Re: [Freevo-users] Re: Better DVB support (C programmers needed)
On Wed 07.07.2004 at 02:45:55PM +0200, Dirk Meyer wrote: It would be nice if you could put all your dvb stuff on a website or something like that. I would like to take a look at it and maybe improve it. Maybe someone else, too. Here is a modified version of tzap that records one channel to either a file or Thomas' timeshift buffer (and also records subtitles if there are some: put the subtitle PID to ~/.tzap/channels.conf just after the audio PID; separate them with a :, as usual). http://www.mit.jyu.fi/~mweber/leffakone/software/dvbrecord/dvbrecord-1.0.tar.gz I tested the timeshift buffer, and it works well (except when it crashes while seeking, I don't know exactly why :) But I think that it would be better to use a plain file for timeshifting, since 256 MB of buffer gives a timeshift of about 7 minutes in DVB, which is not much. Since the thingy uses memory mapped IO, I don't know how it would perform if the buffer is much bigger than the available memory (e.g. set a buffer of 1GB when you have 256 MB or RAM); I don't know if RAM is actually used in the process, or if the file is simply mapped to virtual memory, but no real memory is used (except for normal disk buffering). The .tar.gz alos contains an attempt to modify ts2ps (from libmpegdvb) so that it takes into account the subtitle stream if there is one. It's not perfect yet, and it should be modified to accept several audio streams and several subtitle streams. Matthieu -- (~._.~)Matthieu Weber - Université de Jyväskylä (~._.~) ( ? )email : [EMAIL PROTECTED] ( ? ) ()- -() public key id : 452AE0AD ()- -() (_)-(_) Humor ist, wenn man trotzdem lacht (Germain Muller) (_)-(_) --- This SF.Net email sponsored by Black Hat Briefings Training. Attend Black Hat Briefings Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com ___ Freevo-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freevo-users
Re: [Freevo-users] Time Shifting AND commercial Breakdetection?
(This message was rejected by sourceforge yesterday, so I resend it) On Sat 10.07.2004 at 10:54:01AM -0400, Jason Tackaberry wrote: On Sat, 2004-07-10 at 10:00 +0200, Fons van der Beek wrote: 3. When the player reaches a black frame, the controlling (timeshift) process can tell it to seek to the appropriate place (either to the next black frame, or end of buffer, whichever is first), and since timeshift also knows what timestamps are where in the buffer, seeking will be a lot closer to the mark, since MPlayer's mpeg seeking is hopeless. MPlayer seeking can be fixed. I managed to get precise position seeking in MPEG TS yesterday (even thouhg the stream contains a PTS reset), and the same procedure can be applied to MPEG PS. It needs a bit fine tuning, but the main thing is there. If you can send me your latest black frame detection patch, I'll see about working this up. As soon as it's ready for release! Update: It's ready now: http://www.mit.jyu.fi/~mweber/leffakone/software/mplayer/MPlayer-1.0pre4+TS_seek+Black_frames_detection Matthieu -- (~._.~)Matthieu Weber - Université de Jyväskylä (~._.~) ( ? )email : [EMAIL PROTECTED] ( ? ) ()- -() public key id : 452AE0AD ()- -() (_)-(_) Humor ist, wenn man trotzdem lacht (Germain Muller) (_)-(_) --- This SF.Net email sponsored by Black Hat Briefings Training. Attend Black Hat Briefings Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com ___ Freevo-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freevo-users
Re: [Freevo-users] Commercial break detection
(This message was rejected by sourceforge yesterday, I resend it) On Sat 10.07.2004 at 01:33:03AM +0200, Fons van der Beek wrote: Hi all, I did some work on detection of commercial breaks: i tried several methods: - Black frame detection from Matthieu Weber, i tried this one on recordings from dutch channels, it worked EXTREMELY well: on 6 hours of recorded shows not one error, a smooth detection!, it takes some time, but what the heck, that's why the nice command was invented. -Logo detection, i created a program which detects a Logo transmitted by the serviceprovider, some screenshots of detected logo's are available on http://213.10.252.115/commercial.pdf .but as i watched the shows, i noticed that even when te logo detection would be 100% error free: 1) The service providers blend in logos seconds after the show starts! or the service providers starts with an animated logo, after seconds it went static 2) The service providers use logo's WITHIN commercials to promote their own marchandise or show previews of upcoming shows. The only time logo detection would be usefull is when it can be used to forecast if a given period is a movie: suppose the minimum commercial length = 3 minutes, take a shot each three minutes, if on consequent shots there is a logo, then the timespan between the shots is 99,9% for certain not a commercial. (Taking a screenshot every 3 minutes takes allmost no time (1 hour movie, 2-3 minutes)) black frame detection takes more time (of course depending on cpu speed) We could sample the recording and find areas with a higher probability of commercials, and then use the black frame detection to find the precise location of the commercial breaks. If you can send me the code of your logo detection, I could merge it with my black frame detection filter. -Soundlevel, the simple method I used was not usable and i doubt if this would ever work (a war scene might be detected as commercial. if a more suffisticated method would work, i dont know. it's also more time/cpu consuming as i thought. I'm no expert in sound processing, but I suppose that measuring simply the soundlevel is not enough, you must also (or rather ?) measure the variation of the sound around the average level (some kind of standard deviation of the sound level). My guess is that during the commercials, the standard deviation is much lower than during the show (i.e. the sound is equally loud all the time, whereas during the show it is varying much more). But what about talk-shows? http://toonarchive.com/karin/ uses all 3 methods and merges the results in order to determine where the commercials are. It's for Windows, but the software is GPL, is someone cares to take a look at it and take the algorithms out... Matthieu -- (~._.~)Matthieu Weber - Université de Jyväskylä (~._.~) ( ? )email : [EMAIL PROTECTED] ( ? ) ()- -() public key id : 452AE0AD ()- -() (_)-(_) Humor ist, wenn man trotzdem lacht (Germain Muller) (_)-(_) --- This SF.Net email sponsored by Black Hat Briefings Training. Attend Black Hat Briefings Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com ___ Freevo-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freevo-users
RE: [Freevo-users] Blue screen with mplayer when rebooting.
