h
@@ -29,7 +29,7 @@
#include libavutil/version.h
#define LIBAVCODEC_VERSION_MAJOR 56
-#define LIBAVCODEC_VERSION_MINOR 13
+#define LIBAVCODEC_VERSION_MINOR 14
#define LIBAVCODEC_VERSION_MICRO 100
#define LIBAVCODEC_VERSION_INT AV_VERSION_INT(LIBAVCODEC_VERSION_MAJOR, \
--
1.7.0.4
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] doc/filters: Add ascii graphics to clarify what the currently implemented tinterlace modes do
On Thu, Dec 04, 2014 at 04:03:00PM +, tim nicholson wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 04/12/14 08:24, tim nicholson wrote: On 02/12/14 17:13, Michael Niedermayer wrote: On Tue, Dec 02, 2014 at 05:02:49PM +, tim nicholson wrote: On 02/12/14 15:33, Michael Niedermayer wrote: Signed-off-by: Michael Niedermayer michae...@gmx.at --- doc/filters.texi | 131 ++ 1 file changed, 131 insertions(+) diff --git a/doc/filters.texi b/doc/filters.texi index 8c16c7a..0ea3955 100644 --- a/doc/filters.texi +++ b/doc/filters.texi @@ -8537,33 +8537,164 @@ Available values are: @item merge, 0 Move odd frames into the upper field, even into the lower field, generating a double height frame at half frame rate. +@example + -- time +Input: +Frame 1 Frame 2 Frame 3 Frame 4 + +1 2 3 4 +1 2 3 4 +1 2 3 4 +1 2 3 4 + +Output: +1 3 +2 4 +1 3 +2 4 +1 3 +2 4 +1 3 +2 4 [..] I find the ascii graphic style of the interlace filter more understandable how do you suggest above (and the other modes) would be documented in that style ? Sorry for the delayed reply ISP issue... I think the main difference is going left to right, not top to bottom for input to output, this then matches not only interlace but the general filtergraph descriptions. I will try and see if I can give an example if you like. Attached, follows style of existing interlace filter ascii art. [...] - -- Tim. Key Fingerprint 38CF DB09 3ED0 F607 8B67 6CED 0C0B FC44 8B0B FC83 -BEGIN PGP SIGNATURE- Version: GnuPG v2 iQEcBAEBAgAGBQJUgIWzAAoJEAwL/ESLC/yDGVMIAL/B/xLvAMPYUUW63RHxTiCP XQpKYKtfsFR6+QaC+pql1WSl8mYvlFRJUYLOm6ESd9JS72gLv6UTs1rLdITKaRdo SkSVLBWiVAdY8nzjNESjxWnolelM1JzcEopnfqT1hcxeGnyAx+IMHBd3pW0+r/6U F1BSnTM3yCELBvS/t7d4ZlpZkw4IDO2Lx0qseKi6CvQA/8ehfP9TcRbfdVgotcTf 4zorBilGKhZgaqpftxz8Kxbcldv/xrCczBOWVbB3sZRp4A6zLI9Xuc1c2hIyOia9 EbvHOPmyEBBKdoa31aK8pjFf8gqXF4pM+ZbeaYfR00/WRJC2PbB+tKBQIXXNX4M= =cLiK -END PGP SIGNATURE- From 365723f7bae6bd5e233f6c5550826cd9aeb956f2 Mon Sep 17 00:00:00 2001 From: Tim Nicholson tim.nichol...@bbc.co.uk Date: Thu, 4 Dec 2014 15:55:59 + Subject: [PATCH] doc/filters: Add ascii graphics to clarify what the currently implemented tinterlace modes do Signed-off-by: Tim Nicholson tim.nichol...@bbc.co.uk --- doc/filters.texi | 151 +++ 1 file changed, 151 insertions(+) diff --git a/doc/filters.texi b/doc/filters.texi index 8c16c7a..01f1f63 100644 --- a/doc/filters.texi +++ b/doc/filters.texi @@ -8537,26 +8537,116 @@ Available values are: @item merge, 0 Move odd frames into the upper field, even into the lower field, generating a double height frame at half frame rate. +@example + OriginalOriginal New Frame + Frame 'j' Frame 'j+1' + == === == +Line 0 Frame 'j' Line 0 +Line 0 Frame 'j+1' Line 1 +Line 1 - Frame 'j' Line 2 +Line 1 Frame 'j+1' Line 3 +Line 2 Frame 'j' Line 4 +Line 2 Frame 'j+1' Line 5 +Line 3 - Frame 'j' Line 6 +Line 3 Frame 'j+1' Line 7 +@end example @item drop_odd, 1 Only output even frames, odd frames are dropped, generating a frame with unchanged height at half frame rate. +@example + OriginalOriginal New Frame + Frame 'j' Frame 'j+1' + == === == +Line 0 +Line 0 Frame 'j+1' Line 0 +Line 1 +Line 1 Frame 'j+1' Line 1 +Line 2 +Line 2 Frame 'j+1' Line 2 +Line 3 +Line 3 Frame 'j+1' Line 3 +@end example @item drop_even, 2 Only output odd frames, even frames are dropped, generating a frame with unchanged height at half frame rate. +@example + OriginalOriginal New Frame + Frame 'j' Frame 'j+1' + == === == +Line 0 Frame 'j' Line 0 +Line 0 +Line 1 - Frame
Re: [FFmpeg-devel] [PATCH 4/4] avcodec/pngdec: use memcpy instead of byte loops for P frames.
On Thu, Dec 04, 2014 at 03:28:32PM +0100, Benoit Fouet wrote:
Hi,
Le 03/12/2014 14:31, Benoit Fouet a écrit :
Rely on the way memcpy is optimized for one's system instead of looping
on a byte buffer for buffer copies to handle P frames.
