Anton Khirnov [Tue, 29 Apr 2014 06:00:03 +0000 (08:00 +0200)]
lavc: move CODEC_FLAG_NORMALIZE_AQP to mpegvideo
Anton Khirnov [Tue, 29 Apr 2014 05:50:16 +0000 (07:50 +0200)]
lavc: deprecate CODEC_FLAG_INPUT_PRESERVED
The addition of reference-counted frames makes it pointless.
Anton Khirnov [Sun, 27 Apr 2014 11:40:11 +0000 (13:40 +0200)]
matroska: add the Opus mapping
Anton Khirnov [Tue, 29 Apr 2014 10:26:38 +0000 (12:26 +0200)]
matroskaenc: write CodecDelay
Alessandro Ghedini [Tue, 29 Apr 2014 16:53:16 +0000 (18:53 +0200)]
lavfi: add Bauer stereo-to-binaural audio filter
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Anton Khirnov [Tue, 29 Apr 2014 10:03:13 +0000 (12:03 +0200)]
matroskadec: read the CodecDelay element
Luca Barbato [Sat, 26 Apr 2014 11:32:03 +0000 (13:32 +0200)]
librtmp: Map native options to librtmp ones when possible
Makes the two protocols nearly seamless.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Luca Barbato [Tue, 29 Apr 2014 16:20:17 +0000 (18:20 +0200)]
hls: Sync the file number with the start sequence
Luca Barbato [Tue, 29 Apr 2014 16:37:01 +0000 (18:37 +0200)]
hls: Factor the sequence computation
Luca Barbato [Tue, 29 Apr 2014 08:13:34 +0000 (10:13 +0200)]
hls: Print start_number as first sequence value
The option now behaves as expected.
Luca Barbato [Tue, 29 Apr 2014 08:28:11 +0000 (10:28 +0200)]
segment: Report the current media sequence
Useful for debugging mostly.
Luca Barbato [Tue, 29 Apr 2014 08:07:03 +0000 (10:07 +0200)]
hls: Report the current media sequence
Useful for debugging mostly.
Hendrik Leppkes [Mon, 28 Apr 2014 21:38:05 +0000 (23:38 +0200)]
avconv_dxva2: define all used GUIDs directly instead of relying on the dxva2api.h header
This avoids issues with inconsistent headers with different versions of mingw-w64 and older MSVC variants.
Fixes FATE builds of MSVC 2010.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Anton Khirnov [Tue, 29 Apr 2014 09:13:43 +0000 (11:13 +0200)]
h264: drop commented out cruft
Anton Khirnov [Thu, 5 Dec 2013 20:17:00 +0000 (21:17 +0100)]
mpegvideo: remove unused MpegEncContext.b4_stride
Anton Khirnov [Thu, 5 Dec 2013 20:14:40 +0000 (21:14 +0100)]
mpegvideo: move the MpegEncContext fields used from arm asm to the beginning
This should reduce the frequency with which the offsets need to be
updated.
Luca Barbato [Tue, 18 Mar 2014 20:19:00 +0000 (21:19 +0100)]
hls: Add an option to prepend a baseurl to the playlist entries
Useful to generate playlists with absolute paths.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Enrique Arizón Benito [Tue, 18 Mar 2014 20:18:59 +0000 (21:18 +0100)]
segment: Add an option to prepend a string to the list entries
Useful to generate lists with absolute urls.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Anton Khirnov [Mon, 28 Apr 2014 21:00:38 +0000 (23:00 +0200)]
configure: rework dxva in avconv handling
Move the dxva check after the dependencies have been detected, make sure
the ole32 library exists.
Fixes build on non-Windows.
Hendrik Leppkes [Tue, 22 Apr 2014 13:22:53 +0000 (15:22 +0200)]
avconv: add support for DXVA2 decoding
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Michael Niedermayer [Tue, 22 Apr 2014 13:22:52 +0000 (15:22 +0200)]
hwaccel: fix dxva2 & vaapi loop filter parameters
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Hendrik Leppkes [Tue, 22 Apr 2014 13:22:51 +0000 (15:22 +0200)]
dxva2_h264: add a workaround for old Intel GPUs
Old Intel GPUs expect the reference frame index to the actual surface,
instead of the index into RefFrameList as specified by the spec.
