| Age | Commit message (Collapse) | Author |
|---|
|
The lPitch setting of the offscreen buffer was not taken into account,
which let to a garbled image if the video card driver did support YV12 or
YUV2 color formats. This patch fixes bug #72.
|
|
|
|
The test stream sometimes causes the ASF demuxer to report "unknown GUID"
several times then act as if the headers have been received, with the audio
output thread waiting forever for data to be passed to it by the decoder, in
this case the ffmpeg audio decoder.
This particular hang occurs if playback is stopped before the demuxer has
decided that headers have been received (later, and we'd have the problem
which the parent cset of this cset fixes); having the demuxer control insert
empty audio buffers where it would otherwise wait forever for the audio
output thread to receive some data.
Test stream: mmsh://213.92.19.8:80/radiodeejay?MSWMExt=.asf
|
|
playback stopped.
|
|
|
|
libmms will always fail to request media with URIs containing percent-encoded
characters. This is because the path component in the MMS URI should be
decoded before it is sent to the server.
http://download.microsoft.com/download/9/5/E/95EF66AF-9026-4BB0-A41D-A4F81802D92C/%5BMS-MMSP%5D.pdf
(page 48)
|
|
|
|
|
|
|
|
These are known to be present in some nvidia graphics hardware.
|
|
These drivers use "NV* Video Texture" instead of "* Textured Video".
|
|
|
|
|
|
|
|
in our plugins (for now).
|
|
|
|
|
|
|
|
|
|
|
|
When it comes to FLAC audio files, seeking relies on seekpoints which are
not always present, and even when they are, sometimes it fails. Also, as far
as I can see, xine is unable to play a FLAC stream starting at an arbitrary
position.
Other players (namely mplayer) do not rely on seekpoints when they handle
FLAC files and they don't suffer from these problems.
With this patch, time-based seeking doesn't change, while position-based
seeking is completely independent from seekpoints.
|
|
|
|
- goom initialization
- matroska playing recent files with AAC
- replace free() by ffmpeg's av_free() in ff decoders
|
|
Add warning flags to the DEBUG_CFLAGS too.
|
|
|
|
We don't want to have warnings conditional to extreme optimisations:
if the compiler supports them, they should be enabled.
The reason why they were conditional is that the code testing for GCC
presence enabled both warnings and optimisations, and the warnings
were carried over when the optimisations were made optional.
|
|
When passing CC="ccache gcc", we don't want to quote "$CC" calls,
otherwise the command won't be found.
|
|
former.
The CC_CHECK_CFLAGS macro is already present in attributes.m4, so no
need to keep the old version we had around, especially since the new
one caches its results too.
|
|
equivalent.
I wanted this because I'm going to reuse attribute.m4 with that macro
for a couple other projects.
This also removes AC_TRY_LDFLAGS since it's not used any longer
(replaced by CC_CHECK_LDFLAGS, which also caches results).
|
|
This should make it more readable, and avoid a few extra test. Also
check for the -Werror= style for GCC 4.3 instead of the old
-Werror-implicit-function-declaration.
|
|
This adds variants of CC_CHECK_CFLAGS that append the flags
automatically, and a few comments.
|
|
There is needed some ui<->lib interaction: used caca display is optionally delivered from ui to the caca vo plugin.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This fixes several bugs/incompatibilities in spudvb decoder.
You can find various test samples at http://hftom.free.fr/video_samples/
--HG--
extra : transplant_source : %CD%CE%3A%3F%B6%8BN%FD2%5D%DB%9A%AB%AF%C2%E1%2B%1A%B1%23
|
|
The playback pointer did occasionally overrun the write pointer, which
results in audible gaps in playback. To solve this, the slotted ringbuffer
was replaced by a simple ringbuffer and the service threads now checks
periodically if there is still data in the buffer instead of checking at the
slots borders.
--HG--
extra : transplant_source : %A1%12%60%B9%E4%AB%E7%7C%D7%D6%BD%7C%C6M%C9%94s%7F%BE%91
|
|
If neither YV12 nor YUV2 format is supported by DirectX, a secondary buffer
with the same pixelformat as the primary buffer is used. However, the pixel
format of the primary was not passed to CreateSurface, instead a 16 bit
pixelformat was created, which is the reason why the video was broken on
24bit or 32bit desktops.
--HG--
extra : transplant_source : %A6-%ADwyY%EE%C8%26%E2%5E%2A%83%0A%0B/%CBM%23%0F
|
|
Without this patch, the video freezes.
Available corrupted sample:
http://hftom.free.fr/video_samples/corrupted_video.m2t
--HG--
extra : transplant_source : %86g%9A%B1%AF%12L%7E%3EN%8C%0FT%D2%D8%3B%7Dv%F0%14
|
|
To place SDL output inside an existing window, the SDL_WINDOWID environement
variable has to be set. This was done by the SDL video out pluging if
HAVE_X11 was defined, but not for WIN32, where it works as well.
--HG--
extra : transplant_source : E%169%A2%B4%93%3CY%07%A9%9F%1C%E0%B8-%14m4%A2%11
|
|
let raw video out plugin handle cropping.
--HG--
extra : transplant_source : %F3l%EFUI%28%2C%A0%0C%3E%AD%EA%EC%80%D1%9A%C0/%20%E5
|
|
|
|
Date: Sun, 28 Sep 2008 20:17:54 +0200
The channel order of aac 5.1 audio is wrong during playback. IIRC, the
internal channels order for this kind of streams is the same as dts: dts
output is reordered according to alsa specs (and ac3 5.1 also works), but
aac is not and - for instance - front center is send to front left.
The audio channels configuration table should be something like this
## ---------------------
## | Config: 5.1 Ch |
## ---- ---------------- -------------- --------------
## | Ch | AAC/DTS | ALSA | AC3 |
## ---- ---------------- --------------- --------------
## | 00 | Center front | Left front | Left front |
## | 01 | Left front | Right front | Center |
## | 02 | Right front | Left back | Right front |
## | 03 | Left back | Right back | Left back |
## | 04 | Right back | Center | Right back |
## | 05 | LFE | LFE | LFE |
## ---- ---------------- --------------- --------------
|
|
Date: Sat, 28 Jun 2008 17:29:59 +0200
This patch adds this to the FLAC demuxer.
|
|
|
|
It was not possible to compile some parts of XINE-LIB with NLS enabled. It's
also a good idea to explicitely enable --enable-stdcall-fixup for solving a
huge amount of warnings. Attached patch fixes these problems.
|
|
|