---
libavcodec/pngdec.c | 25 +
1 file changed, 13 insertions(+), 12 deletions(-)
diff --git a/libavcodec/pngdec.c b/libavcodec/pngdec.c
index 7e7b285..e6b7593 100644
--- a/libavcodec/pngdec.c
+++ b/libavcodec/pngdec.c
@@ -884,8 +884,7 @@ static int handle_p_frame_apng(AVCodecContext *avctx,
PNGDecContext *s,
ff_thread_await_progress(s-previous_picture, INT_MAX, 0);
for (j = 0; j s-y_offset; j++) {
-for (i = 0; i ls; i++)
-pd[i] = pd_last[i];
+memcpy(pd, pd_last, ls);
pd += s-image_linesize;
pd_last += s-image_linesize;
}
@@ -907,8 +906,9 @@ static int handle_p_frame_apng(AVCodecContext *avctx,
PNGDecContext *s,
}
for (j = s-y_offset; j s-y_offset + s-cur_h; j++) {
-for (i = 0; i s-x_offset * s-bpp; i++)
-pd[i] = pd_last[i];
+i = s-x_offset * s-bpp;
+if (i)
+memcpy(pd, pd_last, i);
for (; i (s-x_offset + s-cur_w) * s-bpp; i += s-bpp) {
uint8_t alpha = pd[i+ai];
@@ -930,26 +930,27 @@ static int handle_p_frame_apng(AVCodecContext *avctx,
PNGDecContext *s,
break;
}
}
-for (; i ls; i++)
-pd[i] = pd_last[i];
+if (ls - i)
+memcpy(pd+i, pd_last+i, ls - i);
pd += s-image_linesize;
pd_last += s-image_linesize;
pd_last_region += s-image_linesize;
}
} else {
for (j = s-y_offset; j s-y_offset + s-cur_h; j++) {
-for (i = 0; i s-x_offset * s-bpp; i++)
-pd[i] = pd_last[i];
-for (i = (s-x_offset + s-cur_w) * s-bpp; i ls; i++)
-pd[i] = pd_last[i];
+int end_offset = (s-x_offset + s-cur_w) * s-bpp;
+int end_len= ls - end_offset;
+if (s-x_offset)
+memcpy(pd, pd_last, s-x_offset * s-bpp);
+if (end_len)
+memcpy(pd+end_offset, pd_last+end_offset, end_len);
pd += s-image_linesize;
pd_last += s-image_linesize;
}
}
for (j = s-y_offset + s-cur_h; j s-height; j++) {
-for (i = 0; i ls; i++)
-pd[i] = pd_last[i];
+memcpy(pd, pd_last, ls);
pd += s-image_linesize;
pd_last += s-image_linesize;
}
Unless there is some more remarks on this one, I think it can be
applied too.
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] Incorrect aspect ratio in MPEG2 essence
On Thu, Dec 04, 2014 at 01:11:08PM +0100, Erik Johansson wrote: Hi, Setting setting sample_aspect_ratio in AVContext does not produce the expected aspect index value in the MPEG2 sequence header. Problem seems to be around line 263 in mpeg12enc.c. Setting the sample_aspect_ratio to 16/9 leads the current error minimization routine to determine that error -= (1LL32)*(221/100)*(1080/1920) is smaller than error -= (1LL32)*(16/9)*(1080/1920) thus picking aspect index value 4 rather than 3. Proposed fix in attached patch. how can this bug be reproduced ? [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB Avoid a single point of failure, be that a person or equipment. signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] doc/filters: Add ascii graphics to clarify what the currently implemented tinterlace modes do
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 04/12/14 16:09, Clément Bœsch wrote: [..] This representation requires way more efforts to me to understand than the one from Michael... Just shows how different minds think/work ;) As the saying goes, one man's meat is another man's poison. I was going for consistency with current work (which I happened to find clear) and I liked having the extra detail of line numbers. [...] - -- Tim. Key Fingerprint 38CF DB09 3ED0 F607 8B67 6CED 0C0B FC44 8B0B FC83 -BEGIN PGP SIGNATURE- Version: GnuPG v2 iQEcBAEBAgAGBQJUgJJlAAoJEAwL/ESLC/yDmdQH/1BHNlTjS9omO0g28P7+TAXu vpGI7kUxJWFH7txUHhRNf33lHbFFmYdKeb/lIfuiPFXvOtUgeTnmARf8RPIN82e4 7LPFIuN9BFuu78GpyjY/ZEQOUhsT8WK/5c/xWzgJDTo87WsqmQcZe1TuJW945U16 mEZnPfCoYnV4OO/Bu68FJjEMHX4MafIB5VKxhK28/6ymbSJl5v8h9uyFzEZ3TZgF 48bSSYNM4T1KwkJPQsDN+0X5YXeAjA+ZTbB1K5qBQAOZYY4Qj2oCSetl09G6XynG FjeYUxsDCwI97/+x2i+0K6UIBj9Ey/iJRUat1QP4F1FoHQKz5WHpMN5/4XYblU0= =ulRP -END PGP SIGNATURE- ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] web: Update Bower dependencies
Did you check that the latest versions of the dependencies still make the website look the same? I remember having retro-compatibility issues with Bootstrap in the past, but that was a long time ago, so it's probably OK. -- db0 - Deby Barbara Lepage http://db0.fr/ ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] imc: make issue non-fatal
On Thu, Dec 04, 2014 at 12:04:50AM +0100, Christophe Gisquet wrote: Until a noticeably distorted output or crash is reported, it's better ignoring the error, as this results in better output on the affected samples. (ticket #3886) On the other hand, I have a hard time understanding the purpose of that check. I would have expected maybe get_bits_count(q-gb) cw_len, but the next parsing step, inverse_quant_coeff, does not even check what's left. -- Christophe imc.c | 15 ++- 1 file changed, 6 insertions(+), 9 deletions(-) 4334ed60bfabb1f4679940664aeece3d4274a22a 0001-imc-make-issue-non-fatal.patch From 90b5642bd53b24f66f1c36c070b4c490dd72a6cf Mon Sep 17 00:00:00 2001 From: Christophe Gisquet christophe.gisq...@gmail.com Date: Wed, 3 Dec 2014 23:52:52 +0100 Subject: [PATCH] imc: make issue non-fatal The faulty samples actually sound fine when ignoring this issue. For ticket #3886, more samples are decoded. applied also i see at least one bug in the original code, will fix it thanks [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB Those who are too smart to engage in politics are punished by being governed by those who are dumber. -- Plato 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] replaced av_dv_codec_profile by av_dv_codec_profile2 in encoder and dv muxers
On Thu, Dec 04, 2014 at 02:48:15PM +0100, Steve Jiekak wrote: Signed-off-by: Steve Jiekak devaure...@gmail.com --- libavcodec/dvenc.c |2 +- libavformat/dv.c|4 ++-- libavformat/dvenc.c |3 ++- 3 files changed, 5 insertions(+), 4 deletions(-) applied thx [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB If you think the mosad wants you dead since a long time then you are either wrong or dead since a long time. signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 1/2] add av_dv_codec_profile2 : uses framerate to select best matching profile. default on first matching profile
On Thu, Dec 04, 2014 at 02:48:14PM +0100, Steve Jiekak wrote: Signed-off-by: Steve Jiekak devaure...@gmail.com --- doc/APIchanges |3 +++ libavcodec/dv_profile.c | 24 ++-- libavcodec/dv_profile.h |6 ++ libavcodec/version.h|2 +- 4 files changed, 32 insertions(+), 3 deletions(-) applied thanks [...] -- 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] web: Update Bower dependencies
On Thu, Dec 4, 2014 at 4:22 AM, Michael Niedermayer michae...@gmx.at wrote: On Wed, Dec 03, 2014 at 08:56:11PM -0800, Timothy Gu wrote: On Tue, Nov 25, 2014 at 7:50 PM, Timothy Gu timothyg...@gmail.com wrote: Also explicitly add jQuery as a dependency. It is copied from the Makefile. Signed-off-by: Timothy Gu timothyg...@gmail.com --- bower.json| 9 +- htdocs/css/bootstrap.min.css | 4 +- htdocs/css/font-awesome.min.css | 4 +- htdocs/fonts/FontAwesome.otf | Bin 75188 - 85908 bytes htdocs/fonts/fontawesome-webfont.eot | Bin 72449 - 56006 bytes htdocs/fonts/fontawesome-webfont.svg | 942 +- htdocs/fonts/fontawesome-webfont.ttf | Bin 141564 - 112160 bytes htdocs/fonts/fontawesome-webfont.woff | Bin 83760 - 65452 bytes htdocs/js/bootstrap.min.js| 5 +- 9 files changed, 491 insertions(+), 473 deletions(-) Ping. Applying: web: Update Bower dependencies fatal: corrupt patch at line 273 patching file bower.json patching file htdocs/css/bootstrap.min.css It's probably just your mail client, as I tried applying the mail and it applies cleanly. Trying again with MIME attachment. You can also merge it from https://github.com/TimothyGu/ffmpeg-web.git branch master. Timothy ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] web: Update Bower dependencies
On Thu, Dec 4, 2014 at 9:07 AM, db0 company db0comp...@gmail.com wrote: Did you check that the latest versions of the dependencies still make the website look the same? I did, and it's OK. Timothy ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] web: Update Bower dependencies
On Thu, Dec 04, 2014 at 12:55:05PM -0800, Timothy Gu wrote: And with patch From 51ff20706034f2d74e32a643727c71f404aeebaa Mon Sep 17 00:00:00 2001 From: Timothy Gu timothyg...@gmail.com Date: Tue, 25 Nov 2014 19:50:06 -0800 Subject: [PATCH] web: Update Bower dependencies Also explicitly add jQuery as a dependency. It is copied from the Makefile. applied [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB Observe your enemies, for they first find out your faults. -- Antisthenes signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] libavformat/mxfdec.c refactor resolving MultiDescriptor and remove essence group hack
On Tue, 2014-12-02 at 16:43 -0800, Mark Reid wrote:
I think this is a better way to deal with single frame essence data then my
previous way.
