May 2011 - MPlayer-dev-eng - lists.mplayerhq.hu

Homepage DOCS needs update
by Ingo Brückl 24 Apr '14

24 Apr '14

If it's already on the to-do list ignore, but the documentation on the homepage doesn't get updated. Ingo

3 12

[PATCH] add support of compression algorithm 3 in mkv demuxer
by Alexander Beregalov 02 Jan '12

02 Jan '12

Specs from http://www.matroska.org/technical/specs/ : ContentCompAlgo The compression algorithm used. Algorithms that have been specified so far are: 0 - zlib, 1 - bzlib, 2 - lzo1x, 3 - Header Stripping. ContentCompSettings Settings that might be needed by the decompressor. For Header Stripping (ContentCompAlgo=3), the bytes that were removed from the beggining of each frames of the track. I have such file, it works for me. | + A track | + Track number: 2 | + Track UID: 2 | + Track type: audio | + Lacing flag: 1 | + Default flag: 1 | + Codec ID: A_AC3 | + Content encodings | + Content encoding | + Content compression | + Algorithm: 3 (header removal) | + Settings: length 2, data: 0x0b 0x77 See an attachment. Please comment.

7 21

DxVA in MPlayer
by Georgi Petrov 26 Jun '11

26 Jun '11

Hello, I'm Georgi Petrov and a year ago I developed the Direct 3D renderer. Now I'm interested in implementing DxVA support in MPlayer. Yesterday I started to investigate the amount of work and any difficulties I may have and I came up with the following questions: 1. Since DxVA 2 requires EVR renderer (the current Direct 3D one won't work), I suppose that the best way to start is implementing a working EVR renderer first. If so, are there anybody who is familiar at least a little with it? The only thing that bothers me is that EVR is supposed to be used in a DirectShow chain and MPlayer has nothing to do with DirectShow... In short - does anybody know is it feasible to develop the EVR renderer first and test it with the current (non-accelerated) internal MPlayer codecs? 2. Once the EVR custom presenter is ready, FFmpeg already has the needed infrastructure for offloading computations for H.264, VC-1, MPEG-2 and MPEG-4 ASP. Excuse me if I don't get it, but each one of these standards should have an "accelerated" implementation, which from MPlayer's point of view is a new codec. Then inside the codec there are places, where one can hook and transfer the decoding part to the DxVA/VA API/VDAPU at any given stage (IDCT, MO and so on). Is this right? I'm starting to look at the code right now, but a higher abstract point of view is welcome. 3. As far as I understand this approach is used by VA API and VDPAU. Can I reuse their infrastructure as well? On this page I see a VA API patches: http://www.splitted-desktop.com/~gbeauchesne/mplayer-vaapi/ Are they part of MPlayer already or not? 4. Can somebody, who has already worked on VA API/VDPAU implementations give me some clue? How much work is involved in getting the whole chain work? Are most of the puzzle pieces in place or I will have to dig deeper into understanding how H.264 / VC-1 and so on work on a pure codec level? 5. I am trying to figure out a realistic estimate about how much time and hard work this is about, so any help in this direction is welcome. Greeting, Georgi

9 46

Re: [MPlayer-dev-eng] [PATCH] MNG output support for MPlayer
by Stefan Schuermans 18 Jun '11

18 Jun '11

Clément Bœsch wrote: > On Thu, May 12, 2011 at 12:02:38AM +0200, Stefan Schuermans wrote: >>>> + /* we are initialized */ >>>> + vomng->b_is_init = 1; >>>> + >>> It took me quiet a few time to realize what the b meant… please >>> avoid that notation. >> Renamed to "is_init". > > Thank you. Of course, if you had done it for all the other i_* and p_* > that would have been awesome, but I don't really care. Your patch > looks fine. I removed all the prefixes, because I do not want my code to make MPlayer less awesome that it is. :-) >> + for (y = 0; y < hf; ++y) > > Just a nitpick while I'm at it: y++ is preferred (don't waste your > time re-sending a patch for this). Since this is not ++C code it > doesn't affect at all the generated code. As I'm sending a new version of the patch anyway, I also changed all standalone prefix increments to postfix increments. Best regards, Stefan