This workaround should be set when using one of the "ClearVideo" decoder
devices.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Hendrik Leppkes [Tue, 22 Apr 2014 13:22:50 +0000 (15:22 +0200)]
dxva2_h264: set the correct ref frame index in the long slice struct
The latest H.264 DXVA specification states that the index in this
structure should refer to a valid entry in the RefFrameList of the picture
parameter structure, and not to the actual surface index.
Fixes H.264 DXVA2 decoding on recent Intel GPUs (tested on Sandy and Ivy)
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Anton Khirnov [Fri, 25 Apr 2014 14:33:27 +0000 (16:33 +0200)]
lavc: eliminate tb_unreliable()
If framerate estimation is enabled, it makes sense to always apply it,
instead of limiting it to a few specific cases.
Alessandro Ghedini [Mon, 21 Apr 2014 17:10:52 +0000 (19:10 +0200)]
vc1: Do not return an error when skipping b frames
This caused mpv (and possibly others) to fallback to software decoding after
seeking a VC1 stream.
Bug-Id: 667
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Miles Gould [Mon, 21 Apr 2014 13:46:22 +0000 (14:46 +0100)]
mov: Emit the correct tags for clcp tracks
Bug-Id: 664
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Aidan Skinner [Mon, 21 Apr 2014 13:47:38 +0000 (14:47 +0100)]
mov: Write prof section of tapt tag
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Derek Buitenhuis [Wed, 23 Apr 2014 16:15:13 +0000 (12:15 -0400)]
fate: Add fic-in-avi test
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Janne Grunau [Mon, 21 Apr 2014 14:13:28 +0000 (16:13 +0200)]
aarch64: NEON optimized FIR audio resampling
Optimized for the default filter length 16.
30% faster opus silk decoding.
Janne Grunau [Mon, 21 Apr 2014 14:12:21 +0000 (16:12 +0200)]
lavr: define ResampleContext in resample.h
Required for arch optimized resampling.
Janne Grunau [Wed, 23 Apr 2014 10:19:59 +0000 (12:19 +0200)]
lavu: add CHK_OFFS as AV_CHECK_OFFSET to check struct member offsets
Luca Barbato [Wed, 23 Apr 2014 19:20:17 +0000 (21:20 +0200)]
flv: Do not mangle dts values for negative cts
Some applications really mean to send negative pts.
Luca Barbato [Wed, 23 Apr 2014 19:19:27 +0000 (21:19 +0200)]
flv: Warn only once
No point in sending the message multiple time.
Luca Barbato [Wed, 23 Apr 2014 19:04:20 +0000 (21:04 +0200)]
flv: Improve log messages
Messages should start with a capital letter and possibly end with a "."
if they are statements.
Kostya Shishkov [Wed, 23 Apr 2014 18:02:37 +0000 (20:02 +0200)]
on2avc: change a comment at #endif to match actual define
Kostya Shishkov [Sun, 9 Jun 2013 18:03:33 +0000 (20:03 +0200)]
On2 AVC decoder
Derek Buitenhuis [Wed, 23 Apr 2014 16:53:00 +0000 (17:53 +0100)]
fic: Make warning message more accurate
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Derek Buitenhuis [Wed, 23 Apr 2014 16:18:36 +0000 (12:18 -0400)]
fic: Remove redundant clips
The equations can't overflow or underflow anyway.
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Derek Buitenhuis [Tue, 22 Apr 2014 20:42:11 +0000 (16:42 -0400)]
fic: Simplify alpha blending
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Michael Niedermayer [Tue, 22 Apr 2014 20:44:14 +0000 (22:44 +0200)]
avutil: move av_get_time_base_q() after include rational.h
Fix compilation
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Derek Buitenhuis [Tue, 22 Apr 2014 05:26:07 +0000 (07:26 +0200)]
avutil: Add av_get_time_base_q()
This fixes usage of AV_TIME_BASE_Q in C++ applications, which
cannot use compound literals directly in their code.
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Derek Buitenhuis [Sat, 19 Apr 2014 22:46:52 +0000 (18:46 -0400)]
fic: Support rendering cursors
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Derek Buitenhuis [Sat, 19 Apr 2014 19:40:35 +0000 (15:40 -0400)]
fic: Use proper quantization matrix index
The matrices are not zigzagged.
Fixes artefacting.
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Janne Grunau [Sun, 20 Apr 2014 15:57:36 +0000 (17:57 +0200)]
aarch64: NEON vorbis_inverse_coupling
From the ARMv7 NEON version. 16 times faster as the C version, overall
more than 12% faster vorbis decoding on Apple's A7.
