Back to Home / #mythtv / 2003 / 08 / Prev Day | Next Day
#mythtv IRC Logs for 2003-08-22

---Logopened Fri Aug 22 00:00:03 2003
00:20-!-just1nux [] has joined #mythtv
00:20<just1nux>chutt: you around?
00:22<just1nux>cool. do you use the playback offsets?
00:22<Chutt>the over/underscan compensation?
00:23<Chutt>on my dev machine i do
00:23<just1nux>what do you have them set to?
00:23<Chutt>5% each, no x/y offset
00:23<just1nux>odd numbers display fine, but even im getting real messed up shifted colors and a line at the bottom
00:23<Chutt>xv driver bug
00:24<Chutt>doesn't scale/crop properly
00:24<just1nux>it was driving me nuts and i never saw any posts on the lists about it
00:24<Chutt>well, it almost does -- it's using uninitialized ram
00:24<just1nux>i always had 2 on there and it was fine, but now its not
00:24<Chutt>what video card?
00:24<Chutt>that was mostly fixed in newer nvidia drivers
00:25<Chutt>was a problem before
00:25<just1nux>NVidia GeForceMX 440 i tihnk
00:25<Chutt>make sure you're using the next to newest driver
00:25<just1nux>i havent updated my nvidia drive on the machine in a while
00:25-!-brad_ [] has joined #mythtv
00:26<Chutt>get 1.0-4363
00:26<just1nux>ok, ill do that, for now i just uped it to 3 and all is well
00:26<Chutt>the newest version breaks the experimental a/v sync stuff
00:26-!-hfb [] has joined #mythtv
00:27<just1nux>i never had that turned on, what does that do?
00:28<just1nux>oh, another thought about the OSD, for the "O" option (which i really like btw) some arrows to show that more can be crolled through horizontally might be nice.
00:29<just1nux>not sure if vertical ones are neccessary though.
00:30<Chutt>the experimental a/v sync makes the video smoother on nvidia cards
00:30<mikegrb>Chutt: re the O option, would it be difficult to have the frontend query the current time, all channels when you enter browse mode? On a remote frontend every time I hit the channel up/down the video freezes for several secs
00:31<mikegrb>alternatively would increasing some buffer values for the live tv help out with this?
00:31<just1nux>yeah that bothers me to, is it because of loading the icons?
00:31<just1nux>seems like
00:31<mikegrb>just1nux: I haven't thought of that I might try turning them off... if they are off for the guide will they be off there too?
00:32<just1nux>the word "experimental" in the experimental a/v scared me off. Myth is experimental enough.
00:32<mikegrb>I'd check real quick but my xbox is compiling mythgame
00:32<mikegrb>just1nux: it uses nvidia specific stuff for smother video
00:32<mikegrb>er smoother
00:33<mikegrb>made a big difference for me
00:33<just1nux>mikegrb, i only get delays in the video when changing channels when it needs a new icon, when switching channels that use the same icon, no delay.
00:33<just1nux>meant to ask chutt about that
00:34<just1nux>also meant to ask why it still shows the icon of the previous channel if the new channel has no icon
00:34<mikegrb>hmm I'll have to try turning them off in the guide though I like them, it would prob be much easier to cache the images, perhaps the ability for frontend setup to specify a local path for icons
00:34<just1nux>they are cached
00:34<mikegrb>well I meant all the time <g>
00:34<just1nux>in ~/.mythtv/icons/ or something
00:35<mikegrb>just1nux: all I've got is the theme cache
00:36<mikegrb>on the frontend that is on the same machine as the backend they are there though <g>
00:36<just1nux>must be served by the backend...cause of the delay?
00:37<mikegrb>yes, that's what I'm thinking
00:37<mikegrb>noce theme by the way, I think I've told you before but just wanted to make sure ;)
00:38<mikegrb>er nice
00:38<mikegrb>I would look into that now but I actually have to work tommorow, then drive a few hrs when I get off <g>
00:38<just1nux>itmust use the ones form the backend regardless of if the frontend has them, cuz my frontend machine has them (it used to be the backend) and i still get the loading delay
00:39<mikegrb>once this compile is done I'll try turning them off for the guide, see if that works as a temp fix
00:39<mikegrb>when I get back from myrtle beach that will probably be the first thing I look into
00:42<just1nux>looks like that checkbox has no effect on the OSD, mines not checked, yet the icons are showing
00:45<mikegrb>here's what I'm thinking, A setting to specify a local directory (setup -> tv -> general or maybe playback) where channel images can be found, if dir not there or not defined acts as it does now, if the directory exists, it will check there for the image and not even mess with asking the backend
00:45<mikegrb>it would require manually placing the images there
00:45<Chutt>just copy your ~/.mythtv/icons
00:45<Chutt>or whatever it is
00:45<just1nux>true, im trying my theme, minus the icons
00:46<mikegrb>that will work chutt? :-)
00:46<just1nux>it doesnt find them though.
00:46<Chutt>sure it will
00:46<Chutt>it will if you have the same username on both machines
00:47<just1nux>for the record, im using the same username on both machines, i have the same icon folder with all the icons.
00:47<Chutt>that should work fine, then
00:47<mikegrb>same username running the frontend or same username for the backend and frontend?
00:47<just1nux>theme minus icons = no delay...just tested it
00:49-!-brewt [] has joined #mythtv
00:49-!-mechou [] has joined #mythtv
00:49<just1nux>my frontend machine used to be the backend machine, so its all right where it should be.
00:49<mikegrb>thanks chutt that did it for me
00:50<just1nux>wait lemme think maybe its not the same username
00:50<brewt>hey, anyone have a clue on how to fix this error i get with the frontend?
00:50<brewt>QTime::setHMS Invalid time 70:00:00.000
00:50<mikegrb>just1nux: my mythfilldatabase/backend runs as root... frontend on same machine = dif user... remote frontend (xbox) runs frontend as root, coppied icons to /root/.mythtv/channels on xbox and :))
00:50<brewt>QDate::setYMD: Invalid date 2003-82-01
00:51<brewt>it's stored in the database properly, but it seems to be parsing it out wrong
00:51<mikegrb>this should fix the problem mechou was having too
00:51<mikegrb>mechou: you around?
00:51<mechou>yup, just signed on.
00:52<just1nux>YUP that was it
00:52<mechou>what were you guys talking about
00:52<just1nux>i dont wanna run my frontend as root
00:54<mechou>mikegrb, just signed on a minute or so ago. Are you saying I ought to checkout cvs?
00:54<mikegrb>mechou: the hitting i twice when watching a recording issue
00:54<mikegrb>mechou: the delay is from the channel icon
00:55<mechou>yeah, I figured that, but I saw the word "fixed" somewhere :)?
00:55<mikegrb>mechou: if frontend is running as same user name as the backend you can cp the ~/.mythtv/channels dir to the remote frontend
00:55<mikegrb>makes the browsing during live tv work smoothly too
00:57<mechou>mikegrb, cool. But this is only a temp workaround, right?
00:57<mikegrb>well I dunno, it fixes the problem
00:57<markw>anyone see problems with xmltv and dtv?
00:58<markw>it's segfaulting here.
00:58<markw>I can select the cable company and it works, but dtv segfaults.
00:59-!-Kuwanger [] has joined #mythtv
00:59<mechou>mikegrb, I don't want to sound ungrateful, but that basically means for networked mythtv, you got to sync ~/mythv (or have it mounted under NFS or something....)
01:00<Kuwanger>During record, is the live buffer tooken into consideration for space measuring? I've got under 1GB free space on my backend's recording partition and mythbackend failed to record some shows.
01:00* markw wonders how much traffic would be generated on the network.