4 17

[PATCH] Add size based support for -endpos in MPlayer
by Clément Bœsch 11 Jun '11

11 Jun '11

A small patch to add size based support for -endpos in MPlayer. I tested it with random videos, audio only files and -loop option and it seems to work. Anything else I should test? However, I don't understand why mpctx->sh_audio is tested when mpctx->sh_video is NULL; is there a way to call the code around mplayer.c:L3698 without having neither mpctx->sh_video and mpctx->sh_audio? And then, does this mean mpctx->stream could be NULL? -- Clément B. Not sent from a jesusPhone.

2 12

[PATCH 1/5] Remove some useless static zero initializations.
by Clément Bœsch 11 Jun '11

11 Jun '11

--- mplayer.c | 48 ++++++++++++++++++++++++------------------------ 1 files changed, 24 insertions(+), 24 deletions(-) diff --git a/mplayer.c b/mplayer.c index c946b4a..501e2d4 100644 --- a/mplayer.c +++ b/mplayer.c @@ -162,7 +162,7 @@ static int cfg_include(m_option_t *conf, char *filename){ return m_config_parse_config_file(mconfig, filename); } -static int max_framesize=0; +static int max_framesize; int noconsolecontrols=0; //**************************************************************************// @@ -188,16 +188,16 @@ int fixed_vo=0; // benchmark: double video_time_usage=0; double vout_time_usage=0; -static double audio_time_usage=0; -static int total_time_usage_start=0; -static int total_frame_cnt=0; -static int drop_frame_cnt=0; // total number of dropped frames +static double audio_time_usage; +static int total_time_usage_start; +static int total_frame_cnt; +static int drop_frame_cnt; // total number of dropped frames int benchmark=0; // options: #define DEFAULT_STARTUP_DECODE_RETRY 8 int auto_quality=0; -static int output_quality=0; +static int output_quality; float playback_speed=1.0; @@ -207,7 +207,7 @@ int use_gui=0; int enqueue=0; #endif -static int list_properties = 0; +static int list_properties; int osd_level=1; // if nonzero, hide current OSD contents when GetTimerMS() reaches this @@ -218,12 +218,12 @@ int osd_fractions = 0; // determines how fractions of seconds are displayed int term_osd = 1; static char* term_osd_esc = "\x1b[A\r\x1b[K"; -static char* playing_msg = NULL; +static char* playing_msg; // seek: static double seek_to_sec; -static off_t seek_to_byte=0; -static off_t step_sec=0; -static int loop_seek=0; +static off_t seek_to_byte; +static off_t step_sec; +static int loop_seek; static m_time_size_t end_at = { .type = END_AT_NONE, .pos = 0 }; @@ -251,7 +251,7 @@ int dvdsub_lang_id; int vobsub_id=-1; char* audio_lang=NULL; char* dvdsub_lang=NULL; -static char* spudec_ifo=NULL; +static char* spudec_ifo; char* filename=NULL; //"MI2-Trailer.avi"; int forced_subs_only=0; int file_filter=1; @@ -271,14 +271,14 @@ int capture_dump = 0; // A-V sync: static float default_max_pts_correction=-1; static float max_pts_correction=0;//default_max_pts_correction; -static float c_total=0; +static float c_total; float audio_delay=0; -static int ignore_start=0; +static int ignore_start; -static int softsleep=0; +static int softsleep; double force_fps=0; -static int force_srate=0; +static int force_srate; static int audio_output_format=-1; // AF_FORMAT_UNKNOWN int frame_dropping=0; // option 0=no drop 1= drop vo 2= drop decode static int play_n_frames=-1; @@ -309,9 +309,9 @@ static const vf_info_t* const libmenu_vfs[] = { &vf_info_menu, NULL }; -static vf_instance_t* vf_menu = NULL; +static vf_instance_t* vf_menu; int use_menu = 0; -static char* menu_cfg = NULL; +static char* menu_cfg; static char* menu_root = "main"; #endif @@ -337,14 +337,14 @@ int edl_start_pts = 0; ///< Automatically add/sub this from EDL start/stop pos int use_filedir_conf; int use_filename_title; -static unsigned int initialized_flags=0; +static unsigned int initialized_flags; /// step size of mixer changes int volstep = 3; #ifdef CONFIG_CRASH_DEBUG static char *prog_path; -static int crash_debug = 0; +static int crash_debug; #endif /* This header requires all the global variable declarations. */ @@ -759,7 +759,7 @@ static void child_sighandler(int x){ #endif static void exit_sighandler(int x){ - static int sig_count=0; + static int sig_count; #ifdef CONFIG_CRASH_DEBUG if (!crash_debug || x != SIGTRAP) #endif @@ -1351,7 +1351,7 @@ struct mp_osd_msg { }; /// OSD message stack. -static mp_osd_msg_t* osd_msg_stack = NULL; +static mp_osd_msg_t* osd_msg_stack; /** * \brief Add a message on the OSD message stack @@ -1439,7 +1439,7 @@ static void clear_osd_msgs(void) { static mp_osd_msg_t* get_osd_msg(void) { mp_osd_msg_t *msg,*prev,*last = NULL; - static unsigned last_update = 0; + static unsigned last_update; unsigned now = GetTimerMS(); unsigned diff; char hidden_dec_done = 0; @@ -2039,7 +2039,7 @@ static void adjust_sync_and_print_status(int between_frames, float timing_error) v_pts = mpctx->sh_video->pts; { - static int drop_message=0; + static int drop_message; double AV_delay = a_pts - audio_delay - v_pts; double x; // not a good idea to do A-V correction with with bogus values -- 1.7.4.4

