April 2006 - MPlayer-dev-eng - lists.mplayerhq.hu

patch backlog maintenance
by Diego Biurrun 19 Jun '08

19 Jun '08

Folks, I'm awfully overloaded. Could somebody take over the maintenance of the patch backlog that I took over from Attila some time ago? All you need to do is keep track of all unapplied patches in an mbox or something and make it available from time to time. Sending out periodic reminders from time to time is a plus. Thanks Diego

3 13

Synchronization of players in LAN
by J.A. Gutierrez 18 Jun '08

18 Jun '08

Hello I'm trying to synchronise several players running on several computers connected to a dedicated 10/100 Ethernet switch. Each one reads a different movie (video only) from its internal hard disk, and resolution will be 768x576. I'm thinking of a simple solution: - Add some code at mplayer startup so it will wait for an UDP broadcast packet (sent by a central master in the LAN). This way all the players would start at the same time. - Use NTP to keep in sync all the computers' clocks, and trust mplayer's rtc synchronisation feature. With this simple setup, could we be sure all the players would keep the synchronisation? Or should the master computer track them all and take control on their speeds? How difficult would be to add code so all the player synchronise using the LAN? Or should I use some external clock connected to all serial ports? Or use a totally different approach? Thanks. -- finger spd(a)shiva.cps.unizar.es for PGP / .mailcap tip of the day: / La vida es una carcel application/ms-tnef; cat '%s' > /dev/null / con las puertas abiertas text/x-vcard; cat '%s' > /dev/null / (A. Calamaro)

14 40

[PATCH] Use XScreenSaverSuspend if supported
by Ismail Donmez 28 Dec '07

28 Dec '07

Hi, X.org 7.1 will feature a new function called, XScreenSaverSuspend which can suspend/resume screensavers. This is way more better than the current hacks we have. The patch is by Fredrik Höglund ( fredrik at kde.org ), I am proxying for him since he is not subscribed. Please review/comment or just apply :-) Regards, ismail -- <Terwou> \\o \o> <o/ o//

8 29

[PATCH] updated diff for libdvdread-0.9.5
by Emanuele Giaquinta 15 Aug '07

15 Aug '07

Hi, Attached is the patchset for libdvdread, updated for version 0.9.5. -- Emanuele Giaquinta