On Mon, 12 Jul 2004, Jon Morris wrote: I have a small issue with Mplayer in Freevo. When I first reboot, and try to use Mplayer to either watch TV or record a show, the screen is only blue. If I load tvtime and exit before loading freevo, mplayer works just fine. Thanks, I just checked the lsmod and it is in there: Module Size Used by snd_pcm_oss40740 0 tuner 15380 0 tda9887 6020 0 bttv 129292 0 Hmm, any clues in the dmesg related to bttv? What distro kernel version are you using? I see v4l2 enabled. Try using TV_DRIVER='v4l2' (redefine in /etc/freevo/local_config.py). V4l2-based drivers have some issues when called using v4l interface (mplayer default). Check also the freevo.conf file in /etc/freevo that the norm and chanlist values are correct (e.g., ntsc/pal, us-cable/europe-west, etc.) Also, what was the command sent to mplayer? I assume it's using the TV input. A blue screen suggests that mplayer wasn't able to tune into the channel correctly the first time. T.C. Wan Tat Chee (Lecturer) School of Computer Science, Univ. of Science Malaysia, 11800 USM, Penang, Malaysia. Rm.625 Ofc Ph: +604 653-3888 x 3617 NRG Lab Admin: +604 659-4757 Rm.601-E Ofc Ph: +604 653-4396 Internet: [EMAIL PROTECTED]Web: http://nrg.cs.usm.my/~tcwan GPG Key : http://nrg.cs.usm.my/~tcwan/tcw_gpg-20030322.asc F'print : DCF2 B9B2 FA4D 1208 AD59 14CA 9A8F F54D B2C4 63C7 --- This SF.Net email sponsored by Black Hat Briefings Training. Attend Black Hat Briefings Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com ___ Freevo-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freevo-users
[Freevo-users] Re: Better DVB support (C programmers needed)
Matthieu Weber wrote: On Mon 12.07.2004 at 10:04:10PM +0200, Dirk Meyer wrote: Matthieu Weber wrote: http://www.mit.jyu.fi/~mweber/leffakone/software/mplayer/mplayer-commercial_removal.tar.gz Hi, it's not working for me. The detection looks good, I have an edl with the correct timestamps. Maybe...they seem to start at 0, but the recording isn't. Also mplayer doesn't skip the commercials. I used it with the -edl parameter of mplayer. Anything else? This older version would not work if the recording doesn't start at 0 (i.e. if you are using e.g. a TS from DVB). Use the newest version: http://www.mit.jyu.fi/~mweber/leffakone/software/mplayer/MPlayer-1.0pre4+TS_seek+Black_frames_detection Apply the patch with patch -p1 MPlayer-1.0pre4+TS_seek+Black_frames_detection from within the MPlayer-1.0pre4 directory. This patch must be applied to MPlayer sources which were *NOT* patched with the older patch. The edl is not written. I see some output of your plugin, but there is no edl :-( Dischi -- And now for something completely different. --- This SF.Net email sponsored by Black Hat Briefings Training. Attend Black Hat Briefings Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com ___ Freevo-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freevo-users
[Freevo-users] Re: Better DVB support (C programmers needed)
Matthieu Weber wrote: I tested the timeshift buffer, and it works well (except when it crashes while seeking, I don't know exactly why :) I maybe know, more later in this mail. But I think that it would be better to use a plain file for timeshifting, since 256 MB of buffer gives a timeshift of about 7 minutes in DVB, which is not much. OK, there is a small plugin in WIP/Dischi called mplayer_dvb_timeshift.py. Put it into src/tv/plugins, activate it and remove the old tv.mplayer plugin. It is a mix between video.player and tv.mplayer, only using dvb. It calls mplayer dumpstream to dump the stream and than mplayer to play it (after 1 MB buffer to be save). It works good, you can seek and pause. One problem: you may not seek forward. I use mmpython to get the maximum position to seek to to avoid seeking behind the stream. But your seek patch has a small big which may also cause Thomas ringbuffer to crash: first mplayer seeks and after that you correct the position (I guess). That is bad, because mplayer seeks after the end and you get no chance to correct it. So what we need here is a seek patch that seeks correct in the first place. Later we could make it possible to define the dump program (making all dump-to-mpeg programs working) and maybe also the show program, make it possible to choose between xine and mplayer. Try it, Dischi -- Software isn't released, it's allowed to escape. --- This SF.Net email sponsored by Black Hat Briefings Training. Attend Black Hat Briefings Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com ___ Freevo-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freevo-users
[Freevo-users] recording server
Where is recording server file located? --- This SF.Net email sponsored by Black Hat Briefings Training. Attend Black Hat Briefings Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com ___ Freevo-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freevo-users
RE: [Freevo-users] Blue screen with mplayer when rebooting.