7 44

CrystalHD & 1080p mkv memory leaks.
by wallak＠free.fr 04 Jun '11

04 Jun '11

Hello All, I was trying the last mplayer / ffmpeg tree with my bcm970015 board. Everything is fine on a .ts file, but I have 1MB/s memory leaks when I'm using the mkv container. I was unable to debug this situation. Any idea to solve this issue ? Best Regards,

4 7

Small configure improvements
by Ivan Kalvachev 31 May '11

31 May '11

Two small improvements that are already done by ffmpeg configure. 1. Check if the compiler works at all. I had a debug session, where x86 was detected as bigendian and crosscompile. It turned out the compiler was broken and failing checks lead to strange situation. BTW I accept proposals for better messages, but I won't give you credit for them :P 2. Add "-fno-tree-vectorize". It have been used in ffmpeg from quite some time. This option have history of miscompilation, while its speed advantages are rather questionable. We have recent report that gcc 4.6.0 and "-march=corei7-avx -O3" miscompiles ac3dec.c . Compile check is used for detection of the option as it is rather recent. I intend to commit them soon.

2 2

[PATCH] Quick and dirty fix build after BUILD_ROOT got removed from FFmpeg (798b264)
by Clément Bœsch 30 May '11

30 May '11

I'll commit the attached patch as a temporary fix in a few hours if no one comes up with something better. Regards, -- Clément B.

2 3

swab is in string.h on win32 and declaration conflicts
by Gianluigi Tiesi 29 May '11

29 May '11

Hi, on mingw32 swab is defined is string.h thus not detected by configure that looks in unistd.h the signature of mplayer provided differs from the one in string.h: _CRTIMP void __cdecl __MINGW_NOTHROW swab (const char*, char*, size_t); but mplayer declares it as: void swab(const void *from, void *to, ssize_t n); ssize_t is signed while size_t unsigned a possible solution is include also string.h in the configure check Regards -- Gianluigi Tiesi <sherpya(a)netfarm.it> EDP Project Leader Netfarm S.r.l. - http://www.netfarm.it/ Free Software: http://oss.netfarm.it/

6 16