Janne Grunau [Sat, 19 Apr 2014 16:17:23 +0000 (18:17 +0200)]
aarch64: NEON fixed/floating point MPADSP apply_window
30%/25% (fixed/float) faster mp3 decoding on Apple's A7. The floating
point decoder is approximately 7% faster.
Janne Grunau [Wed, 16 Apr 2014 21:47:32 +0000 (23:47 +0200)]
aarch64: NEON float to s16 audio conversion
Janne Grunau [Tue, 15 Apr 2014 16:35:57 +0000 (18:35 +0200)]
aarch64: NEON float (i)MDCT
Approximately as fast as the ARM NEON version on Apple's A7.
Janne Grunau [Wed, 26 Mar 2014 14:20:42 +0000 (15:20 +0100)]
aarch64: NEON float FFT
Approximately as fast as the ARM NEON version on Apple's A7.
Anton Khirnov [Sun, 13 Apr 2014 05:05:27 +0000 (07:05 +0200)]
lavf: do not use the parser duration for video
The parser has no way of knowing video duration, and therefore no video
parsers set it.
Anton Khirnov [Sat, 12 Apr 2014 19:55:46 +0000 (21:55 +0200)]
avconv: always reset packet pts after decoding an audio frame
Currently, if a decoder sets AVFrame.pts, we'd send the same timestamp
to it twice, which is wrong.
Anton Khirnov [Fri, 11 Apr 2014 15:47:51 +0000 (17:47 +0200)]
lavc: improve AVCodecContext.delay doxy
Anton Khirnov [Sun, 13 Apr 2014 08:43:45 +0000 (10:43 +0200)]
txd: do not set the codec timebase.
It is not supposed to be changed from outside of lavc.
Set the stream timebase and average framerate instead.
Michael Niedermayer [Fri, 18 Apr 2014 21:11:31 +0000 (23:11 +0200)]
mpegvideo: Use the current_picture pts
The picture slot can be recycled by select_input_picture and
only current_picture is populated with the valid pts.
Unbreak timestamps when in cbr mode.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Michael Niedermayer [Wed, 16 Apr 2014 00:06:37 +0000 (02:06 +0200)]
avcodec/fic: Fix return value check
Fixes part of Ticket3466
Found-by: Andrey_Karpov / PVS-Studio
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Michael Niedermayer [Sat, 15 Feb 2014 16:19:59 +0000 (17:19 +0100)]
avcodec/fic: clear slice_data
Fixes artifacts
Fixes use of freed memory
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Michael Niedermayer [Sat, 15 Feb 2014 16:19:32 +0000 (17:19 +0100)]
avcodec/fic: fix slice checks
fix integer overflows
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Kostya Shishkov [Tue, 1 Apr 2014 19:00:47 +0000 (21:00 +0200)]
fic: use correct IDCT
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
Martin Storsjö [Sat, 19 Apr 2014 15:51:38 +0000 (18:51 +0300)]
mvc: Specify the pixel format for the mv-mvc* tests
Also set the RGBA pixel format correctly as the native endian format,
which is what it returns.
This fixes the tests on big endian.
Signed-off-by: Martin Storsjö <martin@martin.st>
Vittorio Giovara [Thu, 13 Mar 2014 16:16:52 +0000 (17:16 +0100)]
fate: add tests for SGI RLE and MVC1&2 decoders
Peter Ross [Mon, 17 Dec 2012 09:37:54 +0000 (20:37 +1100)]
Silicon Graphics Movie demuxer
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
Peter Ross [Mon, 17 Dec 2012 09:37:54 +0000 (20:37 +1100)]
Silicon Graphics Motion Video Compressor 1 & 2 decoder
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
Peter Ross [Mon, 17 Dec 2012 09:37:54 +0000 (20:37 +1100)]
Silicon Graphics RLE 8-bit video decoder
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
Janne Grunau [Fri, 18 Apr 2014 12:07:52 +0000 (12:07 +0000)]
h264: codec reinit: remove statements without effect
avctx->coded_{height,width} will always equal h->{height,width} since
init_dimensions() does that explicitly, Size changes are detected by
changes in mb_{height,width} earlier and propagated through the
needs_reinit variable.