7 20

NOVIRUS Bug 188 - [Patch] add vob subtitle to mencoder.
by ha shao 30 Apr '07

30 Apr '07

Vobsub is quite popular but mencoder don't support it. This patch add the -vobsub -vobsubid to mencode. The codes were mostly copy & paste from mplayer's vobsub sessions (FIXME: tags included). I don't really know what I was doing but it simply works for me. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com Index: cfg-common.h =================================================================== RCS file: /cvsroot/mplayer/main/cfg-common.h,v retrieving revision 1.137 diff -u -r1.137 cfg-common.h --- cfg-common.h 18 Oct 2004 20:41:05 -0000 1.137 +++ cfg-common.h 2 Jan 2005 10:24:51 -0000 @@ -251,6 +251,9 @@ {"sub-bg-alpha", &sub_bg_alpha, CONF_TYPE_INT, CONF_RANGE, 0, 255, NULL}, {"sub-no-text-pp", &sub_no_text_pp, CONF_TYPE_FLAG, 0, 0, 1, NULL}, {"sub-fuzziness", &sub_match_fuzziness, CONF_TYPE_INT, CONF_RANGE, 0, 2, NULL}, + {"vobsub", &vobsub_name, CONF_TYPE_STRING, 0, 0, 0, NULL}, + {"vobsubid", &vobsub_id, CONF_TYPE_INT, CONF_RANGE, 0, 31, NULL}, + #endif #ifdef USE_OSD {"font", &font_name, CONF_TYPE_STRING, 0, 0, 0, NULL}, Index: cfg-mplayer.h =================================================================== RCS file: /cvsroot/mplayer/main/cfg-mplayer.h,v retrieving revision 1.236 diff -u -r1.236 cfg-mplayer.h --- cfg-mplayer.h 27 Dec 2004 19:34:42 -0000 1.236 +++ cfg-mplayer.h 2 Jan 2005 10:24:51 -0000 @@ -327,10 +327,6 @@ {"menu", "OSD menu support was not compiled in.\n", CONF_TYPE_PRINT,0, 0, 0, NULL}, #endif - // these should be moved to -common, and supported in MEncoder - {"vobsub", &vobsub_name, CONF_TYPE_STRING, 0, 0, 0, NULL}, - {"vobsubid", &vobsub_id, CONF_TYPE_INT, CONF_RANGE, 0, 31, NULL}, - {"sstep", &step_sec, CONF_TYPE_INT, CONF_MIN, 0, 0, NULL}, // set a-v distance, should be moved to -common and supported in MEncoder Index: mencoder.c =================================================================== RCS file: /cvsroot/mplayer/main/mencoder.c,v retrieving revision 1.259 diff -u -r1.259 mencoder.c --- mencoder.c 27 Dec 2004 17:30:12 -0000 1.259 +++ mencoder.c 2 Jan 2005 10:24:53 -0000 @@ -196,6 +196,7 @@ float sub_delay=0; float sub_fps=0; int sub_auto = 0; +char *vobsub_name=NULL; int subcc_enabled=0; int suboverlap_enabled = 1; @@ -489,6 +490,28 @@ vo_init_osd(); + //==================== Open VOB-Sub ============================ + if (vobsub_name){ + vo_vobsub=vobsub_open(vobsub_name,spudec_ifo,1,&vo_spudec); + if(vo_vobsub==NULL) + mp_msg(MSGT_CPLAYER,MSGL_ERR,MSGTR_CantLoadSub,vobsub_name); + }else if(sub_auto && filename && (strlen(filename)>=5)){ + /* try to autodetect vobsub from movie filename ::atmos */ + char *buf = malloc((strlen(filename)-3) * sizeof(char)); + memset(buf,0,strlen(filename)-3); // make sure string is terminated + strncpy(buf, filename, strlen(filename)-4); + vo_vobsub=vobsub_open(buf,spudec_ifo,0,&vo_spudec); + free(buf); + } + + if(vo_vobsub){ + sub_auto=0; // don't do autosub for textsubs if vobsub found + // inited_flags|=INITED_VOBSUB; + vobsub_set_from_lang(vo_vobsub, dvdsub_lang); + // check if vobsub requested only to display forced subtitles + forced_subs_only=vobsub_get_forced_subs_flag(vo_vobsub); + } + stream=open_stream(filename,0,&file_format); if(!stream){ @@ -1131,7 +1154,11 @@ else sscanf(seek_to_sec, "%f", &d); - demux_seek(demuxer, d, 1); + if (demux_seek(demuxer, d, 1)){ + /* FIXME there should be real seeking for vobsub */ + if (vo_vobsub) + vobsub_seek(vo_vobsub, d_video->pts); + } // there is 2 way to handle the -ss option in 3-pass mode: // > 1. do the first pass for the whole file, and use -ss for 2nd/3rd pases only // > 2. do all the 3 passes with the same -ss value @@ -1548,14 +1575,41 @@ // DVD sub: if(vo_spudec||vobsub_writer){ unsigned char* packet=NULL; - int len; - while((len=ds_get_packet_sub(d_dvdsub,&packet))>0){ - mp_msg(MSGT_MENCODER,MSGL_V,"\rDVD sub: len=%d v_pts=%5.3f s_pts=%5.3f \n",len,sh_video->pts,d_dvdsub->pts); - if (vo_spudec) - spudec_assemble(vo_spudec,packet,len,90000*d_dvdsub->pts); - if (vobsub_writer) - vobsub_out_output(vobsub_writer,packet,len,mux_v->timer + d_dvdsub->pts - sh_video->pts); - } + int len, timestamp; + // Get a sub packet from the dvd or a vobsub and make a timestamp relative to sh_video->timer + while(1) { + // FIXME: messed up timer and pts????????? + // Vobsub + if(vo_vobsub) { + if(sh_video->pts+sub_delay>=0) { + len = vobsub_get_packet(vo_vobsub,sh_video->pts+sub_delay,(void**)&packet,&timestamp); + if(len > 0) { + timestamp -= (sh_video->pts + sub_delay - sh_video->timer)*90000; + mp_dbg(MSGT_CPLAYER,MSGL_V,"\rVOB sub: len=%d v_pts=%5.3f v_timer=%5.3f sub=%5.3f ts=%d \n",len,sh_video->pts,sh_video->timer,timestamp / 90000.0,timestamp); + } + } + } else { + // DVD sub + len=ds_get_packet_sub(d_dvdsub,&packet); + if(len > 0) { + float x = d_dvdsub->pts - sh_video->pts; + if (x < -10 || x > 10) // prevent missing subs on pts reset + timestamp = 90000*(sh_video->timer + d_dvdsub->pts + sub_delay - sh_video->pts); + else + timestamp = 90000*(sh_video->timer + sub_delay); + + mp_msg(MSGT_MENCODER,MSGL_V,"\rDVD sub: len=%d v_pts=%5.3f s_pts=%5.3f ts=%d \n",len,sh_video->pts,d_dvdsub->pts,timestamp); + } + } + if (len <= 0 || !packet) break; + if (timestamp < 0) timestamp = 0; + else{ + if (vo_spudec) + spudec_assemble(vo_spudec,packet,len,timestamp); + if (vobsub_writer) + vobsub_out_output(vobsub_writer,packet,len,mux_v->timer + d_dvdsub->pts - sh_video->pts); + } + } if (vo_spudec) { spudec_heartbeat(vo_spudec,90000*sh_video->pts); vo_osd_changed(OSDTYPE_SPU); @@ -1603,6 +1657,11 @@ mp_msg(MSGT_MENCODER,MSGL_FATAL,MSGTR_ErrorWritingFile, out_filename); mencoder_exit(1, NULL); } + +if(vo_vobsub) + vobsub_close(vo_vobsub); + vo_vobsub = NULL; + if(vobsub_writer) vobsub_out_close(vobsub_writer);