---
libavformat/mxfdec.c | 62
1 file changed, 38 insertions(+), 24 deletions(-)
diff --git a/libavformat/mxfdec.c b/libavformat/mxfdec.c
index 0c88a8a..6c104b9 100644
--- a/libavformat/mxfdec.c
+++ b/libavformat/mxfdec.c
@@ -1508,6 +1508,32 @@ static MXFPackage*
mxf_resolve_source_package(MXFContext *mxf, UID package_uid)
return NULL;
}
+static MXFDescriptor* mxf_resolve_multidescriptor(MXFContext *mxf,
MXFDescriptor *descriptor, int track_id)
+{
+MXFDescriptor *sub_descriptor = NULL;
+int i;
+
+if (!descriptor)
+return NULL;
+
+if (descriptor-type == MultipleDescriptor) {
+for (i = 0; i descriptor-sub_descriptors_count; i++) {
+sub_descriptor = mxf_resolve_strong_ref(mxf,
descriptor-sub_descriptors_refs[i], Descriptor);
+
+if (!sub_descriptor) {
+av_log(mxf-fc, AV_LOG_ERROR, could not resolve sub
descriptor strong ref\n);
+continue;
+}
+if (sub_descriptor-linked_track_id == track_id) {
+return sub_descriptor;
+}
+}
+} else if (descriptor-type == Descriptor)
+return descriptor;
+
+return NULL;
+}
+
static MXFStructuralComponent* mxf_resolve_essence_group_choice(MXFContext
*mxf, MXFEssenceGroup *essence_group)
{
MXFStructuralComponent *component = NULL;
@@ -1529,12 +1555,8 @@ static MXFStructuralComponent*
mxf_resolve_essence_group_choice(MXFContext *mxf,
continue;
descriptor = mxf_resolve_strong_ref(mxf, package-descriptor_ref,
Descriptor);
-if (descriptor){
-/* HACK: force the duration of the component to match the
duration of the descriptor */
-if (descriptor-duration != AV_NOPTS_VALUE)
-component-duration = descriptor-duration;
+if (descriptor)
return component;
-}
}
return NULL;
}
@@ -1735,7 +1757,17 @@ static int mxf_parse_structural_metadata(MXFContext
*mxf)
}
st-id = source_track-track_id;
st-priv_data = source_track;
-source_track-original_duration = st-duration = component-duration;
+
+source_package-descriptor = mxf_resolve_strong_ref(mxf,
source_package-descriptor_ref, AnyType);
+descriptor = mxf_resolve_multidescriptor(mxf,
source_package-descriptor, source_track-track_id);
+
+/* A SourceClip from a EssenceGroup may only be a single frame of
essence data. The clips duration is then how many
+ * frames its suppose to repeat for. Descriptor-duration, if
present, contains the real duration of the essence data */
+if (descriptor descriptor-duration != AV_NOPTS_VALUE)
+source_track-original_duration = st-duration =
FFMIN(descriptor-duration, component-duration);
+else
+source_track-original_duration = st-duration =
component-duration;
+
if (st-duration == -1)
st-duration = AV_NOPTS_VALUE;
st-start_time = component-start_position;
@@ -1758,24 +1790,6 @@ static int mxf_parse_structural_metadata(MXFContext
*mxf)
codec_ul = mxf_get_codec_ul(ff_mxf_data_definition_uls,
source_track-sequence-data_definition_ul);
st-codec-codec_type = codec_ul-id;
-source_package-descriptor = mxf_resolve_strong_ref(mxf,
source_package-descriptor_ref, AnyType);
-if (source_package-descriptor) {
-if (source_package-descriptor-type == MultipleDescriptor) {
-for (j = 0; j
source_package-descriptor-sub_descriptors_count; j++) {
-MXFDescriptor *sub_descriptor =
mxf_resolve_strong_ref(mxf,
source_package-descriptor-sub_descriptors_refs[j], Descriptor);
-
-if (!sub_descriptor) {
-av_log(mxf-fc, AV_LOG_ERROR, could not resolve sub
descriptor strong ref\n);
-continue;
-}
-if (sub_descriptor-linked_track_id ==
source_track-track_id) {
-descriptor = sub_descriptor;
-break;
-}
-}
-} else if (source_package-descriptor-type == Descriptor)
-descriptor = source_package-descriptor;
-}
if (!descriptor) {
av_log(mxf-fc, AV_LOG_INFO, source track %d: stream %d, no
descriptor found\n, source_track-track_id, st-index);
continue;
Seems alright..
/Tomas
signature.asc
Description: This is a digitally signed message part
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
[FFmpeg-devel] parallelization over outputs
Hello, I was wondering if ffmpeg parallelizes over outputs, i.e. if I have one input, but multiple outputs, with different codecs and/or filters, since these outputs are independent from each other they could be rendered/encoded in different threads (or thread groups), is this the case? Best Daniel ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 3/3] lavc/libxvid: workaround for bug in libxvidcore
On 04.12.2014 12:42, Clément Bœsch wrote:
On Thu, Dec 04, 2014 at 12:39:29AM +0100, Lukasz Marek wrote:
On 24.11.2014 01:17, Lukasz Marek wrote:
libxvidcore calculate number of threads basing on video height.
If height is small enough it allocates 0 bytes long memory and
writes to it.