01:05-!-just1nux [] has quit ["later"]
01:15<mikegrb>mechou: well it shouldn't need to be done very often
01:16* Kuwanger hopes that it was a space issue. :/
01:16-!-Kuwanger [] has left #mythtv []
01:18-!-tmk [] has joined #mythtv
01:24-!-brainless [] has quit [Read error: 104 (Connection reset by peer)]
01:24<bline>hey tmk
01:29-!-brewt [] has quit ["[BX] Chester Cheeta uses BitchX. Ayeuhayueuhayueuh!"]
01:34<tmk>ima go watch me some mythTV
01:46<Morph>wonder why im not getting video from mythtv..
01:55<Chutt>what's it do?
01:56<Chutt>ye olde black screen?
01:56<Morph>thats the one.
01:56<Morph>I wonder if im using the wrong msp3400 module..
01:56<Chutt>any error messages on the console?
01:56<Morph>dont believe so. lemme double check.
02:01<Chutt>sure it's on the right input? =)
02:01<Morph>think so.
02:01* Morph checks.
02:01<Chutt>i'm too tired to think of any other cause of it
02:01<Morph>heh. understood.
02:08<Morph>doesnt appear to be the wrong input..
02:08<Morph>when the black screen pops up, I cant do anything.
02:11<Chutt>are you absolutely sure there's no errors from either mythbackend or the frontend? =)
02:11-!-mecraw [] has joined #mythtv
02:11<Chutt>that's really what most people who have that problem get
02:11<Morph>ya know. im gonna check mythbackend..this happened to me before :)
02:11<Chutt>if you're still having problems tomorrow, i'll try to debug more then
02:12<Chutt>gotta sleep
02:12-!-mechou [] has quit [Read error: 110 (Connection timed out)]
02:12<Morph>Nod. thanks. later.
02:17<Morph>heh. it worked when I start it by hand.
02:28<markw>what's the recommended mythtv wm?
02:30<Timon>markw: Read the docs
02:31-!-Drikus [] has joined #mythtv
02:32-!-hfb is now known as hfb_zzzzzz
02:32<markw>Timon: I have and it refers to kde that comes with redhat.
02:32<Morph>the docs I read mention fvwm..
02:33<Morph>dont rememeber seeing a recommendation on kde.
02:33<markw>Morph: I remember seeing something about fvwm
02:33<markw>section 5 or so has the redhat kde thing, has to do with qt
02:34<Morph>thats probably related to building
02:34<bline>windowmaker plays well with myth
02:34<Morph>openbox too.
02:40<markw>no configfiles dir. :)
02:40<markw>which has the .fvwmrc and twmrc files.
02:50-!-Markos [] has quit ["Client exiting"]
02:55<Morph>anyone using a pvr350?
02:59<markw>mecraw: thanks, already found it.
02:59<markw>Morph: configing a 250 right now.
03:01<mecraw>markw: fyi, i had focus problems with fvwm and mythtv when i tried it, but i'm no fvwm expert
03:01<markw>mecraw: my buddy is using fvwm
03:01<markw>seems to work fine, he has everything running, lirc, pvr-250, auto boot/login xmltv...
03:01<markw>he sucks. :)
03:02<mecraw>that bastard
03:29<markw>got tv...
03:29* paulproteus doesn't use a window manager with Myth
03:37* Morph watches Contact via mythvideo
03:37<Morph>any of you guys using an epia board?
03:43<tmk>morph: i've got a 350
03:44<Morph>tmk: heh. I was gonna ask how support for the mpeg decoder was..I suppose you would be the one to ask :)
03:44<tmk>it works using the decoder-alpha.tar.gz
03:44<tmk>but you have to do it manually
03:44<tmk>ie not in myth
03:44<tmk>i'm working on more robust support now
03:44<tmk>also currently you can't decode while encoding
03:45<Morph>Nod. thats about where I thought it was.
03:45<Morph>btw, how was your trip to africa?
03:45<tmk>but the output works great
03:45<tmk>i was just watching Man. U vs bolton
03:45<tmk>top notch quality
03:45<tmk>africa was awesome
03:45<Morph>thats great :)
03:45<Morph>on both accounts.
03:46<tmk>helped build an orphanage/did puppet shows for some kids
03:46<tmk>worked on some linux stuff for the boat
03:46<Morph>thats cool.
03:46<tmk>but now i must sleep
03:46-!-tmk [] has quit []
04:47-!-linagee_ [] has joined #mythtv
04:47-!-linagee [] has quit [Connection reset by peer]
04:50-!-Ripp [] has quit [Read error: 110 (Connection timed out)]
05:08-!-FryGuy [] has quit [Read error: 104 (Connection reset by peer)]
05:38-!-bline [] has quit ["Client exiting"]
05:44-!-radsaq [] has quit [Read error: 104 (Connection reset by peer)]
05:44-!-radsaq [] has joined #mythtv
06:04-!-bline [] has joined #mythtv
07:38-!-o_cee\sleep [] has quit [Ping timeout: 14400 seconds]
08:15-!-jkolb [] has joined #mythtv
08:28-!-Rule [] has joined #mythtv
08:55-!-Rule [] has quit ["bye"]
09:20-!-Ripp [] has joined #mythtv
09:20-!-Markos [] has joined #mythtv
10:02<rkulagow>chutt, are you here?
10:02-!-jkolb [] has quit [Read error: 104 (Connection reset by peer)]
10:07<rkulagow>mdz / chutt: i just committed a boo-boo; didn't carefully look at what was happening and committed my Makefile along with an updated keys.txt. How do you roll-back the change in a commit with CVS?
10:07<mdz>rkulagow: I'll fix it
10:07<rkulagow>thanks. sorry about that.
10:08<mdz>shouldn't matter, really
10:08<mdz>qmake will regenerate it
10:08<mdz>Chutt commits his makefile all the time :-)
10:09<mdz>I usually either do this, to revert a specific change:
10:09<mdz>cvs diff -u -r1.31 -r1.30 Makefile |patch
10:09<mdz>or if you just want to completely replace with an earlier version, do this:
10:09<mdz>cvs update -p -r1.30 Makefile > Makefile
10:20<rkulagow>mdz: thanks.
10:35-!-mecraw [] has quit [Read error: 104 (Connection reset by peer)]
10:45-!-jkolb [] has joined #mythtv
10:51-!-hfb_zzzzzz [] has quit ["Client exiting"]
11:14-!-schultmc [] has joined #mythtv
11:23-!-hfb [] has joined #mythtv
12:04-!-mecraw [~mecraw@] has joined #mythtv
12:04-!-jkolb [] has quit [Read error: 104 (Connection reset by peer)]
12:07<ByteNik>Who was the one two days ago talking about making a MythTV bootsplash?
12:15<mecraw> ByteNik: it was o_cee (
12:15<ByteNik>k, thanks
12:16-!-thor_ [1000@] has joined #mythtv
12:22-!-Rule [] has joined #mythtv
12:26-!-dopez [] has joined #mythtv
12:28-!-jkolb [] has joined #mythtv
12:28-!-thor_ [1000@] has quit ["using sirc version 2.211+KSIRC/1.2.1"]
12:46-!-ByteNik [] has quit [Read error: 104 (Connection reset by peer)]
12:58-!-FryGuy [] has joined #mythtv
13:07-!-bbeattie [] has joined #mythtv
13:10<bbeattie>Chutt: has anyone brought up the idea of having a myth box going into sleep mode between recordings?
13:16-!-sfr [] has joined #mythtv
13:19-!-brainless [] has joined #mythtv
13:19<brainless>where can i find a doc for making themes for mythtv?