5 4

Vidix driver for FX5500
by Pascal Yu 24 Jan '07

24 Jan '07

Hi All, Just subscribed to this mplayer developer list. I'm using mplayer inside GEEXBOX. As Khyron, geexbox developer, suggested I should check the list here for a possible patch.(not asking for a patch) I recently switched to a FX5500 256M card for its SVideo out and possible vidix driver, later found vidix driver doesn't work for this card, only vesa. But the PIII CPU inside the box is not so powerful. Did few changed for nvidia_vid.c and pci.db enabled it to recgonize FX5500, also 256M instead of 128M plus some memory alignment. Added several lines to video filter expand, finished the last two features mentioned in mplayer Doc but missing from its source support, aspect and roundup. This is also a work around fix for some diagonally lines show in some media that I played, I think this issue could be common for NV30 architech or NV34 or NV34B, which my card is. Should I send the modification I made, eveyone interested? Thanks, Pascal

3 2

vidix status??
by Reimar Döffinger 24 Jan '07

24 Jan '07

Hi, I've been trying to sync out vidix with the CVS versions, just to see that some of our files are newer (e.g. unichrome_vid), some are older and some a completely different! And most of the changes are only cosmetics. So what to do about it?? Greetings, Reimar DÃ¶ffinger

5 14

[PATCH] Maintain subtitles in pts order in spudec.c
by Wolfram Gloger 24 Jan '07

24 Jan '07

Hi! I think subtitles in spudec.c should be maintained in start_pts order. E.g. when detecting timestamps and substreams in demux_mpg_open(), a subtitle from near the end of the input file can be read, which subsequently "blocks" the list -- the first packet is never taken in spudec_heartbeat(). Alternatively, demux_mpg_open() could at its end flush the queue with spudec_reset(), but I think the following is also very simple and preferable, because it is also tolerant against subtitles in the wrong order in a broken input file. Regards, Wolfram. diff -urb MPlayer-1.0pre7/spudec.c MPlayer-1.0pre7-wg/spudec.c --- MPlayer-1.0pre7/spudec.c Sat Jan 8 22:06:04 2005 +++ MPlayer-1.0pre7-wg/spudec.c Fri Sep 23 14:19:29 2005 @@ -95,11 +95,22 @@ static void spudec_queue_packet(spudec_handle_t *this, packet_t *packet) { - if (this->queue_head == NULL) + packet_t *last = 0, *p = this->queue_head; + while (p && p->start_pts<packet->start_pts) { + last = p; + p = p->next; + } + if (last) { + packet->next = last->next; + last->next = packet; + if (last == this->queue_tail) + this->queue_tail = packet; + } else { + packet->next = p; this->queue_head = packet; - else - this->queue_tail->next = packet; + if (!p) this->queue_tail = packet; + } } static packet_t *spudec_dequeue_packet(spudec_handle_t *this)