Setting thread_count to 0 uses 1 thread and skips bugged code.
Signed-off-by: Lukasz Marek lukasz.m.lu...@gmail.com
---
libavcodec/libxvid.c | 11 +++
1 file changed, 11 insertions(+)
diff --git a/libavcodec/libxvid.c b/libavcodec/libxvid.c
index 70d52b9..efe7b8a 100644
--- a/libavcodec/libxvid.c
+++ b/libavcodec/libxvid.c
@@ -471,6 +471,17 @@ static av_cold int xvid_encode_init(AVCodecContext *avctx)
xvid_enc_create.num_zones = 0;
xvid_enc_create.num_threads = avctx-thread_count;
+/* workaround for a bug in libxvidcore */
+if (avctx-height 18) {
+if (avctx-thread_count 2) {
+xvid_enc_create.num_threads = 0;
+} else {
+av_log(avctx, AV_LOG_ERROR,
+ Too small height for threads 1.);
+ret = AVERROR(EINVAL);
+goto fail;
+}
+}
xvid_enc_create.plugins = plugins;
xvid_enc_create.num_plugins = 0;
Please comment this patch. xvid accepted fix for it.
Can you add a version check, or at least a comment saying what versions
are affected by this?
Added a check. it was introduced in 1.3.0 and still present in 1.3.3.
Next should be fixed.
From 1bfe00284fdb3dba072147df1efce03c7d29dd35 Mon Sep 17 00:00:00 2001
From: Lukasz Marek lukasz.m.lu...@gmail.com
Date: Mon, 24 Nov 2014 01:12:06 +0100
Subject: [PATCH] lavc/libxvid: workaround for bug in libxvidcore
libxvidcore calculate number of threads basing on video height.
If height is small enough it allocates 0 bytes long memory and
writes to it.
Setting thread_count to 0 uses 1 thread and skips bugged code.
Signed-off-by: Lukasz Marek lukasz.m.lu...@gmail.com
---
libavcodec/libxvid.c | 13 +
1 file changed, 13 insertions(+)
diff --git a/libavcodec/libxvid.c b/libavcodec/libxvid.c
index 046d2f7..64572da 100644
--- a/libavcodec/libxvid.c
+++ b/libavcodec/libxvid.c
@@ -471,6 +471,19 @@ static av_cold int xvid_encode_init(AVCodecContext *avctx)
xvid_enc_create.num_zones = 0;
xvid_enc_create.num_threads = avctx-thread_count;
+#if (XVID_VERSION = 0x010303) (XVID_VERSION = 0x010300)
+/* workaround for a bug in libxvidcore */
+if (avctx-height = 16) {
+if (avctx-thread_count 2) {
+xvid_enc_create.num_threads = 0;
+} else {
+av_log(avctx, AV_LOG_ERROR,
+ Too small height for threads 1.);
+ret = AVERROR(EINVAL);
+goto fail;
+}
+}
+#endif
xvid_enc_create.plugins = plugins;
xvid_enc_create.num_plugins = 0;
--
1.9.1
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH] libavformat/mxfdec.c refactor resolving MultiDescriptor and remove essence group hack
On Thu, Dec 04, 2014 at 11:11:48PM +0100, Tomas Härdin wrote:
On Tue, 2014-12-02 at 16:43 -0800, Mark Reid wrote:
I think this is a better way to deal with single frame essence data then my
previous way.
---
libavformat/mxfdec.c | 62
1 file changed, 38 insertions(+), 24 deletions(-)
diff --git a/libavformat/mxfdec.c b/libavformat/mxfdec.c
index 0c88a8a..6c104b9 100644
--- a/libavformat/mxfdec.c
+++ b/libavformat/mxfdec.c
@@ -1508,6 +1508,32 @@ static MXFPackage*
mxf_resolve_source_package(MXFContext *mxf, UID package_uid)
return NULL;
}
+static MXFDescriptor* mxf_resolve_multidescriptor(MXFContext *mxf,
MXFDescriptor *descriptor, int track_id)
+{
+MXFDescriptor *sub_descriptor = NULL;
+int i;
+
+if (!descriptor)
+return NULL;
+
+if (descriptor-type == MultipleDescriptor) {
+for (i = 0; i descriptor-sub_descriptors_count; i++) {
+sub_descriptor = mxf_resolve_strong_ref(mxf,
descriptor-sub_descriptors_refs[i], Descriptor);
+
+if (!sub_descriptor) {
+av_log(mxf-fc, AV_LOG_ERROR, could not resolve sub
descriptor strong ref\n);
+continue;
+}
+if (sub_descriptor-linked_track_id == track_id) {
+return sub_descriptor;
+}
+}
+} else if (descriptor-type == Descriptor)
+return descriptor;
+
+return NULL;
+}
+
static MXFStructuralComponent* mxf_resolve_essence_group_choice(MXFContext
*mxf, MXFEssenceGroup *essence_group)
{
MXFStructuralComponent *component = NULL;
@@ -1529,12 +1555,8 @@ static MXFStructuralComponent*
mxf_resolve_essence_group_choice(MXFContext *mxf,
continue;
descriptor = mxf_resolve_strong_ref(mxf, package-descriptor_ref,
Descriptor);
-if (descriptor){
-/* HACK: force the duration of the component to match the
duration of the descriptor */
-if (descriptor-duration != AV_NOPTS_VALUE)
-component-duration = descriptor-duration;
+if (descriptor)
return component;
-}
}
return NULL;
}
@@ -1735,7 +1757,17 @@ static int mxf_parse_structural_metadata(MXFContext
*mxf)
}
st-id = source_track-track_id;
st-priv_data = source_track;
-source_track-original_duration = st-duration =
component-duration;
+
+source_package-descriptor = mxf_resolve_strong_ref(mxf,
source_package-descriptor_ref, AnyType);
+descriptor = mxf_resolve_multidescriptor(mxf,
source_package-descriptor, source_track-track_id);
+
+/* A SourceClip from a EssenceGroup may only be a single frame of
essence data. The clips duration is then how many
+ * frames its suppose to repeat for. Descriptor-duration, if
present, contains the real duration of the essence data */
+if (descriptor descriptor-duration != AV_NOPTS_VALUE)
+source_track-original_duration = st-duration =
FFMIN(descriptor-duration, component-duration);
+else
+source_track-original_duration = st-duration =
component-duration;
+
if (st-duration == -1)
st-duration = AV_NOPTS_VALUE;
st-start_time = component-start_position;
@@ -1758,24 +1790,6 @@ static int mxf_parse_structural_metadata(MXFContext
*mxf)
codec_ul = mxf_get_codec_ul(ff_mxf_data_definition_uls,
source_track-sequence-data_definition_ul);
st-codec-codec_type = codec_ul-id;
-source_package-descriptor = mxf_resolve_strong_ref(mxf,
source_package-descriptor_ref, AnyType);
-if (source_package-descriptor) {
-if (source_package-descriptor-type == MultipleDescriptor) {