13:20<brainless>i am trying to make one
13:20<sfr>brainless there's a site describing it, search the gossamer ml archive
13:27<brainless>sfr: thx. i got the ui.xml file
13:31<FryGuy>bbeattie: yes, someone has already done that
13:32<bbeattie>FryGuy: .. got some info on it? (I've never seen the options in myth)
13:33<FryGuy>it's a patch AFAIK
13:33<FryGuy>i'm trying to find it for you
13:36-!-tmk [] has joined #mythtv
13:39<FryGuy>the patch is in the middle
13:40<FryGuy>it might be old though, so YMMV
13:40<bbeattie>.. any reason it isn't in myth now?
13:40<bbeattie>why still a patch?
13:41<FryGuy>read the thread :)
13:43<FryGuy>here's another
13:43<FryGuy>newer I think
13:44<bbeattie>so far it just appears that someone hasn'
13:44<bbeattie>hasn't created a patch that works correctly, but no reason it shouldn't that I can see yet
13:49-!-Captain_Murdoch_ [] has joined #mythtv
13:54-!-StarHeart [] has joined #mythtv
14:03-!-bbeattie [] has quit ["Client Exiting"]
14:10-!-bbeattie [] has joined #mythtv
14:13<rkulagow>bbeattie: does the pchdtv board have the equiv. of btaudio support on the analog side?
14:13-!-dopez [] has quit [Remote closed the connection]
14:17<bbeattie>erm, well,
14:17<bbeattie>I think.
14:18<rkulagow>do you have physical posession of a card?
14:18<rkulagow>does the PCI edge connector have one notch in it, or two?
14:18<bbeattie>I don't have the btaudio driver loaded, and I use the card fine
14:20<bbeattie>I'm saying 1, but look at, there is a picture of the card there
14:20<Morph>anyone tried this viaexp thing?
14:21<rkulagow>right, but i was just wondering if what's on the web site is what's shipping. retail stuff often says stuff like "board appearance may change", so the picture isn't always accurate.
14:21-!-brainless [] has quit []
14:21<bbeattie>it's exact.
14:21<Chutt>morph, i don't run one of the distributions they provide kernel modules for
14:22<Chutt>i was _trying_ to get it going a couple nights ago
14:22<rkulagow>bbeattie: ok, thanks. can't use it then.
14:22<bbeattie>too bad, how come?
14:24<rkulagow>the slots that i have are 3.3V, which means that they need the PCI to be notched near the metal bracket, vs. 5V cards which are notched at the other end of the connector. oh well, i'm sure that i could cobble together a PC if i had to, but i think there are only a few FTA HDTV broadcasters around here.
14:24<Morph>Chutt: Yeah. I noticed they only even seem to wanna build for RH and MDK.
14:24<Morph>Chutt: apparently someone on the via forums is close to getting the CLE266 decoder working.
14:26-!-dopez [] has joined #mythtv
14:26<Chutt>no, he's not really :p
14:26<Morph>well :) Closer than anyone else. :)
14:26<bbeattie>the card is PCI v2.2, not sure if that helps any.
14:26<Chutt>he's done a decompile of the library
14:27<Chutt>which doesn't really do very much
14:27<Chutt>as that seems to just talk to the kernel modules
14:28<Chutt>you've had problems with dvd playback on the m10k?
14:28<Chutt>i wouldn't think it'd be that bad
14:28<Morph>its not unplayable. just a bit jerky.
14:29<Chutt>those cpus are really too slow
14:31<Chutt>i need to find a new power supply and/or case for my mini-itx
14:31<Chutt>well, for my existing one, at least
14:31<Morph>Nod finding a decent power supply for those is a challenge.
14:32<rkulagow>chutt: why don't you just wait until they send you one for winning that contest?
14:32<tmk>bah. i'm going to win!
14:32<Chutt>the power supply fan is getting louder and louder
14:33<tmk>is it heat-triggered?
14:33<rkulagow>tmk: you should be in ivtv-devel fixing my compile problems, not goofing off in #mythtv. :)
14:33<jkolb>I need to get new fans for my case & ps. They're both really loud.
14:34-!-Ripp [] has quit [Connection timed out]
14:35<tmk>what problems are these? i just joined
14:35<rkulagow>oh, sorry. see you there in a minute then.
14:35<Chutt>someone just emailed me, lemme summarize
14:36<Chutt>guy's trying to convince his parents to let him install mythtv instead of a tivo
14:37<Morph>what are his arguments?
14:37<Chutt>"So, to help me out, what can I tell my family about the MythTV stability, update process, and defect process? And, and type of testing is done on MythTV? Is everything tested thoroughly and using what guidelines?"
14:38<Chutt>yet another email i'll be ignoring
14:38<tmk> /dev/null just got a little fuller
14:38<Morph>translation: ``Can my family blame you when their recording of Friends doesnt work?``
14:39<Chutt>tmk, that's the beauty of /dev/null, it just eats and eats and eats and never gets full =)
14:39<Morph>the carney wilson of the file system.
14:40<tmk>i wonder if dev/null's code is just
14:41<tmk>size_t read(struct file_p filp, const char *buf, size_t count) {return count;}
14:44-!-o_cee [] has joined #mythtv
14:44* o_cee can't compile a fresh cvs of mythvideo..
14:45<o_cee>i did make clean and make distclean.. is there anything more i can do?
14:45<Morph>why wont it compile? do you get errors?
14:46<Morph>would you like to share them? :)
14:46<o_cee>:) let's spam a little
14:46<o_cee>g++ -c -pipe -Wall -W -O3 -march=pentium3 -fomit-frame-pointer -D_REENTRANT -fPIC -D_GNU_SOURCE -DPREFIX=\"/usr/
14:46<o_cee>NO_DEBUG -DQT_THREAD_SUPPORT -DQT_PLUGIN -I/usr/qt/3/mkspecs/linux-g++ -I. -I/usr/local/include -I/usr/qt/3/inclu
14:46<o_cee>oc.o fileassoc.cpp
14:46<o_cee>UIType*MythThemedDialog::widget_with_current_focus' is private
14:46<o_cee>fileassoc.cpp:346: within this context
14:46<o_cee>UIType*MythThemedDialog::widget_with_current_focus' is private
14:46<o_cee>fileassoc.cpp:351: within this context
14:46<o_cee>UIType*MythThemedDialog::widget_with_current_focus' is private
14:46<Chutt>you don't have a new enough version of mythtv installed.
14:46<o_cee>fileassoc.cpp:576: within this context
14:46<o_cee>make[1]: *** [fileassoc.o] Error 1
14:46<o_cee>make[1]: Leaving directory `/home/mythtv/mythtvcvs/mythvideo/mythvideo'
14:46<o_cee>make: *** [sub-mythvideo] Error 2
14:47<o_cee>aaah, installed, no. haven't compiled it yet :)
14:47<Morph>well there ya go.
14:47<o_cee>i'll start off with it then
14:47<o_cee>another thing while i'm at it. is xvmc worth trying yet?
14:47<Morph>never tried it.
14:47<Chutt>not really
14:47<Chutt>no OSD
14:48<o_cee>oh, okay.. what's it supposed to do anyway?
14:48<Chutt>some small amounts of hardware accellerated mpeg2 decoding
14:48<o_cee>hm, okay.. thought it was more exciting :)
14:48<rkulagow>found out that if you use native lirc support in the main myth compile, you've got to enable it in at least mythdvd in _its_ as well.
14:49<Chutt>rkulagow, just in mythdvd's
14:50<o_cee>also another problem.. i'm using gentoo. the alsasound script doesn't startup snd-mixer/pcm-oss.. don't know why, and if i put them in modules.autoload, they'll get loaded before alsasound and alsasound will not start because of that... anyone who's got any suggestions for that?
14:50<Morph>speaking of mythvideo, do the standard myth volume controls work in mythvideo too?