-Original Message- From: [EMAIL PROTECTED] on behalf of Wan Tat Chee Sent: Tue 7/13/2004 1:20 AM To: [EMAIL PROTECTED] Cc: Subject: RE: [Freevo-users] Blue screen with mplayer when rebooting. Hmm, any clues in the dmesg related to bttv? What distro kernel version are you using? I see v4l2 enabled. Try using TV_DRIVER='v4l2' (redefine in /etc/freevo/local_config.py). V4l2-based drivers have some issues when called using v4l interface (mplayer default). Check also the freevo.conf file in /etc/freevo that the norm and chanlist values are correct (e.g., ntsc/pal, us-cable/europe-west, etc.) Also, what was the command sent to mplayer? I assume it's using the TV input. A blue screen suggests that mplayer wasn't able to tune into the channel correctly the first time. T.C. Wan Tat Chee (Lecturer) School of Computer Science, Univ. of Science Malaysia, 11800 USM, Penang, Malaysia. Rm.625 Ofc Ph: +604 653-3888 x 3617 NRG Lab Admin: +604 659-4757 Rm.601-E Ofc Ph: +604 653-4396 Internet: [EMAIL PROTECTED]Web: http://nrg.cs.usm.my/~tcwan GPG Key : http://nrg.cs.usm.my/~tcwan/tcw_gpg-20030322.asc F'print : DCF2 B9B2 FA4D 1208 AD59 14CA 9A8F F54D B2C4 63C7 -- There are a few errors in the dmesg output, but nothing that stands out as a problem to me: Linux video capture interface: v1.00 bttv: driver version 0.9.14 loaded bttv: using 8 buffers with 2080k (520 pages) each for capture bttv: Bt8xx card found (0). bttv0: Bt878 (rev 17) at :00:08.0, irq: 11, latency: 32, mmio: 0xde00 bttv0: detected: Pinnacle PCTV [card=39], PCI subsystem ID is 11bd:0012 bttv0: using: Pinnacle PCTV Studio/Rave [card=39,insmod option] bttv0: gpio: en=, out= in=00ffefff [init] bttv0: i2c: checking for MSP34xx @ 0x80... not found bttv0: pinnacle/mt: id=5 info=NTSC / mono radio=no bttv0: using tuner=33 bttv0: i2c: checking for MSP34xx @ 0x80... not found bttv0: i2c: checking for TDA9875 @ 0xb0... not found bttv0: i2c: checking for TDA7432 @ 0x8a... not found tda9887: chip found @ 0x86 tuner: chip found at addr 0xc0 i2c-bus bt878 #0 [sw] tuner: type set to 33 (MT20xx universal) by bt878 #0 [sw] tuner: microtune: companycode=3cbf part=42 rev=99 tuner: microtune MT2050 found, OK bttv0: registered device video0 bttv0: registered device vbi0 bttv0: PLL: 28636363 = 35468950 .. ok bttv0: PLL can sleep, using XTAL (28636363). I am using Fedorca Core 2, Kernel 2.6.5-1.358. I am using the v4l2 driver in Freevo. I have all the settings in the local_config.py set to ntsc and us-cable... Here is the command sent to mplayer when the recordserver kicks off: [*RecordServer*] popen /usr/bin/mencoder tv:// -tv driver=v4l2:input=0:norm=ntsc:channel=1:chanlist= us-cable:width=640:height=480:outfmt=yuy2:device=/dev/video0:adevice=/dev/dsp:audiorate=32000:forceaudio:forcechan=1:buffe rsize=64 -ovc lavc -lavcopts vcodec=mpeg4:vbitrate=1200:keyint=30 -oac mp3lame -lameopts br=128:cbr:mode=3 -ffourcc divx - endpos 147 -o /usr/share/freevo/record/07-13_06:20_Manual_Record.avi I am currently setup to use tvtime for watching TV in Freevo. I found that if I try to watch TV using tvtime in Freevo, I also get a blue screen, and the message No Signal. If I run tvtime from the shell, I get a picture. When I exit, both tvtime and mplayer work fine in Freevo... Thanks again for the help... It is greatly appreciated. Jon