2 3

[PATCH] Various updates to GIF demuxer
by John Koleszar 16 Jan '07

16 Jan '07

Hi, Attached patch adds support for several animated gif features that either weren't supported by or outright crashed mplayer, including variable frame sizes, interlacing, transparency, and some alternative frame update methods. Also fixed a bug extracting timestamps, which was probably introduced by a change in giflib, since the code in question dates back to rev 1.1. Samples (courtesy of gifanimations.com) here: ftp://opensource:opensource@ftp.on2.com/gifanimations_com-samples.tar.bz2 Comments/testing welcome. Cheers, John Index: libmpdemux/demux_gif.c =================================================================== RCS file: /cvsroot/mplayer/main/libmpdemux/demux_gif.c,v retrieving revision 1.5 diff -u -r1.5 demux_gif.c --- libmpdemux/demux_gif.c 5 Aug 2005 19:57:46 -0000 1.5 +++ libmpdemux/demux_gif.c 3 Feb 2006 20:50:45 -0000 @@ -19,8 +19,21 @@ #include "stheader.h" #include <gif_lib.h> -static int current_pts = 0; -static unsigned char *pallete = NULL; + +struct gif_priv_s { + GifFileType* gif_hndl; + unsigned char* pallete; + unsigned char* canvas; + unsigned int canvas_sz; +}; +#define gif_hndl (((struct gif_priv_s*)(demuxer->priv))->gif_hndl) +#define pallete (((struct gif_priv_s*)(demuxer->priv))->pallete) +#define canvas (((struct gif_priv_s*)(demuxer->priv))->canvas) +#define canvas_sz (((struct gif_priv_s*)(demuxer->priv))->canvas_sz) + +static int current_pts = 0, + ilace_offset[] = { 0, 4, 2, 1 }, + ilace_jmp[] = { 8, 8, 4, 2 }; #define GIF_SIGNATURE (('G' << 16) | ('I' << 8) | 'F') @@ -40,13 +53,16 @@ static int demux_gif_fill_buffer(demuxer_t *demuxer, demux_stream_t *ds) { - GifFileType *gif = (GifFileType *)demuxer->priv; + GifFileType *gif = gif_hndl; sh_video_t *sh_video = (sh_video_t *)demuxer->video->sh; GifRecordType type = UNDEFINED_RECORD_TYPE; int len = 0; demux_packet_t *dp = NULL; ColorMapObject *effective_map = NULL; char *buf = NULL; + int trans = 0; + int trans_color = 0; + int reaction = 0; while (type != IMAGE_DESC_RECORD_TYPE) { if (DGifGetRecordType(gif, &type) == GIF_ERROR) { @@ -70,8 +86,13 @@ } if (code == 0xF9) { int frametime = 0; - if (p[0] == 4) // is the length correct? - frametime = (p[1] << 8) | p[2]; // set the time, centiseconds + if (p[0] == 4) { // is the length correct? + frametime = (p[3] << 8) | p[2]; // set the time, centiseconds + reaction = (p[1] >> 2) & 0x3; + // Get transparancy info + trans = p[1] & 0x1; + trans_color = p[4]; + } current_pts += frametime; } else if ((code == 0xFE) && (verbose)) { // comment extension // print iff verbose @@ -103,21 +124,35 @@ } len = gif->Image.Width * gif->Image.Height; - dp = new_demux_packet(len); + dp = new_demux_packet(canvas_sz); buf = malloc(len); memset(buf, 0, len); - memset(dp->buffer, 0, len); + memcpy(dp->buffer, canvas, canvas_sz); - if (DGifGetLine(gif, buf, len) == GIF_ERROR) { - PrintGifError(); - return 0; // oops + if(gif->Image.Interlace) { + int i,j; + for(i = 0; i < 4; i++) { + for(j = ilace_offset[i]; j < gif->Image.Height; j += ilace_jmp[i]) { + if (DGifGetLine(gif, buf + (j * gif->Image.Width), gif->Image.Width) + == GIF_ERROR) + { + PrintGifError(); + return 0; // oops + } + } + } + } else { + if (DGifGetLine(gif, buf, len) == GIF_ERROR) { + PrintGifError(); + return 0; // oops + } } effective_map = gif->Image.ColorMap; if (effective_map == NULL) effective_map = gif->SColorMap; { - int y; + int y,x,width,height; // copy the pallete for (y = 0; y < 256; y++) { @@ -127,15 +162,55 @@ pallete[(y * 4) + 3] = 0; } - for (y = 0; y < gif->Image.Height; y++) { + // Use the canvas w/h as a bounding box for the blit + height = gif->Image.Height; + width = gif->Image.Width; + if(gif->Image.Top + gif->Image.Height > gif->SHeight) + height = gif->SHeight - gif->Image.Top; + if(gif->Image.Left + gif->Image.Width > gif->SWidth) + width = gif->SWidth - gif->Image.Left; + + for (y = 0; y < height; y++) { unsigned char *drow = dp->buffer; unsigned char *gbuf = buf + (y * gif->Image.Width); - drow += gif->Image.Width * (y + gif->Image.Top); + drow += gif->SWidth * (y + gif->Image.Top); drow += gif->Image.Left; - memcpy(drow, gbuf, gif->Image.Width); + if(!trans) { + // No transparancy, can do a fast blit + memcpy(drow, gbuf, width); + } else { + // Do a byte by byte blit for all non transparant pixels. + for (x = 0; x < width; x++) { + if(gbuf[x] != trans_color) + drow[x] = gbuf[x]; + } + } } + + switch(reaction) { + case 1: + // Preserve the current frame contents for the next frame. + memcpy(canvas, dp->buffer, canvas_sz); + break; + case 2: + // Fill the image area with the background color + for (y = 0; y < height; y++) { + unsigned char *drow = canvas; + drow += gif->SWidth * (y + gif->Image.Top); + drow += gif->Image.Left; + memset(drow, gif->SBackGroundColor, width); + } + break; + case 3: + // Preserve the previous frame contents for the next frame. + // No action required. + break; + default: + // Wipe the whole canvas with the background color + memset(canvas, gif->SBackGroundColor, canvas_sz); + } } free(buf); @@ -199,16 +274,19 @@ sh_video->bih->biCompression = sh_video->format; sh_video->bih->biBitCount = 8; sh_video->bih->biPlanes = 2; - pallete = (unsigned char *)(sh_video->bih + 1); + demuxer->priv = malloc(sizeof(struct gif_priv_s)); + pallete = (unsigned char *)(sh_video->bih + 1); + canvas_sz = gif->SWidth * gif->SHeight; + gif_hndl = gif; + canvas = malloc(canvas_sz); + memset(canvas, gif->SBackGroundColor, canvas_sz); - demuxer->priv = gif; - return demuxer; } static void demux_close_gif(demuxer_t* demuxer) { - GifFileType *gif = (GifFileType *)demuxer->priv; + GifFileType *gif = gif_hndl; if(!gif) return; @@ -217,6 +295,8 @@ PrintGifError(); demuxer->stream->fd = 0; + if(demuxer->priv) + free(demuxer->priv); demuxer->priv = NULL; }