14:50<Chutt>not unless you set up whatever you're using to playback with to use those keys
14:50<rkulagow>chutt: yeah, but i just figured it was inherited. i suppose the other modules would need to be modified for native lirc also?
14:50<Morph>thought so.
14:50<Chutt>rkulagow, nope -- mythdvd is the only other module that makes an executible
14:51-!-zzo [] has joined #mythtv
14:52<zzo>running myth in dual-head NOT xinerama with 2 different resoltuions - 800x600 & 1024x768 - myth runs fine on the 800x600 but just shows a matrix of red dots on the 1024x768 screen (this happens in xawtv & all over 'video' apps as well)
14:52<Chutt>someone needs to tell colin armstrong to use windows
14:52<Chutt>he's not ready for linux.
14:52<zzo>if i just run the 1024x768 screen it looks fine - only when running both I cannot run myth on the 1024x768
14:53<Chutt>zzo, if it happens in other apps, then it's nothing to do with mythtv.
14:53<zzo>agreed, but do you know if it's possible to run with that configuration - it's not a myth-specific problem
14:53<Chutt>ask whoever writes your video card drivers.
14:54<zzo>can Xv or whatever only run on 1 screen at a time?
14:56<o_cee>i've got the bootsplash up and running now.. at least the first version. is that something you'd like to put on Chutt?
14:56<Chutt>send it to the list first.
14:57<Morph>damnit. why cant mythtv play scrambled channels yet.
14:57<tmk>Chutt: yeah colin's not getting it
14:58<bbeattie>zzo: get your card yet?
15:01-!-ByteNik [] has joined #mythtv
15:07<o_cee>Chutt: okay. will do.
15:07<o_cee>is it ok to send such big files? haven't got anywhere to put them..
15:07<o_cee>probably less when packed
15:08<tmk>oc: email a link
15:08<o_cee>i haven't got anywhere to put them.
15:08<tmk>heh missed htat
15:08<jkolb>What's 307k?
15:08<tmk>you don't have dsl?
15:08<o_cee>the bootsplashs
15:08<jkolb>What format are they?
15:08<o_cee>but oh, btw. there's something i need to fix up first.
15:08<o_cee>and one cfg
15:09<jkolb>How many @ what res?
15:09<o_cee>silent and verbose @ 800x600
15:09<o_cee>based on the blue theme
15:09<jkolb>That seems large for two 800x600 jpegs.
15:10<Morph>if you want, you can email them to me and I'll put them up on my LUG server.
15:10<o_cee>highes quality, probably not necesarry
15:10<o_cee>sure, we could do that.. later tonight? i need to tidy up the cfg and add a little something to one of the pics
15:11<o_cee>got it
15:11-!-o_cee is now known as o_cee\movie
15:11<o_cee\movie>later :)
15:17<Morph>any of guys in California?
15:18<Morph>which part?
15:18<tmk>silicon valley
15:18<Morph>Southern California.
15:18<tmk>they have computers down there?
15:18<Morph>yes. a few.
15:19<zzo>not as many a up here...
15:19<tmk>darn skippy
15:24-!-tmk [] has quit ["Client exiting"]
15:24<ByteNik>I'm unable to watch livetv... Myth gives me an error... lemme retype:
15:26<ByteNik>X Error: BadMatch (invalid parameter attribute) 8
15:26<ByteNik> Major opcode: 143
15:26<ByteNik> Minor opcode: 19
15:26<ByteNik> Resource id: 0xa00002
15:26<ByteNik>I get that over and over
15:26<ByteNik>And the MythTV screen goes completely black.
15:26<ByteNik>I can watch in mplayer tho
15:28<ByteNik>Any ideas?
15:28<FryGuy>i'm in sacramento
15:33<Morph> :)
16:01<Captain_Murdoch_>Quote from Colin's latest post... "This time I have no idea what to do..." When did he ever?
16:02<ahbritto>Captain_Murdoch_: How are you detecting logos? Minimum intensity of a pixel over several frames?
16:02<Captain_Murdoch_>Should have read futher... "I am not posting the error because it is the typical sound card problem error..."
16:03<Captain_Murdoch_>ahbritto: checking for certain percentages of light and dark pixels in the area where the logo must appear. also has to occur for 5 seconds of time. also looked at doing a historgram on that area and a box slightly outsid the area but not doing that right now. need to see how accurate current method is.
16:04<Captain_Murdoch_>I had code (but lost it somewhere) which would try to detect the borders of the rating box if it existed but I'm not sure if that or this way is more accurate.
16:05<Captain_Murdoch_>I'm talking rating logos. haven't delved into station logos much yet. that will use intensity though.
16:05<Captain_Murdoch_>sorry, typed all that then realized you meant station.
16:05<ahbritto>Thanks. :)
16:05<Captain_Murdoch_>I want to get a method that doesn't require you to hit a key to tell it the logo is present.
16:06<sfr>Captain_Murdoch how do these rating logos look like?
16:07<ahbritto>Well, the intesity thing show work well, and you could have a confidence level based on the percentage of agreeing frames.
16:07<Captain_Murdoch_>either black box with white text or vice versa. I've seen one station use transparent white though.
16:08<ahbritto>sfr: Ratings logos:
16:08<Captain_Murdoch_>I've seen some logo detection code from a program called Karin but haven't looked at it enough or tested it to see if or how well it works.
16:09<Captain_Murdoch_>also some channels will display the rating logo in their little 5-second commercial before the show so that has to be considered as well.
16:10<ahbritto>I love it when your code works. I hate it when it doesn't... When it works, it is so clear how nice things could be. :)
16:10<Captain_Murdoch_>commercial detection is working great for 90-95% of the stuff I watch so I haven't been motivated to work on it much lately.
16:11<Captain_Murdoch_>I've started recording more shows lately since I added more storage though so I've been motivated a little more. :)
16:11<ahbritto>I have much worse.
16:11<ahbritto>Cartoon network works great.
16:11<sfr>Captain_Murdoch does the light/dark percentage method work? are there no logos with a similar distribution?
16:11<Captain_Murdoch_>that reminds me, I meant to record an episode of cops since someone said that failed.
16:12<ahbritto>I will try to note problem programs and post to the list.
16:13<Captain_Murdoch_>sfr, it's not a 60/40 or so method, it's something like 60/20 (since the logo isn't exactly all white/black) or 65/15. still working on it and tossing around ideas. will probably benefit more though with the station logo detection when I get that working.
16:13<Captain_Murdoch_>ahbritto: thanks.
16:15<sfr>Captain_Murdoch what about character recognition? might be overkill
16:15<Captain_Murdoch_>sfr: I'll probably try to rewrite the box-detection code to find the borders of the rating box to help the current method, but sometimes it's a black box with white lettering on a black background so edge detection won't work.
16:15<Captain_Murdoch_>would benefit more from station logo detection than trying to do OCR on the rating logo. :)
16:15<ahbritto>sfr: You want light wieght algorythms for real time processing.
16:16<ahbritto>sfr: also logos can be graphic.
16:16<Captain_Murdoch_>this won't be real-time, only use it in the post-record thread.
16:16<sfr>ahbritto logos are a different area, true
16:17<Captain_Murdoch_>I've seen a webpage/pdf which did really good logo detection but the guy didn't post code. it was a research project or something. I have the url bookmarked and file saved I think.
16:17<Captain_Murdoch_>he had code snippets for some things.
16:18<ByteNik>anyone have any idea about my problem?
16:18<ahbritto>Captain_Murdoch_: I've been thinking about signatures since we last talked.
16:18<sfr>Captain_Murdoch but you could try a vector-based approach, meaning detecting the junction/length of the three parts of the letter Y
16:18<sfr>for example
16:18-!-o_cee\movie [] has quit [Read error: 110 (Connection timed out)]
16:18<ahbritto>You thought signature detection would be expensive.