-for (j = 0; j
source_package-descriptor-sub_descriptors_count; j++) {
-MXFDescriptor *sub_descriptor =
mxf_resolve_strong_ref(mxf,
source_package-descriptor-sub_descriptors_refs[j], Descriptor);
-
-if (!sub_descriptor) {
-av_log(mxf-fc, AV_LOG_ERROR, could not resolve
sub descriptor strong ref\n);
-continue;
-}
-if (sub_descriptor-linked_track_id ==
source_track-track_id) {
-descriptor = sub_descriptor;
-break;
-}
-}
-} else if (source_package-descriptor-type == Descriptor)
-descriptor = source_package-descriptor;
-}
if (!descriptor) {
av_log(mxf-fc, AV_LOG_INFO, source track %d: stream %d, no
descriptor found\n, source_track-track_id, st-index);
continue;
Seems
Re: [FFmpeg-devel] [PATCH 2/2] lavc/options: fix leaks in avcodec_copy_context
On Thu, Dec 04, 2014 at 12:54:30AM +0100, Lukasz Marek wrote: On 27.11.2014 04:17, Michael Niedermayer wrote: On Thu, Nov 27, 2014 at 12:43:57AM +0100, Lukasz Marek wrote: On 24.11.2014 05:16, Lukasz Marek wrote: Signed-off-by: Lukasz Marek lukasz.m.lu...@gmail.com --- libavcodec/options.c | 23 +++ 1 file changed, 11 insertions(+), 12 deletions(-) diff --git a/libavcodec/options.c b/libavcodec/options.c index 7f9fb07..8ba997c 100644 --- a/libavcodec/options.c +++ b/libavcodec/options.c @@ -190,6 +190,11 @@ int avcodec_copy_context(AVCodecContext *dest, const AVCodecContext *src) } av_opt_free(dest); +av_freep(dest-rc_override); +av_freep(dest-intra_matrix); +av_freep(dest-inter_matrix); +av_freep(dest-extradata); +av_freep(dest-subtitle_header); memcpy(dest, src, sizeof(*dest)); av_opt_copy(dest, src); @@ -205,13 +210,7 @@ int avcodec_copy_context(AVCodecContext *dest, const AVCodecContext *src) dest-slice_offset= NULL; dest-hwaccel = NULL; dest-internal= NULL; - -/* reallocate values that should be allocated separately */ -dest-extradata = NULL; -dest-intra_matrix= NULL; -dest-inter_matrix= NULL; -dest-rc_override = NULL; -dest-subtitle_header = NULL; This should stay. Updated version attched options.c | 16 +++- 1 file changed, 11 insertions(+), 5 deletions(-) 971d1769e50d2e853d7359003f8c924092a96e6b 0001-lavc-options-fix-leaks-in-avcodec_copy_context.patch From 722d7e837093212d6e6e6b17814ed408300d25a6 Mon Sep 17 00:00:00 2001 From: Lukasz Marek lukasz.m.lu...@gmail.com Date: Thu, 27 Nov 2014 00:41:16 +0100 Subject: [PATCH] lavc/options: fix leaks in avcodec_copy_context Signed-off-by: Lukasz Marek lukasz.m.lu...@gmail.com --- libavcodec/options.c | 16 +++- 1 file changed, 11 insertions(+), 5 deletions(-) breaks make fate-lavf-yuv4mpeg --- ./tests/ref/lavf/yuv4mpeg 2014-11-26 20:04:48.466295490 +0100 +++ tests/data/fate/lavf-yuv4mpeg 2014-11-27 04:16:07.234916526 +0100 @@ -1,2 +0,0 @@ -ec8178cb152f9cdbfd9cb724d977db2e *./tests/data/lavf/lavf.y4m -3801808 ./tests/data/lavf/lavf.y4m Test lavf-yuv4mpeg failed. Look at tests/data/fate/lavf-yuv4mpeg.err for details. make: *** [fate-lavf-yuv4mpeg] Error 139 I will rerun whole fate, but I assume I can push it if nothing fails? The one mentioned above is OK. ok [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB Asymptotically faster algorithms should always be preferred if you have asymptotical amounts of data signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH 1/3] avformat/hlsenc: added segment filename template
Signed-off-by: Christian Suloway csulo...@globaleagleent.com
---
libavformat/hlsenc.c | 87 +---
1 file changed, 69 insertions(+), 18 deletions(-)
diff --git a/libavformat/hlsenc.c b/libavformat/hlsenc.c
index e13f438..0a48919 100644
--- a/libavformat/hlsenc.c
+++ b/libavformat/hlsenc.c
@@ -33,7 +33,7 @@
#include internal.h
typedef struct HLSSegment {
-char filename[1024];
+char *filename;
double duration; /* in seconds */
int64_t pos;
int64_t size;
@@ -73,14 +73,23 @@ typedef struct HLSContext {
HLSSegment *segments;
HLSSegment *last_segment;
+char *dirname;
char *basename;
char *baseurl;
char *format_options_str;
AVDictionary *format_options;
+char *segment_filename;
+
AVIOContext *pb;
} HLSContext;
+static void hls_free_segment(HLSSegment *en)
+{
+av_freep(en-filename);
+av_freep(en);
+}
+
static int hls_mux_init(AVFormatContext *s)
{
HLSContext *hls = s-priv_data;
@@ -119,7 +128,9 @@ static int hls_append_segment(HLSContext *hls, double
duration, int64_t pos,
if (!en)
return AVERROR(ENOMEM);
-av_strlcpy(en-filename, av_basename(hls-avf-filename),
sizeof(en-filename));
+en-filename = av_strdup(hls-avf-filename);
+if (!en-filename)
+return AVERROR(ENOMEM);
en-duration = duration;
en-pos = pos;
@@ -136,7 +147,7 @@ static int hls_append_segment(HLSContext *hls, double
duration, int64_t pos,
if (hls-max_nb_segments hls-nb_entries = hls-max_nb_segments) {
en = hls-segments;
hls-segments = en-next;
-av_free(en);
+hls_free_segment(en);
} else
hls-nb_entries++;
@@ -152,7 +163,7 @@ static void hls_free_segments(HLSContext *hls)
while(p) {
en = p;
p = p-next;
-av_free(en);
+hls_free_segment(en);
}
}
@@ -186,6 +197,7 @@ static int hls_window(AVFormatContext *s, int last)
sequence);
for (en = hls-segments; en; en = en-next) {
+
avio_printf(hls-pb, #EXTINF:%f,\n, en-duration);
if (hls-flags HLS_SINGLE_FILE)
avio_printf(hls-pb, #EXT-X-BYTERANGE:%PRIi64@%PRIi64\n,
@@ -208,6 +220,8 @@ static int hls_start(AVFormatContext *s)
HLSContext *c = s-priv_data;
AVFormatContext *oc = c-avf;
int err = 0;
+int filename_size;
+char *filename;
if (c-flags HLS_SINGLE_FILE)
av_strlcpy(oc-filename, c-basename,
@@ -220,8 +234,18 @@ static int hls_start(AVFormatContext *s)
}
c-number++;
-if ((err = avio_open2(oc-pb, oc-filename, AVIO_FLAG_WRITE,
- s-interrupt_callback, NULL)) 0)
+filename_size = strlen(c-dirname) + strlen(oc-filename) + 1;
