We need at least one pixel around the logo to use as known points to
interpolate from. So properly declare the band/t attribute has having
a minimum value of 1.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Reviewed-by: Stefano Sabatini <stefasab@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Options "show" and "band" are unrelated and should thus be
independent. However, setting "show" to 1 currently resets "band" to
its default value of 4. While this is documented, this still
surprising and confusing IMHO.
Change this behavior and make "show" and "band" independent from each
other. Update the documentation accordingly.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Reviewed-by: Stefano Sabatini <stefasab@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
The left and right samples are the same for the whole line, so store
their values and don't recompute them for every iteration of "y".
This simple optimization results in a speed improvement between 15%
and 20% in my tests (depending on the logo geometry.)
Result is obviously the same.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Reviewed-by: Stefano Sabatini <stefasab@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
The top left hand corner pixel coordinates are already stored in
logo_x1 and logo_y1 so don't recompute each of them 6 times for every
iteration.
This is a simple code optimization, result is obviously the same. The
performance gain is small (about 2% in my tests) but still good to
have, and the new code is clearer.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Reviewed-by; Stefano Sabatini <stefasab@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
When operating on subsampled chroma planes, some rounding is taking
place. The left and top borders are rounded down while the width and
height are rounded up, so all rounding is done outward to guarantee the
logo area is fully covered.
The problem is that the width and height are counted from the
unrounded left and top borders, respectively. So if the left or top
border position has indeed been rounded down, and the width or height
needs no rounding (up), the position of the the right or bottom border
will be effectively rounded down, i.e. inward.
The issue can easily be seen with a yuv240p input and
-vf delogo=45:45:60:40:show=1 -vframes 1 delogo-bug.png
(or virtually any logo area with odd x and y and even width and
height.) The right and bottom chroma borders (in green) are clearly
off.
In order to fix this, the width and height must be adjusted to include
the bits lost in the rounding of the left and top border positions,
respectively, prior to being themselves rounded up.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
When interpolating, weights are based on relative distances, which
assume square pixels. If a non-1:1 sample aspect ratio is used, it
should be taken into account when comparing distances, because the
human eye and brain care about the picture as it is displayed, not
stored.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Add function avfilter_graph_parse_ptr() and favor it in place of
avfilter_graph_parse(), which will be restored with the old/Libav
signature at the next bump.
If HAVE_INCOMPATIBLE_LIBAV_API is enabled it will use the
Libav-compatible signature for avfilter_graph_parse().
At the next major bump the current implementation of
avfilter_graph_parse() should be dropped in favor of the Libav/old
implementation.
Should address trac ticket #2672.
The original delogo algorithm interpolates both horizontally and
vertically and uses the average to compute the resulting sample. This
works reasonably well when the logo area is almost square. However
when the logo area is significantly larger than high or higher than
large, the result is largely suboptimal.
The issue can be clearly seen by testing the delogo filter with a fake
logo area that is 200 pixels large and 2 pixels high. Vertical
interpolation gives a very good result in that case, horizontal
interpolation gives a very bad result, and the overall result is poor,
because both are given the same weight.
Even when the logo is roughly square, the current algorithm gives poor
results on the borders of the logo area, because it always gives
horizontal and vertical interpolations an equal weight, and this is
suboptimal on borders. For example, in the middle of the left hand
side border of the logo, you want to trust the left known point much
more than the right known point (which the current algorithm already
does) but also much more than the top and bottom known points (which
the current algorithm doesn't do.)
By properly weighting each known point when computing the value of
each interpolated pixel, the visual result is much better, especially
on borders and/or for high or large logo areas.
The algorithm I implemented guarantees that the weight of each of the
4 known points directly depends on its distance to the interpolated
point. It is largely inspired from the original algorithm, the key
difference being that it computes the relative weights globally
instead of separating the vertical and horizontal interpolations and
combining them afterward.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Stefano Sabatini <stefasab@gmail.com>
* 'drawbox_exprs' of https://github.com/mjmvisser/FFmpeg:
enabled expressions on x, y, w, h and t parameters for drawgrid and drawbox, added examples
Reviewed-by: Andrey Utkin
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'eeeb5c291d3f78eaade5b99c2614c7cab0e9be79':
vsrc_movie: do not free avoption variables in uninit()
Conflicts:
libavfilter/src_movie.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
The algorithm works on src and writes to dst, not the other way
around.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Makes it easier to understand that there is no difference in init
callback for movie and amovie. Also saves a few lines of code.