16:19<ahbritto>I was thinking, it could be relatively light wieght, translating a frame to a single 16 bit or 32 bit value.
16:20<Captain_Murdoch_>histogram would be more accurate.
16:20<ahbritto>For a frame?
16:20<Captain_Murdoch_>some guy on the mailing list posted a patch which tried that for scene change detection. that was back in march but he never posted again abut how well it worked.
16:21<Captain_Murdoch_>ahbritto: yeah. the scene change stuff uses a histogram right now to see if it looks like the scene changed frame-to-frame.
16:21<ahbritto>I was thinking commericals could be detected as a sequence of frame singatures.
16:21-!-ByteNik [] has left #mythtv []
16:22<ahbritto>Only commericals repeat in multiple programs.
16:24<Captain_Murdoch_>yeah, but you'd have to figure out what frames to use. and with ntsc video, your values could be different every time you see the same frame of video.
16:24<ahbritto>With only 32 bits of resolution, you would map near images to each other.
16:24<ahbritto>A fingerprint for a commerical would be made from combined signatures of an entire commerical.
16:25-!-Rule [] has quit ["sleep"]
16:25<ahbritto>Once a commerical was identified, you could have a finger print for the first 1, 2, and 5 seconds. And for the entire commerical.
16:26<ahbritto>So, detecting repeat commericals becomes a light wieght real-time taks.
16:26<Captain_Murdoch_>look at the pixel values sometime. they sometimes change frame to frame even if the pixel is suposed to be the same color. in looking at the rating symbol stuff, I would see the same pixel in consecutive frames with a luma ranging from 240 to 255 when they were supposed to be all black.
16:27<ahbritto>So, with 480 lines. you would translate 480/32 = 15 lines per bit.
16:27<Captain_Murdoch_>if the pixel values frame to frame for a given position vary by even 1, that could be 250000 difference in your 32-bit number.
16:27<sfr>Captain_Murdoch give each color a range per R/G/B which is allowed
16:27<Captain_Murdoch_>480 x 480 pixels each off by one even though the frame is supposed to look the same as the previous frame.
16:28<sfr>well if you have R/G/B info
16:28-!-linagee_ is now known as linagee
16:28<Captain_Murdoch_>sfr, so what happens when it teaters on the edge of the range?
16:28<ahbritto>Basically, you want to compress not checksum.
16:29<ahbritto>Perhaps: given a line, take the top for bits of ranges to produce a number.
16:29<Captain_Murdoch_>pixels teatering on the edge will cause your numbers to be off as well, although not by as much. that's part of why I said a histogram might be more accurate.
16:29<sfr>Captain_Murdoch hm, adjust the range 'properly' :)
16:29<ahbritto>You take pixels 0..32 for one number, 16..47 for the next.
16:30* sfr doesn't find teaters via dictd
16:31<Captain_Murdoch_>teater-totter. :)
16:31<Captain_Murdoch_>is that the proper spelling? probably not.
16:32<ahbritto>Given, the color average of several sections, that is very fuzzy, then you go digital to generate a frame signature.
16:32<sfr>Captain_Murdoch and you could compare with this pixels neighbours to judge it's color
16:32<Captain_Murdoch_>range idea might work, that's sort of what a histogram gives you depending on how wide each bucket is.
16:33<sfr>Captain_Murdoch it works (for a completely different application, though)
16:33<ahbritto>Let me try to be more clear: you want fuzzy data to generate digital data.
16:34<Captain_Murdoch_>sounds like coding for someone into programming for video way more than I am. :) probably easier than I think, but not something I'd tackle anytime soon.
16:34<Captain_Murdoch_>ahbritto: yeah sounds right.
16:34<ahbritto>Captain_Murdoch: I could write some psudo code. Unfiortunately, I don't have time to do a full implementation.
16:35<sfr>i might have applicable knowledge but lack the programmers part
16:36-!-paulproteus|lapt [] has quit [Read error: 60 (Operation timed out)]
16:36<ahbritto>I would imagine signature detection would be about as hard as random logo detection.
16:36<Captain_Murdoch_>I've been searching on google multiple times for code or algorithms. found lots of "papers" people wrote on commercial detection but not much actual code.
16:37<Captain_Murdoch_>ahbritto: nah, random logo can use intensity to see if there is a certain area that is always brighter or doesn't change intensity while the rest of a block does.
16:37<sfr>you could decide based on the presence of a logo. in germany during commercials the staion logo isn't shown.
16:38<ahbritto>I've reconsidered, signatures probably 5-10x harder...
16:38<Captain_Murdoch_>ahbritto: if you post examples of what shows don't detect too well, post the channel as well. I did a search on mythweb for cops and found it on about 4-5 channels here.
16:39<ahbritto>Captain_Murdoch_: Well do. I didn't realize you were lacking for data or I would have done it sooner.
16:39<Captain_Murdoch_>sfr: yeah, that's my idea. same for rating logo. so they could be used to help eliminate false positives. if logo or rating logo present then this 30-second spot is not a commercial.
16:39<Captain_Murdoch_>ahbritto: well, I hate it when my code works for me but not for others. would be nice if there's help debugging, but if I can just record a show and test myself to see what's going on it's ok as well.
16:40<Captain_Murdoch_>some stations do weird stuff.
16:40<ahbritto>Captain_Murdoch_: My signature idea could be used to detect MTV videos. So, it could build an archive of videos by constantly recording...
16:41<sfr>ahbritto signature? can you explain it to me?
16:41<ahbritto>I have done code review for somethings. And would be happy to do more, if the code is selt contained.
16:42<ahbritto>sfr: Each frame of video can be compressed into a single number: a signature.
16:42-!-zzo [] has quit ["Leaving"]
16:42<ahbritto>This can be done even if the frame varies slightly.
16:42<sfr>resulting in the same signature?
16:42<Captain_Murdoch_>all commercial skip stuff is in commercial_skip.cpp with methods being called from a few other files during recording and the flagging thread.
16:43<ahbritto>A commerical is a sequence of singatures: which can be hashed into a fingerprint. The whole commerical can be finger printed or just the first few seconds.
16:43<ahbritto>sfr: Yes, even if the frame varies slightly.
16:43<ahbritto>The important part is sequencing eliminates false-positives/negatives.
16:44<Captain_Murdoch_>I think those single-frame signatures will vary unless you have a very large signature.
16:45<ahbritto>sfr: So there are two parts to commerical detection with signatures: non-real-time comparing different programs, real-time looking for previously recognized commericals.
16:45<sfr>ahbritto you want to use the signatures only to detect repetiions of them?
16:45<sfr>ok, answers my question
16:46<sfr>ahbritto real-time looking? to skip commercials while watching tv?
16:46<ahbritto>Captain_Murdoch_: I believe, some stations may skip a frame or two per second to insert an extra commerical, other than that frames should be fairly consistent.
16:47<ahbritto>sfr: Real-time = Start watching late, skiping commericals till caught up, perhaps, lowering audio while caught up.
16:47<Captain_Murdoch_>hmm, this place is selling Myth boxes...
16:48<ahbritto>that is lowering audio during detected commericals while caught up.
16:48<Captain_Murdoch_>ahbritto: not with ntsc. I think the comment people make about ntsc is "Never The Same Color".
16:48<sfr>still might work with PAL ;)
16:49<ahbritto>Captain_Murdoch_: Need a real fuzzy algorythm, I think you are thinking much too digitally...
16:49<Captain_Murdoch_>$40 for a printed copy of the howto...
16:50<Captain_Murdoch_>ahbritto: probably. :) I'm by no means a video expert..
16:50<ahbritto>For example: given 32 pixels, average them then only consider the top 4 bits.
16:50<ahbritto>Captain_Murdoch_: I'm no expert either.