+filename = av_malloc(filename_size);
+if (!filename)
+return AVERROR(ENOMEM);
+*filename = '\0';
+av_strlcat(filename, c-dirname, filename_size);
+av_strlcat(filename, oc-filename, filename_size);
+
+err = avio_open2(oc-pb, filename, AVIO_FLAG_WRITE,
+ s-interrupt_callback, NULL);
+av_free(filename);
+if (err 0)
return err;
if (oc-oformat-priv_class oc-priv_data)
@@ -237,15 +261,13 @@ static int hls_write_header(AVFormatContext *s)
char *p;
const char *pattern = %d.ts;
AVDictionary *options = NULL;
-int basename_size = strlen(s-filename) + strlen(pattern) + 1;
+int basename_size;
+char *basename;
hls-sequence = hls-start_sequence;
hls-recording_time = hls-time * AV_TIME_BASE;
hls-start_pts = AV_NOPTS_VALUE;
-if (hls-flags HLS_SINGLE_FILE)
-pattern = .ts;
-
if (hls-format_options_str) {
ret = av_dict_parse_string(hls-format_options,
hls-format_options_str, =, :, 0);
if (ret 0) {
@@ -270,21 +292,45 @@ static int hls_write_header(AVFormatContext *s)
goto fail;
}
-hls-basename = av_malloc(basename_size);
-
-if (!hls-basename) {
+hls-dirname = av_strdup(s-filename);
+if (!hls-dirname) {
ret = AVERROR(ENOMEM);
goto fail;
}
-strcpy(hls-basename, s-filename);
+basename = (char *)av_basename(hls-dirname);
-p = strrchr(hls-basename, '.');
+if (hls-segment_filename) {
+hls-basename = av_strdup(av_basename(hls-segment_filename));
+if (!hls-basename) {
+ret = AVERROR(ENOMEM);
+goto fail;
+}
+if (strlen(hls-basename) != strlen(hls-segment_filename)) {
+av_log(hls, AV_LOG_ERROR, invalid segment filename %s\n,
+ hls-segment_filename);
+ret = AVERROR(EINVAL);
+goto fail;
+}
+} else {
+if (hls-flags HLS_SINGLE_FILE)
+pattern = .ts;
-if (p)
-*p = '\0';
+basename_size = strlen(basename) + strlen(pattern) + 1;
+hls-basename = av_malloc(basename_size);
+
[FFmpeg-devel] [PATCH 3/3] avformat/hlsenc: added HLS encryption
HLS encryption with key URL key containing file specified with
-hls_key_info_file (key info updates when modified). Option for IV from
segment number or random generation (-hls_random_iv).
Signed-off-by: Christian Suloway csulo...@globaleagleent.com
---
libavformat/hlsenc.c | 225 +--
1 file changed, 218 insertions(+), 7 deletions(-)
diff --git a/libavformat/hlsenc.c b/libavformat/hlsenc.c
index 11817a9..c7bd1be 100644
--- a/libavformat/hlsenc.c
+++ b/libavformat/hlsenc.c
@@ -33,16 +33,23 @@
#include libavutil/avstring.h
#include libavutil/opt.h
#include libavutil/log.h
+#include libavutil/lfg.h
+#include libavutil/random_seed.h
#include avformat.h
#include internal.h
+#define BLOCKSIZE 16
+
typedef struct HLSSegment {
char *filename;
double duration; /* in seconds */
int64_t pos;
int64_t size;
+char *key_uri;
+char *iv_string;
+
struct HLSSegment *next;
} HLSSegment;
@@ -88,12 +95,25 @@ typedef struct HLSContext {
char *segment_filename;
+char *key_info_file;
+int random_iv;
+int encrypt;
+
+char *key_file;
+char *key_uri;
+char *key_string;
+char *iv_string;
+
+AVLFG *lfg;
+
AVIOContext *pb;
} HLSContext;
static void hls_free_segment(HLSSegment *en)
{
av_freep(en-filename);
+av_freep(en-key_uri);
+av_freep(en-iv_string);
av_freep(en);
}
@@ -143,6 +163,137 @@ static int hls_delete_old_segments(HLSContext *hls) {
return 0;
}
+static int hls_encryption_init(AVFormatContext *s)
+{
+HLSContext *hls = s-priv_data;
+
+hls-key_file = NULL;
+hls-key_uri = NULL;
+hls-key_string = NULL;
+hls-iv_string = NULL;
+
+if (hls-key_info_file) {
+hls-encrypt = 1;
+} else {
+hls-encrypt = 0;
+return 0;
+}
+
+if (hls-random_iv) {
+hls-lfg = av_malloc(sizeof(AVLFG));
+av_lfg_init(hls-lfg, av_get_random_seed());
+} else
+hls-lfg = NULL;
+
+return 0;
+}
+
+static int hls_encryption_start(HLSContext *hls) {
+
+uint8_t *iv = NULL;
+int ret, i, j, rotate_iv = 0;
+unsigned int u;
+AVIOContext *pb = NULL;
+AVIOContext *dyn_buf = NULL;
+uint8_t buf[1024], *tmp;
+char *p, *tstr, *saveptr = NULL;
+char key[BLOCKSIZE], *key_string;
+
+if ((ret = avio_open(pb, hls-key_info_file, AVIO_FLAG_READ)) 0) {
+av_log(hls, AV_LOG_ERROR, error opening key info file %s\n,
+ hls-key_info_file);
+return ret;
+}
+ret = avio_open_dyn_buf(dyn_buf);
+if (ret 0) {
+avio_closep(pb);
+return ret;
+}
+
+while ((ret = avio_read(pb, buf, sizeof(buf))) 0)
+avio_write(dyn_buf, buf, ret);
+avio_w8(dyn_buf, 0);
+avio_closep(pb);
+
+if ((ret = avio_close_dyn_buf(dyn_buf, tmp)) 0)
+return ret;
+
+p = tmp;
+if (!(tstr = av_strtok(p, \n, saveptr)) || !*tstr) {
+av_log(hls, AV_LOG_ERROR, no key URL specified in key info file %s\n,
+ hls-key_info_file);
+return AVERROR(EINVAL);
+}
+p = NULL;
+av_free(hls-key_uri);
+hls-key_uri = av_strdup(tstr);
+if (!hls-key_uri)
+return AVERROR(ENOMEM);
+if (!(tstr = av_strtok(p, \n, saveptr)) || !*tstr) {
+av_log(hls, AV_LOG_ERROR, no key file specified in key info file
%s\n,
+ hls-key_info_file);
+return AVERROR(EINVAL);
+}
+av_free(hls-key_file);
+hls-key_file = av_strdup(tstr);
+if (!hls-key_file)
+return AVERROR(ENOMEM);
+av_free(tmp);
+
+if ((ret = avio_open(pb, hls-key_file, AVIO_FLAG_READ)) 0) {
+av_log(hls, AV_LOG_ERROR, error opening key file %s\n,
+ hls-key_file);
+return ret;
+}
+
+ret = avio_read(pb, key, sizeof(key));
+avio_closep(pb);
+if (ret sizeof(key)) {
+av_log(hls, AV_LOG_ERROR, error reading key file %s\n,
+ hls-key_file);
+return ret;
+}
+
+key_string = av_mallocz(BLOCKSIZE*2 + 1);
+if (!key_string)
+return AVERROR(ENOMEM);
+ff_data_to_hex(key_string, key, BLOCKSIZE, 0);
+if (!hls-key_string || strncmp(key_string, hls-key_string, BLOCKSIZE*2))
{
+av_free(hls-key_string);
+hls-key_string = key_string;
+rotate_iv = 1;
+} else {
+av_free(key_string);
+}
+
+if (!hls-random_iv) {