Signed-off-by: Alexander Strasser <eclipse7@gmx.net>
The drawbox video filter is drawing lines one pixel thinner than
requested. The default thickness is 4 pixel but in fact the lines
drawn by default are only 3 pixel wide.
Change the comparisons in the code to fix this off-by-one bug. Also
change the default thickness from 4 to 3 to minimize the unexpected
changes from the user's perspective.
As I was already touching these lines, I also removed the "maximum" in
the thickness parameter description, as I don't think it was adding
any value and I even found it confusing.
Reviewed-by: Andrey Utkin <andrey.krieger.utkin@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Based on the libavfilter SOC filter by Vitor Sessak, with the following additions:
* integer arithmetic
* bilinear interpolation
* RGB path
* configurable parametric angle, output width and height
Address trac issue #1500.
See thread:
Subject: [FFmpeg-devel] [WIP] rotate filter(s)
Date: 2010-10-03 17:35:49 GMT
w/h from input frame are not copied in av_frame_copy_props(). This
commit avoids a mismatch between aligned_[wh] and outbuf->{width,height}
(and thus avoids triggering an assert in avfilter because of this).
Do not read padding or out-of-buffer values when computing the output
value for a pixel close to the image buffer edge.
This avoids non visible artifacts which affected the output checksum.
See thread:
Subject: [FFmpeg-devel] [PATCH] lavfi/mp/mcdeint: avoid uninited data read
Date: Thu, 30 May 2013 18:57:14 +0200
Make code slightly faster, simpler, clearer.
The filter is still slow as hell, and that change won't cause any
visible performance improvement (it still takes more than one minute to
process a single 1080p frame on a Core 2 here).
Call NEXT() only once since it got slower due to the overflow condition
introduced in 91b46145.
interp_trilinear: 1462 → 1280 decicycles
interp_tetrahedral: 1188 → 1097 decicycles
Tested on a Core2, GCC 4.8.
This more evenly distributes the load between threads
This also fixes the chroma filtering where the filter was applied twice
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Fixes crash on SSE2
Fixes Ticket2443
Note, its possible to restructure the code to avoid the 16pixel limitation,
and such change is welcome!
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'cebdedca57d95834a8f8098c7b6a322a1163e26b':
prores: decode alpha plane when it's present
avfilter: check a malloc in avfilter_link().
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'd79bd6040599ef017dc45d5997fcc959ab1f71f6':
lavfi: factorize freeing a link
Conflicts:
libavfilter/avfilter.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '1dd35611433c224e2231ed7ae26b5b560863a3cb':
vf_format: do not use the AVFilterFormats internals.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '16a645adeb758207346a4bbf66766f02734c461e':
vf_pixdesctest: make config_props work properly when called multiple times.
vf_hqdn3d: make config_props work properly when called multiple times.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '3fb29588a27a711132106b924e27b53789a58dcb':
vf_drawtext: don't leak the expressions.
vf_crop: make config_props work properly when called multiple times.
vf_setdar: make config_props work properly when called multiple times.
Conflicts:
libavfilter/vf_aspect.c
libavfilter/vf_drawtext.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '4c205f42c86ccefa093c59434669af34ad14a52b':
vf_drawbox: make config_props work properly when called multiple times.
vf_drawtext: do not reset the frame number in config_input.
vf_fps: move initializing pts from config_props to init.
Conflicts:
libavfilter/vf_drawbox.c
libavfilter/vf_drawtext.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'd371c3c2e2830d9783465ecfe1ab7d93351083b7':
vf_frei0r: make config_props work properly when called multiple times.
vf_gradfun: make config_props work properly when called multiple times.
vf_lut: make config_props work properly when called multiple times.