Vittorio Giovara [Tue, 15 Apr 2014 13:47:30 +0000 (15:47 +0200)]
utils: add yvyu422 to avcodec_align_dimensions2
Vittorio Giovara [Sat, 19 Apr 2014 01:02:26 +0000 (03:02 +0200)]
filtfmts: remove unused lavf include
Vittorio Giovara [Fri, 11 Apr 2014 09:54:15 +0000 (11:54 +0200)]
lavfi: name anonymous structs
Martin Storsjö [Fri, 18 Apr 2014 20:40:59 +0000 (23:40 +0300)]
configure: Handle armcc 5.0
The first line of the "armcc --vsn" output in one version is
"Product: ARM Compiler 5.04".
Signed-off-by: Martin Storsjö <martin@martin.st>
Alessandro Ghedini [Sun, 13 Apr 2014 12:23:57 +0000 (14:23 +0200)]
mp3dec: decode more data from Info header
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Alessandro Ghedini [Mon, 14 Apr 2014 08:18:53 +0000 (10:18 +0200)]
crc: add ANSI CRC16 LE
This adds the reversed byte-order version of the ANSI CRC16.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Alessandro Ghedini [Sat, 12 Apr 2014 16:25:41 +0000 (18:25 +0200)]
mp3dec: move XING/Info and VBRI parsing into their own functions
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Alessandro Ghedini [Sat, 12 Apr 2014 16:25:40 +0000 (18:25 +0200)]
replaygain: allow exporting already decoded replaygain values
This adds a function to export raw replaygain values (i.e. in the (u)int32_t
form). It first checks whether AV_PKT_DATA_REPLAYGAIN side data is present, in
which case it does nothing.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Anton Khirnov [Sat, 22 Feb 2014 19:27:44 +0000 (20:27 +0100)]
resample: remove an unneeded context variable
Luca Barbato [Sat, 12 Apr 2014 14:02:50 +0000 (16:02 +0200)]
drawtext: Add fontconfig support
Introduce the `font` option and make it optional to pass a fontfile.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Martin Storsjö [Mon, 14 Apr 2014 17:43:04 +0000 (20:43 +0300)]
vp78: Align the intra4x4_pred_mode_top array within VP8Macroblock
This array is written using AV_WN32A, assuming alignment.
This hopefully fixes the failing vp7 fate test on sparc.
Signed-off-by: Martin Storsjö <martin@martin.st>
Niels Möller [Mon, 14 Apr 2014 08:26:57 +0000 (10:26 +0200)]
dcadec: Delete redundant init_get_bits call
No initialization is needed in dca_decode_frame, because the next
thing it does is calling dca_parse_frame_header, which takes care of
the needed initialization.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Stephan Soller [Sat, 12 Apr 2014 23:06:22 +0000 (01:06 +0200)]
rtmpproto: Handle RTMP chunk size packets before the connect packet
In all other cases where ff_rtmp_packet_read is used, the packet returned
is passed to rtmp_parse_result more or less immediately. In this single
case, the content of the packet was required to be a connect packet.
Some clients, e.g. Open Broadcaster Software, send a chunk size packet
before the connect packet. If the first packet is a chunk size packet,
handle it and read another one, requiring this to be a connect packet
instead.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Sun, 13 Apr 2014 10:44:03 +0000 (13:44 +0300)]
rtmpproto: Make sure to pass on the error code if read_connect failed
Previously, if read_connect failed, the ret variable was unmodified
and had the value 0, indicating success, which then was returned from
the rtmp_open function, even though it actually failed.
CC: libav-stable@libav.org
Signed-off-by: Martin Storsjö <martin@martin.st>
James Almer [Sat, 5 Apr 2014 17:00:53 +0000 (14:00 -0300)]
x86/synth_filter: remove the fma3 version ifdefs
This fixes compilation failures with --disable-fma3
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Alessandro Ghedini [Sun, 6 Apr 2014 14:25:08 +0000 (16:25 +0200)]
af_volume: implement replaygain clipping prevention
This adds a new "replaygain_noclip" option to the filter, and, if enabled,
limits the gain applied for tracks where clipping would occur.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Anton Khirnov [Thu, 20 Mar 2014 19:40:24 +0000 (20:40 +0100)]
lavr: allocate the resampling buffer with a positive size
This fixes cases where very few input samples (fewer than needed for one
output sample) are passed to lavr at the beginning.