16:51<ahbritto>Then step 16 pixels and take another sample.
16:51<ahbritto>The 4 bit samples can the be treated digitally.
16:52<ahbritto>can then
16:53<Chutt>captain_murdoch, email them and ask for more details/source, etc =)
16:53<Captain_Murdoch_>have to think about different capture resolutions as well. I record some shows at low-res and some at high so pixels would have to take that into account.
16:53<Captain_Murdoch_>Chutt: yeah, like what happens when xmltv changes their page. they say for 995 you get online email support.
16:54<Chutt>i haven't heard of them before, at least
16:55<ahbritto>I think my pvr250 comes with titantv listings if I use windows. I haven't heard much about xmltv and that...
16:55<Chutt>titantv doesn't provide a data feed, as far as i know
16:55<Chutt>they're web-only
16:55<Chutt>and not in a format that would be easily parsed.
16:57<Captain_Murdoch_>I found that page by searching for commercial detection code on google. they came up 8th on the front page of results.
16:57<ahbritto>"Schedule your TV recordings with WinTV-Scheduler. The Electronic Program Guides in the U.S., in Europe and iEPG in Japan, are Internet based program guides which work with WinTV-PVR to help schedule your TV recordings"
16:57<ahbritto>This is from:
16:58<ahbritto>I guess the key is "to help sechedule"
16:58<Captain_Murdoch_>ahbritto: check out for more info like what you're talking about. they detect commercials and even know which one was found.
16:59-!-jkolb [] has quit []
16:59<ahbritto>Gee, hope they don't file for any silly patents. :/
17:00<Captain_Murdoch_>even have realplayer clips of commercials you can play.
17:01<Chutt>stuff like that's patented, of course :p
17:03<Captain_Murdoch_>and probably funded in-part by the taxpayers since it's a university
17:03* Captain_Murdoch_ goes to doublecheck that to see if said "institute" instead. :)
17:04<Chutt>and it'd be a tiny bit of a conflict of interest if there was anything like that in mythtv
17:05<Captain_Murdoch_>The guy that wrote Karin is the one that wrote one of those papers on logo detection that I saved.
17:05<ahbritto>"Our ComDetect product utilizes VidWatch's patented real-time video content analysis technology..."
17:07<ahbritto>Chutt: Do you accept paypal donations/thank yous?
17:14<Captain_Murdoch_>cool. just thought of something. part of the reason I couldn't get logo detection working well earlier was because I didn't have a way to detect when there had been enough motion. now I can use the scene change stuff to determine that. if the histogram says there's over X percent change then I should have enough info to see if a logo is present.
17:15<Captain_Murdoch_>re-reading that doc I posted above made me realize that.
17:15<Chutt>snapstream 3.2's changelog: Join a recording (MPEG2) in progress
17:15<Chutt>someday, the windows programs will catch up
17:16<Captain_Murdoch_>final sentence in this guy's document referring to the example algorithms he gave.. "Hopefully, they will play a role in the movement for creating commercial-free television."
17:16<Captain_Murdoch_>little did he know back in 2001... :)
17:16<sfr>Captain_Murdoch what's the format of the frames you're analyzing in commercial_skip.cpp?
17:16<Captain_Murdoch_>yuv, myth's standard format.
17:17<Captain_Murdoch_>so I'm only looking at the Y (luma)
17:19-!-Justin_ [] has quit [Read error: 104 (Connection reset by peer)]
17:19-!-Justin__ [] has joined #mythtv
17:20-!-Justin__ is now known as Justin_
17:20<sfr>Captain_Murdoch, so how would i access the color data?
17:21<Captain_Murdoch_>it's yuv, so if you want rgb you have to convert it.
17:22<Captain_Murdoch_>luma is enough for blank-frame, rating symbol, and logo detection so I haven't had a need to convert to color for anything. it's easy to do though and there's code in libavcodec to convert a whole frame or you can do it pixel by pixel.
17:24-!-kja [] has joined #mythtv
17:25<Captain_Murdoch_>Chutt, I'd have to use another email to mail that website about their server. they probably read and ask questions in the mailig lists. maybe that's Colin doing their online support. ;)
17:27<Chutt>someone really needs to suggest a windows alternative
17:27<Chutt>he just keeps digging himself deeper
17:28<Captain_Murdoch_>send him to snapstream.
17:32<kja>Have you peeked at that domain the mail originate from?
17:49-!-kja [] has quit ["..\..\..\"]
17:53-!-kja [] has joined #mythtv
17:58-!-Drikus [] has quit ["toedeledoki"]
17:58<Captain_Murdoch_>kja, you mean their www page? if so, guess we're lucky it's not Stripes asking the questions.
18:02<kja>yea, that's what i meant :))
18:05<Captain_Murdoch_>must be colin's website as he's got the only Christmas list online.
18:07<Captain_Murdoch_>time to head home. bbl...
18:07-!-Captain_Murdoch_ [] has quit []
18:44-!-sc00p__ is now known as sc00p
18:45-!-jkolb [] has joined #mythtv
18:54-!-o_cee [] has joined #mythtv
18:57<o_cee>has anyone got bootsplash setup?
19:03<kja>belive there was someone around here a couple of days ago, working on that..
19:04<o_cee>that'd be me. just wanted to know if anyone had it setup who'd actually try the pics
19:05<kja>ahh, sorry for not remembering your nick
19:05<o_cee>np :)
19:12<o_cee>got down the filesize to 80kb.. is that ok to send to the list?
19:14<kja>no, max is 40k i think, better use a url..
19:14<o_cee>okay. i'll email that other guy and have him put it up
19:15<o_cee>if anyone wants them right away i can email them directly
19:16-!-sfr [] has quit [Remote closed the connection]
19:16<kja>sure, dunno when i'll get time to test (working on dvb stuff right now)
19:17<o_cee>who are you?
19:17<o_cee>i mean, i always see people's "real" names in themailinglists
19:19-!-term [] has quit [Read error: 104 (Connection reset by peer)]
19:19<kja>yea, not easy to figure that out
19:21-!-schultmc [] has quit ["Client exiting"]
19:24<o_cee>no one else who wants it right away?
19:25<kja>way, those looked nice, really!!
19:26<o_cee>looks alot nicer then the normal startup
19:26<Morph>o_cee: just sent you an email.
19:27<o_cee>go it
19:27<o_cee>sending the mail to the usr list
19:28<Morph>wait a minute.
19:28<Morph>your tar.gz is screwy :)
19:28<Morph>ya. how did you tar this up?
19:28<o_cee>tar -cf mythtv-bootsplash-0.1.tar *
19:28* o_cee is no good at that
19:29<Morph>oh okay.
19:29<o_cee>oohkay? :)
19:29<Morph>the .tgz on the end threw me.
19:29<Morph>fixed the file, removed the tgz on the end.
19:29<o_cee>me didn't add that?
19:30<o_cee>where the hell did that come from
19:30<Morph>its on the end of the attachment.
19:30<Morph>there are nice.
19:30<o_cee>hhe, must have been my apache
19:30<o_cee>dragon mythtv # ls
19:30<Morph>so this is to hide kernel messages right?
19:31<o_cee>you get a nice progress bar as well
19:31<Morph>I might try this out later.
19:31<o_cee>and it's also possible to have animations
19:31<Morph>hows it work?
19:31<o_cee>it's .mng animations
19:31<o_cee>but that's the part that works the worse
19:32<Morph>No No. I mean how does it hide the kernel messages? whats the setup for using it?
19:33<Morph>ya. that looks like its it.
19:33<Morph>hmm this looks like what LPP used it be.
19:34<o_cee>think lpp is dead right?
19:34<o_cee>this is the new thing :)
19:34<Morph>Nod I think so.