4 12

[RFC] 6-channel AAC and channel reordering
by Corey Hickey 09 Nov '06

09 Nov '06

The problem: 6-channel AAC has a different channel order than MPlayer outputs. MPlayer wants: 0 - front left 1 - front right 2 - rear left 3 - rear right 4 - front center 5 - lfe MPlayer gets: 0 - front center 1 - front left 2 - front right 3 - rear left 4 - rear right 5 - lfe -------------------------------------------------------------------------- For a test source you can probably use any recent hd trailer from Apple. Or: http://movies.apple.com/movies/universal/king_kong/king_kong-tlr_h480p.mov This works fine because AAC does the downmixing internally: $ mplayer king_kong-tlr_h480p.mov -channels 2 This sounds odd -- the channels are definitely mixed up: $ mplayer king_kong-tlr_h480p.mov -channels 6 If you don't have a 5.1 speaker setup (I don't either), then you can do a quick-n-dirty downmix: $ mplayer king_kong-tlr_h480p.mov -channels 6 \ -af pan=2:0.4:0:0:0.4:0.2:0:0:0.2:0.3:0.3:0.5:0.5 ...or an hrtf one: $ mplayer king_kong-tlr_h480p.mov -channels 6 -af hrtf Either way, you'll hear that the channel position isn't right. -------------------------------------------------------------------------- The attached patch uses af_channels to handle the reordering. ad_faad.c:init() sets up a routing map and af.c:af_init() uses that map to put a channels filter at the beginning of the audio chain. Try the above examples with a patched mplayer -- they should work correctly. As far as I know the patch doesn't break anything. Probably it does, but not that I know of. :) Anyway, I don't consider it finished. Some notes: * It would probably be ideal to have faad handle the reordering, but I don't see a way to do that. * If this kind of approach seems acceptable, I should switch to af_pan to handle upmixing and downmixing where channels have to be split or combined. * I don't have any code to deal with 3- or 4- channel AAC. With af_pan I could probably upmix or downmix these appropriately to match the output channel order. Does anyone have a sample? * This could be applied to audio formats other than AAC, of course. Does anyone have an example of another format that could benefit? -------------------------------------------------------------------------- So, what do you think? Am I barking up the wrong tree or should I keep working in this direction? -Corey Index: libaf/af.c =================================================================== RCS file: /cvsroot/mplayer/main/libaf/af.c,v retrieving revision 1.51 diff -u -r1.51 af.c --- libaf/af.c 20 Oct 2005 09:12:28 -0000 1.51 +++ libaf/af.c 22 Feb 2006 09:04:06 -0000 @@ -355,17 +355,40 @@ // Check if this is the first call if(!s->first){ - // Add all filters in the list (if there are any) - if(!s->cfg.list){ // To make automatic format conversion work - if(!af_append(s,s->first,"dummy")) - return -1; + /* remap channels if the codec/demuxer provides a channel map */ + if (s->chan_map) { + af_control_ext_t arg; + arg.ch = s->input.nch; + af_instance_t *af = NULL; + af=af_append(s, s->first, "channels"); + /* set up af_channels to route */ + if (!af || (AF_OK != af->control(af, AF_CONTROL_CHANNELS_ROUTER, &s->input.nch))) + return -1; + if (!af || (AF_OK != af->control(af, AF_CONTROL_CHANNELS_NR, &s->input.nch))) + return -1; + /* set up each route */ + for (arg.ch = 0; arg.ch < s->input.nch; ++arg.ch) { + arg.arg = s->chan_map + 2*arg.ch; /* two by two... */ + if (!af || (AF_OK != af->control(af, AF_CONTROL_CHANNELS_ROUTING, &arg))) + return -1; + } + if (!af || (AF_OK != af->control(af, AF_CONTROL_CHANNELS, &s->input.nch))) + return -1; + if (AF_OK != af_reinit(s,af)) + return -1; } - else{ + // Add all filters in the list (if there are any) + if (s->cfg.list) { while(s->cfg.list[i]){ if(!af_append(s,s->last,s->cfg.list[i++])) return -1; } } + // To make automatic