CC:libav-stable@libav.org
Anton Khirnov [Tue, 4 Mar 2014 11:03:19 +0000 (12:03 +0100)]
FATE: add tests for more resampling modes
Anton Khirnov [Tue, 4 Mar 2014 21:38:47 +0000 (22:38 +0100)]
resample: implement flushing
Tim Walker [Sat, 12 Apr 2014 20:11:53 +0000 (22:11 +0200)]
dcadec: Use correct channel count in stereo downmix check
s->prim_channels is greater than num_core_channels
when an XCh extension is present in the bitstream.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Tim Walker [Sat, 12 Apr 2014 20:11:52 +0000 (22:11 +0200)]
dcadec: Do not decode the XCh extension when downmixing to stereo
This is neither necessary nor currently supported.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Jean First [Sun, 25 Sep 2011 14:00:04 +0000 (16:00 +0200)]
tiffdec: support predictor type 2 with RGB48 pixel format
Justin Ruggles [Mon, 30 Sep 2013 05:00:20 +0000 (01:00 -0400)]
tiffdec: remove an unneeded variable
Justin Ruggles [Mon, 30 Sep 2013 04:39:51 +0000 (00:39 -0400)]
tiff: use a better name and enum values for PhotometricInterpretation
Also add additional known values and log as missing features.
Justin Ruggles [Mon, 30 Sep 2013 04:11:12 +0000 (00:11 -0400)]
tiffdec: use a single strip if RowsPerStrip is 0
The spec does not specify that 0 is an error condition, and there
are samples which use 0 when the whole image is in one strip.
Justin Ruggles [Sun, 29 Sep 2013 23:47:55 +0000 (19:47 -0400)]
tiffdec: use bytestream2 to simplify overread/overwrite protection
Based on a patch by Paul B Mahol <onemda@gmail.com>
CC:libav-stable@libav.org
Justin Ruggles [Sun, 29 Sep 2013 23:45:57 +0000 (19:45 -0400)]
bytestream: add bytestream2_copy_buffer() functions
This is basically an overread/overwrite-safe memcpy between a
GetByteContext and a PutByteContext.
CC:libav-stable@libav.org
Paul B Mahol [Wed, 21 Mar 2012 00:10:18 +0000 (00:10 +0000)]
bytestream: add functions for accessing size of buffer
Signed-off-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
CC:libav-stable@libav.org
Justin Ruggles [Sat, 28 Sep 2013 16:20:19 +0000 (12:20 -0400)]
tiffdec: use correct data type for palette entries and set alpha to 0xFF
Jean First [Fri, 23 Sep 2011 06:22:31 +0000 (08:22 +0200)]
tiffdec: rename variables for consistency and fix variable shadowing
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
Justin Ruggles [Sat, 5 Apr 2014 18:14:30 +0000 (14:14 -0400)]
swscale: Set alpha to opaque for internal palettes.
Fixes conversion of pal8 to rgb formats with alpha.
Updated references for 2 FATE tests which previously encoded fully
transparent images.
Based on a patch by Baptiste Coudurier <baptiste.coudurier@gmail.com>
Justin Ruggles [Sat, 5 Apr 2014 18:46:22 +0000 (14:46 -0400)]
swscale: fix some undefined signed left shifts
Based on a patch by Michael Niedermayer <michaelni@gmx.at>
Justin Ruggles [Sat, 5 Apr 2014 18:05:19 +0000 (14:05 -0400)]
swscale: fix an implementation-defined unsigned-to-signed conversion
Katerina Barone-Adesi [Sun, 6 Apr 2014 10:32:49 +0000 (12:32 +0200)]
doc: Libavfilter English cleanup
This aims to standardize the English used in the libavfilter
documentation, describing the same thing in the same way more often,
and significantly reduce the amount of outright errors, including
run-on sentences and sentence fragments.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Michael Niedermayer [Wed, 9 Apr 2014 16:22:53 +0000 (18:22 +0200)]
mp3enc: Properly write bitrate value in XING header
Instead of using a fixed bitrate_idx, calculate a matching bitrate for
the XING header.
Using a fixed bitrate_idx causes tools such as file(1) and mediainfo(1)
to report wrong bitrate and bitrate mode when using CBR.
Bug-Id: https://bugs.debian.org/736088
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
Anton Khirnov [Tue, 4 Mar 2014 16:12:33 +0000 (17:12 +0100)]
resample: split linear into its own function
Anton Khirnov [Tue, 4 Mar 2014 15:56:01 +0000 (16:56 +0100)]
resample: add initial padding explicitly
This simplifies the code, since we do not have to deal with a possibly
negative source index anymore.