19:34<Morph>cool. I'll have to check it out.
19:35<o_cee>beware... it might take time to get working.. took me two days. but.. i'm not that good at this stuff :)
19:35<Morph>warning noted :)
19:35<o_cee>running gentoo?
19:36<o_cee>okay, know nothing about it then..
19:37<Morph>you're using gentoo?
19:37<o_cee>got the patches applied with their kernel
19:37<Morph>bbl. home.
19:37<o_cee>not much to get it working really.. hehe
19:37<o_cee>ah, me is going to bed soon
19:38-!-mecraw [~mecraw@] has quit ["Trillian ("]
19:40-!-just1nux [] has joined #mythtv
19:43-!-bbeattie [] has quit ["Client Exiting"]
19:49-!-o_cee [] has quit ["me is off to the bed"]
19:51<just1nux>To anyone listening. I have 3 cards in 1 backend, 2 ivtv, 1 bttv. they all have one input which use the same source. One of the pvr-250's has an additional source. when 3 programs are scheduled for the same time, all from the same source, everything works fine, but if the pvr-250 with the alternate source is using that alternate source, its as if the third card doesnt exist. It will record only from the two pvr-250's and refuse to
19:52<just1nux>ive tv during this time
19:57-!-hadees [] has joined #mythtv
20:03-!-Ripp [] has joined #mythtv
20:16-!-jkolb [] has quit []
20:18-!-bbeattie [] has joined #mythtv
20:19-!-bbeattie [] has quit [Client Quit]
20:19-!-bbeattie [] has joined #mythtv
20:44-!-matt_ [] has joined #mythtv
20:57-!-hfb [] has quit ["Client exiting"]
21:00<matt_>i have my default, live tv, and trascode profiles all set up the same. when i transcoded a recording after marking the commercials, the file was about half the size, and half the quality
21:01<matt_>it looked noticeably worse
21:11-!-bbeattie [] has quit ["Client Exiting"]
21:14-!-ByteNik [] has joined #mythtv
21:14<ByteNik>Hey hey
21:14<ByteNik>Anyone out there at this time that can help me with a problem? :-P
21:14-!-bline [] has quit ["Client exiting"]
21:14<ByteNik>I don't want to spam the channel with it again if no one is around.
21:15<matt_>just ask the question
21:16<ByteNik>[15:24] <ByteNik> I'm unable to watch livetv... Myth gives me an error... lemme retype:
21:16<ByteNik>[15:25] <ByteNik> X Error: BadMatch (invalid parameter attribute) 8
21:16<ByteNik>[15:25] <ByteNik> Major opcode: 143
21:16<ByteNik>[15:25] <ByteNik> Minor opcode: 19
21:16<ByteNik>[15:25] <ByteNik> Resource id: 0xa00002
21:16<ByteNik>[15:26] <ByteNik> I get that over and over
21:16<ByteNik>[15:26] <ByteNik> And the MythTV screen goes completely black.
21:16<ByteNik>What could cause that problem?
21:16<ByteNik>I can watch via mplayer, btw.
21:18* matt_ doesn't know
21:25<kja>Could you try to reset your gui settings at MythFrontend / genral / apperance (2nd screen)...?
21:25<Ripp>hey anyone here use xmltv for dishnetwork listings?
21:25* kja wonders if that would help @ all?
21:25<ByteNik>Tried it. :-/
21:26<kja>reduce your resolution?
21:26<ByteNik>And I even recompiled
21:26<kja>x-change graphics driver?
21:27<ByteNik>to what? :P
21:27* kja is shooting blindly
21:27<kja>sorry, bad sugestion, as your mplayer is working
21:28<kja>try export USE_SHM=1 before mythtv
21:28<ByteNik>What's it do?
21:28<kja>change from pure xv to shared memory
21:28<ByteNik>lemme try
21:29<matt_>oh, that's what i was going to ask...
21:29<matt_>has anyone here gotten ppv listings for directv? tarek's script doesn't work for me
21:29<kja>sorry, that should be export NO_XV
21:36-!-dopez [] has quit [Remote closed the connection]
21:39-!-dopez [] has joined #mythtv
21:43* kja thinks ByteNik is watching tv
21:43<ByteNik>I think not. :P
21:44<kja>you tried that last (no_xv)?
21:44<ByteNik>No workies
21:47<kja>and no other errors/warnings?
21:47<ByteNik>Just the one I pasted
21:50* kja is trying to figure it out, but...
21:55-!-term [] has joined #mythtv
21:57-!-automatic_jack [] has joined #mythtv
21:59<kja>ByteNik: sorry to bother you with small tests, but try export NO_SHM=1 also...
22:00<kja>ByteNik: btw, which graphics card/driver are you using?
22:01<kja>nd x version...
22:02<automatic_jack>I set up MythTV the other night, and I've got it to a point where it will record, but I see nothing but a black screen. I assume this is a problem with my TV tuner (ATI TV Wonder VE). can anyone suggest some things I can do to troubleshoot it?
22:03<kja>ByteNik: and when you try these export ...;mythfrontend, do you se a message similar to: using .. port <number>?
22:04<kja>a..._jack: hmmm, a lot of people see black screens in livetv at the moment...what does mythfrontend output?
22:05<kja>automatic_jack: query if you feel like it's to much
22:06<automatic_jack>kja: what do you mean, what does mythfrontend output?
22:06<automatic_jack>oh, my sound card doesn't seem to be working in that system - mythtv says it can't read /dev/dsp
22:07<automatic_jack>I know I need to fix that in my kernel, but could that cause blank video?
22:07<kja>yea, because it syncs video with audio....
22:08<automatic_jack>so the first thing I should probably do is fix my sound card :)
22:09<kja>that would be a good starting point :)
22:09<automatic_jack>ok, let me go throw this sblive card in there.
22:10<kja>sounds good!!
22:12-!-matt_ [] has quit [Read error: 104 (Connection reset by peer)]
22:15-!-bline [] has joined #mythtv
22:26-!-[-ripp-] [] has joined #mythtv
22:26-!-[-ripp-] is now known as RippNB
22:39-!-pablo__ [] has joined #mythtv
22:39-!-BenB [] has joined #mythtv
22:45* pablo__ wonders if anyone knows why after upgrading to 0.11 his audio on recordings doesn't come in for a minute or so
22:45-!-owen_ [] has joined #mythtv
22:45<owen_>I have an input source that I want to delete. how do I do that?
22:45<Chutt>set it to none
22:46<owen_>video source, that is. where do I set it to none?
22:47<owen_>I created two video sources but I only have one. so I have one associated, and the other not, so mythbackend won't run.
22:47<Chutt>just delete it from the db, then
22:47<owen_>using the mysql frontend?
22:47<owen_>I suck with that :)
22:48<Chutt>i dunno, there's plenty of other tools
22:48<Chutt>phpmyadmin or whatnot
22:55<owen_>ok I think that's all set.
22:55-!-owen_ [] has quit ["[BX] Tony the Tiger uses BitchX. Its Grrrrrrrrreat!"]
22:57-!-owen_ [] has joined #mythtv
22:57<owen_>this is weird.
22:57<owen_>this was working the other night.
22:57<owen_>now I get "Qserversocket: failed to bind or listen to the socket."
22:58<owen_>when I try to run mythbackend
22:58<kja>killl the other mythbackend or reboot to reset the stale socket
23:02<ByteNik>Chutt... I get no video in MythTV but I do get it in mplayer.