+iv = av_mallocz(BLOCKSIZE);
+if (!iv)
+return AVERROR(ENOMEM);
+for (i = 0; i 8; i++)
+iv[BLOCKSIZE - 1 - i ] = (hls-sequence i*8) 0xff;
+} else if (!hls-iv_string || rotate_iv) {
+iv = av_malloc(BLOCKSIZE);
+if (!iv)
+return AVERROR(ENOMEM);
+for (i = 0; i BLOCKSIZE/4; i++) {
+u = av_lfg_get(hls-lfg);
+for (j = 0; j 4; j++)
+iv[i*4 + j] =
[FFmpeg-devel] [PATCH 2/3] avformat/hlsenc: added segment file deletion
Option removes segments no longer in playlist when older than playlist +
segment duration
Signed-off-by: Christian Suloway csulo...@globaleagleent.com
---
libavformat/hlsenc.c | 72
1 file changed, 67 insertions(+), 5 deletions(-)
diff --git a/libavformat/hlsenc.c b/libavformat/hlsenc.c
index 0a48919..11817a9 100644
--- a/libavformat/hlsenc.c
+++ b/libavformat/hlsenc.c
@@ -19,8 +19,13 @@
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
+#include config.h
#include float.h
#include stdint.h
+#if HAVE_UNISTD_H
+#include unistd.h
+#endif
+#include stdio.h
#include libavutil/avassert.h
#include libavutil/mathematics.h
@@ -58,6 +63,7 @@ typedef struct HLSContext {
float time;// Set by a private option.
int max_nb_segments; // Set by a private option.
int wrap; // Set by a private option.
+int delete_segments;
uint32_t flags;// enum HLSFlags
int allowcache;
@@ -72,6 +78,7 @@ typedef struct HLSContext {
HLSSegment *segments;
HLSSegment *last_segment;
+HLSSegment *old_segments;
char *dirname;
char *basename;
@@ -90,6 +97,52 @@ static void hls_free_segment(HLSSegment *en)
av_freep(en);
}
+static int hls_delete_old_segments(HLSContext *hls) {
+
+HLSSegment *segment, *previous_segment = NULL;
+float playlist_duration = 0.0f;
+int path_size;
+char *path;
+
+segment = hls-segments;
+while (segment) {
+playlist_duration += segment-duration;
+segment = segment-next;
+}
+
+segment = hls-old_segments;
+while (segment) {
+playlist_duration -= segment-duration;
+previous_segment = segment;
+segment = segment-next;
+if (playlist_duration = 0.0f) {
+previous_segment-next = NULL;
+break;
+}
+}
+
+while (segment) {
+av_log(hls, AV_LOG_DEBUG, deleting old segment %s\n,
+ segment-filename);
+path_size = strlen(hls-dirname) + strlen(segment-filename) + 1;
+path = av_malloc(path_size);
+if (!path)
+return AVERROR(ENOMEM);
+av_strlcpy(path, hls-dirname, path_size);
+av_strlcat(path, segment-filename, path_size);
+if (unlink(path) 0) {
+av_log(hls, AV_LOG_ERROR, failed to delete old segment %s: %s\n,
+ path, strerror(errno));
+}
+av_free(path);
+previous_segment = segment;
+segment = segment-next;
+hls_free_segment(previous_segment);
+}
+
+return 0;
+}
+
static int hls_mux_init(AVFormatContext *s)
{
HLSContext *hls = s-priv_data;
@@ -124,6 +177,7 @@ static int hls_append_segment(HLSContext *hls, double
duration, int64_t pos,
int64_t size)
{
HLSSegment *en = av_malloc(sizeof(*en));
+int ret;
if (!en)
return AVERROR(ENOMEM);
@@ -147,7 +201,14 @@ static int hls_append_segment(HLSContext *hls, double
duration, int64_t pos,
if (hls-max_nb_segments hls-nb_entries = hls-max_nb_segments) {
en = hls-segments;
hls-segments = en-next;
-hls_free_segment(en);
+
+if (en hls-delete_segments !hls-wrap) {
+en-next = hls-old_segments;
+hls-old_segments = en;
+if ((ret = hls_delete_old_segments(hls)) 0)
+return ret;
+} else
+hls_free_segment(en);
} else
hls-nb_entries++;
@@ -156,9 +217,9 @@ static int hls_append_segment(HLSContext *hls, double
duration, int64_t pos,
return 0;
}
-static void hls_free_segments(HLSContext *hls)
+static void hls_free_segments(HLSSegment *p)
{
-HLSSegment *p = hls-segments, *en;
+HLSSegment *en;
while(p) {
en = p;
@@ -197,7 +258,6 @@ static int hls_window(AVFormatContext *s, int last)
sequence);
for (en = hls-segments; en; en = en-next) {
-
avio_printf(hls-pb, #EXTINF:%f,\n, en-duration);
if (hls-flags HLS_SINGLE_FILE)
avio_printf(hls-pb, #EXT-X-BYTERANGE:%PRIi64@%PRIi64\n,
@@ -442,7 +502,8 @@ static int hls_write_trailer(struct AVFormatContext *s)
hls-avf = NULL;
hls_window(s, 1);
-hls_free_segments(hls);
+hls_free_segments(hls-segments);
+hls_free_segments(hls-old_segments);
av_free(hls-dirname);
avio_close(hls-pb);
@@ -458,6 +519,7 @@ static const AVOption options[] = {
{hls_list_size, set maximum number of playlist entries,
OFFSET(max_nb_segments),AV_OPT_TYPE_INT,{.i64 = 5}, 0, INT_MAX, E},
{hls_ts_options,set hls mpegts list of options for the container format
used for hls, OFFSET(format_options_str), AV_OPT_TYPE_STRING, {.str = NULL},
0, 0,E},
{hls_wrap, set number after which the index wraps, OFFSET(wrap),
AV_OPT_TYPE_INT,{.i64 = 0}, 0,
Re: [FFmpeg-devel] [PATCH] lavc/options: don't copy priv context when it is not AVOption object
On Thu, Dec 04, 2014 at 12:50:06AM +0100, Lukasz Marek wrote: On 03.12.2014 01:28, Michael Niedermayer wrote: On Wed, Dec 03, 2014 at 01:24:55AM +0100, Lukasz Marek wrote: On 03.12.2014 01:15, Michael Niedermayer wrote: On Wed, Dec 03, 2014 at 12:06:47AM +0100, Lukasz Marek wrote: This prevents potential crash when opt API is used without a class. Signed-off-by: Lukasz Marek lukasz.m.lu...@gmail.com --- libavcodec/options.