Conflicts:
libavfilter/vf_lut.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
vf_pad: use the name 's' for the pointer to the private context
vf_overlay: use the name 's' for the pointer to the private context
vf_lut: use the name 's' for the pointer to the private context
Conflicts:
libavfilter/vf_lut.c
libavfilter/vf_overlay.c
libavfilter/vf_pad.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '4753f802c00853859b7b4b8fdb79c35e082cb7f8':
vf_libopencv: use the name 's' for the pointer to the private context
vf_hqdn3d: use the name 's' for the pointer to the private context
vf_hflip: use the name 's' for the pointer to the private context
vf_gradfun: use the name 's' for the pointer to the private context
Conflicts:
libavfilter/vf_gradfun.c
libavfilter/vf_hflip.c
libavfilter/vf_hqdn3d.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'f6b6d6ac42c0ee427fcd3fae4c398b2103b491b8':
vf_frei0r: use the name 's' for the pointer to the private context
vf_format: use the name 's' for the pointer to the private context
vf_fieldorder: use the name 's' for the pointer to the private context
vf_fieldorder: cosmetics, remove an extra linebreak
Conflicts:
libavfilter/vf_frei0r.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'c0279956b3ca3e5fd0a6a25253893d6f07000e68':
vf_fade: use the name 's' for the pointer to the private context
vf_drawtext: use the name 's' for the pointer to the private context
vf_drawbox: use the name 's' for the pointer to the private context
Conflicts:
libavfilter/vf_drawbox.c
libavfilter/vf_drawtext.c
libavfilter/vf_fade.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'ba09675f44612fad9f7169f71b8276beb50a0dcd':
vf_delogo: use the name 's' for the pointer to the private context
vf_cropdetect: use the name 's' for the pointer to the private context
vf_crop: cosmetics, break lines
Conflicts:
libavfilter/vf_delogo.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '7f83959598b6565baa0091e5739dd9091ab7a990':
vf_crop: use the name 's' for the pointer to the private context
vf_boxblur: use the name 's' for the pointer to the private context
vf_blackframe: use the name 's' for the pointer to the private context
Conflicts:
libavfilter/vf_blackframe.c
libavfilter/vf_boxblur.c
libavfilter/vf_crop.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'b3ea76624ad1baab0b6bcc13f3f856be2f958110':
vf_aspect: use the name 's' for the pointer to the private context
Remove commented-out debug #define cruft
Conflicts:
libavcodec/4xm.c
libavcodec/dvdsubdec.c
libavcodec/ituh263dec.c
libavcodec/mpeg12.c
libavfilter/avfilter.c
libavfilter/vf_aspect.c
libavfilter/vf_fieldorder.c
libavformat/rtmpproto.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Always use the special filter for the first and last 3 columns (only).
Changes made in 64ed397 slowed the filter to just under 3/4 of what it
was. This commit restores the speed while maintaining identical output.
For reference, on my Athlon64:
1733222 decicycles in old
2358563 decicycles in new
1727558 decicycles in this
Signed-off-by: Anton Khirnov <anton@khirnov.net>
With the introduction of AVFilterContext->is_disabled, we can simplify
the custom passthrough mode in filters.
This commit is technically a small compat break, but the timeline was
introduced very recently.
Doxy by Stefano Sabatini.
This should fix several issues with odd dimensions inputs.
lut, vflip, pad and crop video filters also need to be checked for such
issues. It's possible sws is also affected.
query_formats() returning EAGAIN is not considered a progress
in the format negotiation.
If the filter returns EAGAIN but did set some of its formats
lists, it could be considered a partial success and counted
as progress in the negotiation. Not counting it is not a
problem because it currently only happens in the first round,
where there will always be some progress on the other filters.
* commit '20c86571ccc71412781d4a4813e4693e0c42aec6':
lavfi: let gcc realign the stack on public graph driving functions
Conflicts:
libavfilter/buffersink.c
libavfilter/buffersrc.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
The functions which actually drive the filter graph by pushing
frames through it need to ensure an aligned stack for SIMD functions.
This fixes a crash in YADIF filter when using a mingw build in a MSVC
application.
Signed-off-by: Martin Storsjö <martin@martin.st>
Note, the design is still flawed, dont expect this to be frame
accurate. float/double first needs to be removed and frames
itself trimmed instead of drop vs nodrop
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* commit 'c22263d3e813d442df8fa5f5ba8993573fe775d8':
graphparser: only print filter arguments if they are non-NULL
af_channelmap: ensure the output channel layout is valid.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
From various tests, there is no evidence that the filter is faster than
the native tinterlace, and it has several more limitations and a few
bugs.
See thread:
Subject: [FFmpeg-devel] [NOPATCH] lavfi/mp: drop tinterlace wrapper
Date: Wed, 5 Sep 2012 00:26:01 +0200