format conversion work + if (!s->first){ + if(!af_append(s,s->first,"dummy")) + return -1; + } } // Init filters @@ -711,3 +734,23 @@ { data->bps = af_fmt2bits(data->format)/8; } + +int *af_set_channel_map(int channels, char *routes){ + int *chan_map, *ptr; + ptr=chan_map=malloc(2 * channels * sizeof(int)); + if (!chan_map) { + mp_msg(MSGT_DEMUX, MSGL_ERR, "set_channel_map: cannot malloc for %d channels\n", channels); + return 0; + } + while (channels-- > 0) { + if (*routes == '\0' || *(routes+1) == '\0') { + mp_msg(MSGT_DEMUX, MSGL_ERR, "set_channel_map: not enough routes\n"); + free(chan_map); + return 0; + } + /* two by two... */ + *ptr++ = *routes++ - '0'; + *ptr++ = *routes++ - '0'; + } + return chan_map; +} Index: libaf/af.h =================================================================== RCS file: /cvsroot/mplayer/main/libaf/af.h,v retrieving revision 1.28 diff -u -r1.28 af.h --- libaf/af.h 1 Oct 2005 12:55:34 -0000 1.28 +++ libaf/af.h 22 Feb 2006 09:04:06 -0000 @@ -104,6 +104,7 @@ af_data_t output; // Configuration for this stream af_cfg_t cfg; + int *chan_map; }af_stream_t; /********************************************* @@ -329,6 +330,16 @@ */ void af_fix_parameters(af_data_t *data); +/** + * \brief set up channel remapping + * \param number of output channels + * \param string of from-to routes + * \return int array of routes + * + * Makes an int array from a string of provided routes. + */ +int *af_set_channel_map(int channels, char *routes); + /** Memory reallocation macro: if a local buffer is used (i.e. if the filter doesn't operate on the incoming buffer this macro must be called to ensure the buffer is big enough. Index: libmpcodecs/ad_faad.c =================================================================== RCS file: /cvsroot/mplayer/main/libmpcodecs/ad_faad.c,v retrieving revision 1.24 diff -u -r1.24 ad_faad.c --- libmpcodecs/ad_faad.c 12 Jan 2006 20:04:34 -0000 1.24 +++ libmpcodecs/ad_faad.c 22 Feb 2006 09:04:06 -0000 @@ -154,6 +154,23 @@ mp_msg(MSGT_DECAUDIO,MSGL_V,"FAAD: Negotiated samplerate: %ldHz channels: %d\n", faac_samplerate, faac_channels); sh->channels = faac_channels; if (audio_output_channels <= 2) sh->channels = faac_channels > 1 ? 2 : 1; + + /* re-map channels */ + switch (sh->channels) { + default: + case 1: /* no action needed */ + case 2: /* no action needed */ + case 3: /* no suitable default behavior? */ + case 4: /* no suitable default behavior? */ + break; + case 5: /* mplayer treats this like 6-channel */ + case 6: + sh->chan_map = af_set_channel_map(6, "04" "10" "21" "32" "43" "55"); + break; + case 7: /* not supported by mplayer? */ + break; + } + sh->samplerate = faac_samplerate; sh->samplesize=2; //sh->o_bps = sh->samplesize*faac_channels*faac_samplerate; Index: libmpcodecs/dec_audio.c =================================================================== RCS file: /cvsroot/mplayer/main/libmpcodecs/dec_audio.c,v retrieving revision 1.50 diff -u -r1.50 dec_audio.c --- libmpcodecs/dec_audio.c 16 Feb 2006 22:43:42 -0000 1.50 +++ libmpcodecs/dec_audio.c 22 Feb 2006 09:04:06 -0000 @@ -309,6 +309,7 @@ // filter config: memcpy(&afs->cfg,&af_cfg,sizeof(af_cfg_t)); + afs->chan_map = sh_audio->chan_map; mp_msg(MSGT_DECAUDIO, MSGL_V, MSGTR_BuildingAudioFilterChain, afs->input.rate,afs->input.nch,af_fmt2str_short(afs->input.format), Index: libmpdemux/stheader.h =================================================================== RCS file: /cvsroot/mplayer/main/libmpdemux/stheader.h,v retrieving revision 1.42 diff -u -r1.42 stheader.h --- libmpdemux/stheader.h 18 Feb 2006 09:26:39 -0000 1.42 +++ libmpdemux/stheader.h 22 Feb 2006 09:04:07 -0000 @@ -18,6 +18,7 @@ int samplerate; int samplesize; int channels; + int *chan_map; int o_bps; // == samplerate*samplesize*channels (uncompr. bytes/sec) int i_bps; // == bitrate (compressed bytes/sec) // in buffers:

5 18