23:02<ByteNik>I do see an error
23:02-!-paulproteus|lapt [] has joined #mythtv
23:03<ByteNik>[21:16] <ByteNik> [15:25] <bytenik> X Error: BadMatch (invalid parameter attribute) 8
23:03<ByteNik>[21:16] <ByteNik> [15:25] <bytenik> Major opcode: 143
23:03<ByteNik>[21:16] <ByteNik> [15:25] <bytenik> Minor opcode: 19
23:03<ByteNik>[21:16] <ByteNik> [15:25] <bytenik> Resource id: 0xa00002
23:03<ByteNik>It shows up over and over
23:07<kja>ByteNik: why did you not bother answering my last questions (you just \0x27)?
23:07<ByteNik>Didn't see them.
23:08<ByteNik>What did you ask?
23:08<kja>[03:59:08] <kja> ByteNik: sorry to bother you with small tests, but try export NO_SHM=1 also...
23:08<kja>[03:59:51] <kja> ByteNik: btw, which graphics card/driver are you using?
23:08<kja>[04:00:52] <kja> nd x version...
23:08<ByteNik>GeForce FX 5200, standard nVidia drivers
23:09<ByteNik>And I did the NO_SHM=1 before you told me to do the other.
23:10<kja>oh, you have read videoout_xv.cpp then?
23:10<kja>I misqouted the source and said:
23:10<kja>[03:27:49] <kja> try export USE_SHM=1 before mythtv
23:11<kja>which should be NO_XV
23:11<ByteNik>I did both of what you said before.
23:11<kja>did you do the third?
23:12<ByteNik>Just before
23:12<ByteNik>No work
23:12<kja>and: [04:02:54] <kja> ByteNik: and when you try these export ...;mythfrontend, do you se a message similar to: using .. port <number>?
23:12<kja>you see that message? those commands didn't work then...
23:13<ByteNik>export NO_SHM=1 && mythfrontend
23:13<ByteNik>Substitute in with the other variables also
23:14<kja>seems right to me...?
23:14* kja is scratching his head...
23:14<ByteNik>No message when I eliminate the &&
23:14-!-owen__ [] has joined #mythtv
23:15-!-owen_ [] has quit [Read error: 113 (No route to host)]
23:15<kja>; then?
23:15<owen__>apparently running mythtv with only 128M of RAM will hang your system?
23:15<owen__>I got video to work, but when I started to change channels the system bogged down quickly and then hung.
23:15<owen__>couldn't ping it or anything.
23:15<kja>without swap or something?
23:15<owen__>oh hmm.
23:16<owen__>when I run free it sees the swap.
23:17<owen__>a total of 375M.
23:17<pablo__>is there any reason why live tv audio would work but recorded tv audio would have problems?
23:17<kja>did the mythtv process grov much?
23:18-!-RippNB [] has quit ["Client exiting"]
23:18* pablo__ is using a pvr250 card on a slow machine (700mhz) if that affects things
23:19<kja>what problems exactly?
23:20<owen__>I couldn't see what happened to my mythtv process.
23:20<owen__>I had to hard reset.
23:20<pablo__>kja: for some reason the audio doesn't come in until a minute late
23:20<kja>does it do it every time?
23:20<pablo__>kja: sometimes a couple minutes
23:21<pablo__>kja: it seems to
23:21<pablo__>kja: but what's odd is that live tv works fine, which i presumed would work the same way (hence the performance issues on slow machine with simultaneous record/playback)
23:21* kja wonders who he's talking to :)
23:22* pablo__ gives the floor to owen ;-)
23:22<kja>no worry, i'll try to direct my msgs (the ones before was directed at owen__)..
23:23<kja>pablo: you have a nvidia adapter?
23:23<pablo__>kja: yeah, geforce2 i think
23:24<pablo__>kja: i also just upgraded 0.10 -> 0.11 which may have something to do with it (it worked fine before)
23:24<kja>pablo: you have turned on experimental a/v sync and are running the lates driver which supports it (4363?)
23:24<owen__>I'm just not sure why my system would show video and then hang after trying to change channels.
23:26<kja>owen: me neither, make's no sense (but it's software ya:), but is it reproducible?
23:26<pablo__>kja: i'm on the latest driver, haven't tried experimental a/v sync since i didn't need it before, though i'll try it now
23:27<kja>pablo: latest is 4496, and that broke the a/v sync :/
23:28<kja>chutt: have you gotten around to mailing that developer @ nvidia?
23:31<pablo__>kja: i think i just grabbed it from cvs, should i grab a known working old version?
23:32<kja>pablo: use the latest cvs version of mythtv, but make sure your nvidia driver is 1.0-4363 (not 1.0-4496)
23:33-!-owen__ [] has quit ["BitchX-1.0c20cvs -- just do it."]
23:33<pablo__>kja: oh, i'm just using X and doing nvidia -> avermedia imicro -> tv
23:34-!-dopez [] has quit ["..."]
23:34<pablo__>kja: errr, the regular nvidia driver included with xfree 4.2.1
23:35-!-owen_ [] has joined #mythtv
23:35<owen_>this is getting annoying :)
23:35<kja>owen: what's up?
23:35<owen_>dja what did you say just before in response to my comment about mythtv dying while changing channels?
23:36<owen_>I was trying to get the tv out on my card to work and it messed up my console :)
23:36<owen_>I need to tackle one problem at a time.
23:36<owen_>kja actually I think it was you that responded to my comment.
23:36<owen_>but I couldn't read it.
23:36<kja>pablo: seems like you are using the regular nv driver that comes with X (no good), go to
23:38<kja>owen: you said 'hmmm' and then a bit later [05:33:15] * owen__ has quit IRC ("BitchX-1.0c20cvs -- just do it.")
23:38<owen_>ok, I got mythtv to run with video.
23:38<owen_>then I tried to change channels, and it worked a little, then eventually hung.
23:38<owen_>and I had to hard-reset the system.
23:39<Chutt>then it's not a mythtv problem, you've got other issues
23:39<owen_>I figured.
23:39<owen_>probably something with the tuner.
23:39<Chutt>a userspace program can't cause your system to hang
23:39<owen_>that's my guess.
23:39<owen_>I know.
23:39<Chutt>yeah, drivers or hardware or whatnot =)
23:39<owen_>ok well here's one other issue.
23:39<owen_>I installed the Debian mythtv package.
23:39<pablo__>kja: but do the x nvidia drivers affect audio sync?
23:40<owen_>and when I boot, it tries to start mythbackend, but gets an error about not being able to create .qt or something.
23:40<owen_>but if I run it manually as the mythtv user it works find.
23:40<Chutt>owen, just ignore that
23:40<owen_>it seems to mess up mythfrontend.
23:41<Chutt>it shouldn't
23:41<pablo__>owen_: fwiw, i always have to restart the mythtv-backend after debian starts it, though i'm not quite sure why
23:41<kja>pablo: they do a better job at it, as long as you go via archives, and downloads 1.0-4363:
23:42<pablo__>kja: though its totally synced up for live tv ,that's what i don't really understand
23:42<pablo__>kja: though i'll try out the nvidia driver
23:43<Chutt>pablo, make sure you're using the latest ivtv driver
23:43<kja>pablo: me nither, although it really should be the same thing, only the player get's fed from multiple places
23:43<Chutt>other than that, i don't know what it could be, the live-tv and recording code is identical
23:44<pablo__>Chutt: is there any possibility that i'm using hardware encoding/decoding on live tv but not on recording?
23:45<Chutt>do you have two cards?
23:46<pablo__>Chutt: nah, just one pvr250 (and i just upgraded from 0.10->0.11 with debian packages, fwiw)
23:46<Chutt>then, no, that's not a possibility -- those cards don't really do analog capture yet in linux
23:50-!-radsaq [] has quit [Connection reset by peer]
23:53-!-BenB [] has quit ["Leaving IRC - dircproxy 1.0.3"]
23:53-!-radsaq [] has joined #mythtv
---Logclosed Sat Aug 23 00:00:07 2003