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavcodec/options.c b/libavcodec/options.c index 7f9fb07..42ef754 100644 --- a/libavcodec/options.c +++ b/libavcodec/options.c @@ -196,7 +196,7 @@ int avcodec_copy_context(AVCodecContext *dest, const AVCodecContext *src) dest-priv_data = orig_priv_data; -if (orig_priv_data) +if (orig_priv_data src-codec src-codec-priv_class) av_opt_copy(orig_priv_data, src-priv_data); src-codec-priv_class --- LGTM the src-codec check is probably ok too but is it NULL ever ? after avcodec_alloc_context3(NULL)? IIRC there was a patch recently that set codec in avcodec_alloc_context3 (via avcodec_get_context_defaults3) when provided. Maybe after that it is not so common case. But still when allocated with NULL then you have NULL there. src context doesn't have to be open neither. ok I'm sorry for doing a mess recently regarding this damn avcodec_copy_context function. I'm sending updated patch. I will send in a moment patchset with test for it. The test doesn't cover everything (it is kinda hard to test it), but it should cover options management. Just to clear, I will ping in one more in a moment and don't bother with any other atm :) options.c |5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) 2667b09ea91f32396f7151900b5fa24a8bcabfef 0001-lavc-options-don-t-copy-priv-context-when-it-is-not-.patch From d8884452f3bd0af45e459c7c2b2549b7a225ed4d Mon Sep 17 00:00:00 2001 From: Lukasz Marek lukasz.m.lu...@gmail.com Date: Tue, 2 Dec 2014 23:24:15 +0100 Subject: [PATCH] lavc/options: don't copy priv context when it is not AVOption object This prevents potential crash when opt API is used without a class. LGTM [...] -- Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB Freedom in capitalist society always remains about the same as it was in ancient Greek republics: Freedom for slave owners. -- Vladimir Lenin signature.asc Description: Digital signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
[FFmpeg-devel] [PATCH] Fixed PNG decoding with Paeth filter for RGBA 64-bits.
Found and fixed an artifact in the last column of decoded RGBA 64-bits PNG images. The code was dealing with a SIMD-optimized version of the function without taking into account that we can have RGB/RGBA images that are respectively 6 or 8 bytes per pixel (not just 3 or 4). Dominique 0001-Fixed-PNG-decoding-with-Paeth-filter-for-RGBA-64-bit.patch Description: 0001-Fixed-PNG-decoding-with-Paeth-filter-for-RGBA-64-bit.patch ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 1/3] avformat/hlsenc: added segment filename template
On Thu, Dec 04, 2014 at 05:05:24PM -0600, Christian Suloway wrote:
Signed-off-by: Christian Suloway csulo...@globaleagleent.com
---
libavformat/hlsenc.c | 87
+---
1 file changed, 69 insertions(+), 18 deletions(-)
missing documentation
diff --git a/libavformat/hlsenc.c b/libavformat/hlsenc.c
index e13f438..0a48919 100644
--- a/libavformat/hlsenc.c
+++ b/libavformat/hlsenc.c
@@ -33,7 +33,7 @@
#include internal.h
typedef struct HLSSegment {
-char filename[1024];
+char *filename;
this looks unrelated
double duration; /* in seconds */
int64_t pos;
int64_t size;
@@ -73,14 +73,23 @@ typedef struct HLSContext {
HLSSegment *segments;
HLSSegment *last_segment;
+char *dirname;
char *basename;
char *baseurl;
char *format_options_str;
why is the dirname needed for allowing the user to set the template ?
[...]
@@ -186,6 +197,7 @@ static int hls_window(AVFormatContext *s, int last)
sequence);
for (en = hls-segments; en; en = en-next) {
+
avio_printf(hls-pb, #EXTINF:%f,\n, en-duration);
if (hls-flags HLS_SINGLE_FILE)
avio_printf(hls-pb, #EXT-X-BYTERANGE:%PRIi64@%PRIi64\n,
stray change
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Those who are too smart to engage in politics are punished by being
governed by those who are dumber. -- Plato
signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Re: [FFmpeg-devel] [PATCH 3/3] lavc/libxvid: workaround for bug in libxvidcore
On Thu, Dec 04, 2014 at 11:33:03PM +0100, Lukasz Marek wrote:
On 04.12.2014 12:42, Clément Bœsch wrote:
On Thu, Dec 04, 2014 at 12:39:29AM +0100, Lukasz Marek wrote:
On 24.11.2014 01:17, Lukasz Marek wrote:
libxvidcore calculate number of threads basing on video height.
If height is small enough it allocates 0 bytes long memory and
writes to it.
Setting thread_count to 0 uses 1 thread and skips bugged code.
Signed-off-by: Lukasz Marek lukasz.m.lu...@gmail.com
---
libavcodec/libxvid.c | 11 +++
1 file changed, 11 insertions(+)
diff --git a/libavcodec/libxvid.c b/libavcodec/libxvid.c
index 70d52b9..efe7b8a 100644
--- a/libavcodec/libxvid.c
+++ b/libavcodec/libxvid.c
@@ -471,6 +471,17 @@ static av_cold int xvid_encode_init(AVCodecContext
*avctx)
xvid_enc_create.num_zones = 0;
xvid_enc_create.num_threads = avctx-thread_count;
+/* workaround for a bug in libxvidcore */
+if (avctx-height 18) {
+if (avctx-thread_count 2) {
+xvid_enc_create.num_threads = 0;
+} else {
+av_log(avctx, AV_LOG_ERROR,
+ Too small height for threads 1.);
+ret = AVERROR(EINVAL);
+goto fail;
+}
+}
xvid_enc_create.plugins = plugins;
xvid_enc_create.num_plugins = 0;
Please comment this patch. xvid accepted fix for it.
Can you add a version check, or at least a comment saying what versions
are affected by this?
Added a check. it was introduced in 1.3.0 and still present in
1.3.3. Next should be fixed.
libxvid.c | 13 +
1 file changed, 13 insertions(+)
de82e17a2e509e21ce8384d913a9394f80b8537d
0001-lavc-libxvid-workaround-for-bug-in-libxvidcore.patch
From 1bfe00284fdb3dba072147df1efce03c7d29dd35 Mon Sep 17 00:00:00 2001
From: Lukasz Marek lukasz.m.lu...@gmail.com
Date: Mon, 24 Nov 2014 01:12:06 +0100
Subject: [PATCH] lavc/libxvid: workaround for bug in libxvidcore
libxvidcore calculate number of threads basing on video height.
If height is small enough it allocates 0 bytes long memory and
writes to it.
Setting thread_count to 0 uses 1 thread and skips bugged code.
Signed-off-by: Lukasz Marek lukasz.m.lu...@gmail.com
applied
thanks
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Observe your enemies, for they first find out your faults. -- Antisthenes
signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
