| *** tpb has joined #timvideos | 00:00 | |
| *** tvCommitBot has joined #timvideos | 00:09 | |
| tvCommitBot | [gst-switch] jhenstridge opened pull request #218: Don't use gcov symbols if coverage is disabled. (master...fix-disable-coverage) http://git.io/vCb1m | 00:09 |
|---|---|---|
| *** tvCommitBot has left #timvideos | 00:09 | |
| *** Bertl_oO is now known as Bertl_zZ | 02:40 | |
| *** tvCommitBot has joined #timvideos | 03:58 | |
| tvCommitBot | [dvsource-v4l2-other] CarlFK pushed 1 new commit to master: http://git.io/vCNvn | 03:58 |
| tvCommitBot | dvsource-v4l2-other/master 8605ea6 Carl Karsten: add all the gstreamer deps | 03:58 |
| *** tvCommitBot has left #timvideos | 03:58 | |
| mithro | CarlFK: ping? | 04:47 |
| CarlFK | mithro: pong | 04:48 |
| mithro | CarlFK: I'll accept a pull request which makes the defaults for the misoc-firmware match https://github.com/timvideos/HDMI2USB-misoc-firmware/issues/111 | 04:48 |
| tpb | Title: defaults for video_matrix · Issue #111 · timvideos/HDMI2USB-misoc-firmware · GitHub (at github.com) | 04:48 |
| CarlFK | untested? ;) | 04:50 |
| mithro | CarlFK: nope | 04:50 |
| mithro | CarlFK: you should only need to modify a little bit of C code | 04:50 |
| CarlFK | and test it | 04:51 |
| mithro | CarlFK: yes, but as you are only changing the C code, you can just do a "make firmware-lm32; make load-lm32" then type "reboot" into the prompt | 04:52 |
| CarlFK | "just" | 04:52 |
| mithro | CarlFK: that is like a 30 second process | 04:52 |
| mithro | CarlFK: as appose to recompiling the gateware which is a 15-30 minute process | 04:53 |
| CarlFK | kinda like I should just be able to use the same gstreamer pipeline on a pie | 04:53 |
| CarlFK | http://paste.ubuntu.com/12839134/ | 04:53 |
| tpb | Title: Ubuntu Pastebin (at paste.ubuntu.com) | 04:53 |
| CarlFK | any idea why I get 9 dv frames and then it just sits doing nothing ? | 04:53 |
| mithro | CarlFK: I would be slightly suspicious of the hardware JPEG decode that appears to be getting added to that pipeline | 04:54 |
| mithro | GstOMXMJPEGDec | 04:54 |
| CarlFK | that's the dvsource-v4l2-other pipeline with dvswitchsink swapped for file sink | 04:54 |
| CarlFK | hmmm | 04:54 |
| mithro | which is coming from the decodebin | 04:54 |
| CarlFK | good catch... i wonder how I surprise that | 04:55 |
| mithro | I wouldn't be surprised if it is choking on our JPEG frames | 04:55 |
| mithro | CarlFK: decodebin will try and create the most "efficient" pipeline | 04:55 |
| mithro | CarlFK: I wouldn't guarantee we are producing entirely 100% strictly compliant jpeg | 04:57 |
| *** techman83|clone has joined #timvideos | 04:58 | |
| *** ChanServ sets mode: +v techman83|clone | 04:58 | |
| *** tumbleweed_ has joined #timvideos | 05:00 | |
| *** se6astia` has joined #timvideos | 05:02 | |
| *** se6astia` is now known as se6astian | 05:02 | |
| *** tumbleweed has quit IRC | 05:04 | |
| *** techman83 has quit IRC | 05:05 | |
| *** xfxf has quit IRC | 05:05 | |
| *** Sewar has quit IRC | 05:05 | |
| *** aps has quit IRC | 05:05 | |
| *** hyades_zombie has quit IRC | 05:05 | |
| *** se6astian|away has quit IRC | 05:05 | |
| *** tumbleweed_ is now known as tumbleweed | 05:05 | |
| *** tumbleweed has joined #timvideos | 05:05 | |
| *** Sewar has joined #timvideos | 05:05 | |
| *** xfxf has joined #timvideos | 05:12 | |
| CarlFK | mithro: sudo apt-get remove gstreamer1.0-omx 'fixed' it.. except now there isn't enough cpu. | 05:19 |
| mithro | CarlFK: yeah - I wouldn't think the RPi would be fast enough to do that | 05:20 |
| *** aps has joined #timvideos | 05:23 | |
| seaLne | what are you trying to do with the rpi? | 05:23 |
| *** aps is now known as Guest46413 | 05:23 | |
| seaLne | mithro: just about to try your fixes | 05:25 |
| mithro | seaLne: great! | 05:26 |
| seaLne | mithro: should i have ended up at a BIOS> prompt? | 05:30 |
| mithro | seaLne: no | 05:30 |
| mithro | seaLne: ending up at the BIOS> prompt means the memtest failed normally | 05:31 |
| mithro | seaLne: can you run "memtest" | 05:31 |
| seaLne | failed | 05:31 |
| mithro | seaLne: can you paste the output into https://github.com/timvideos/HDMI2USB-misoc-firmware/issues/77 ? | 05:32 |
| tpb | Title: DDR Memory is currently broken on the Opsis · Issue #77 · timvideos/HDMI2USB-misoc-firmware · GitHub (at github.com) | 05:32 |
| seaLne | romboot worked though | 05:35 |
| CarlFK | seaLne: converting the mjpg to dv to send to dvswitch | 05:39 |
| seaLne | i managed that ok on an rpi2 | 05:39 |
| seaLne | but ended up running the source on a pc as couldn't get it to build on raspbian | 05:40 |
| seaLne | this for with the camera? | 05:40 |
| CarlFK | with the hdmi2usb | 05:40 |
| seaLne | ah | 05:42 |
| seaLne | you using a rpi2? | 05:42 |
| seaLne | mithro: should i be doing anything with switching modes now for after the initial serial? | 05:44 |
| mithro | seaLne: When you do a "make load-fx2" you should have working serial+webcam (see caveats) - you shouldn't need to use mode-switch | 05:45 |
| seaLne | neither working :( | 05:46 |
| CarlFK | seaLne: i dont think it is a 2. /proc/cpuinfo shows ARMv6-compatible processor rev 7 (v6l) | 05:47 |
| mithro | seaLne: what does lsusb / dmesg show? | 05:48 |
| seaLne | i never got anywhere with the camera on a 1 but the 2 did the converting and throughing accross the network fine | 05:48 |
| seaLne | https://paste.kde.org/pykqpq4zu | 05:49 |
| tpb | Title: KDE Paste (at paste.kde.org) | 05:49 |
| mithro | seaLne: "git describe" | 05:49 |
| seaLne | v0.0.0-429-ged90ee1 | 05:50 |
| seaLne | pretty sure i built the right branch this time, but still not fully awake yet | 05:50 |
| mithro | seaLne: great! that is the same version as me | 05:51 |
| mithro | seaLne: what does "git status" show? | 05:51 |
| seaLne | https://paste.kde.org/pprn0gfmq | 05:53 |
| tpb | Title: KDE Paste (at paste.kde.org) | 05:53 |
| mithro | seaLne: third_party/fx2lib (new commits) | 05:53 |
| seaLne | https://paste.kde.org/patr9ak0q | 05:54 |
| tpb | Title: KDE Paste (at paste.kde.org) | 05:54 |
| mithro | seaLne: I'm confused at why the Makefile didn't push fx2lib forward | 05:55 |
| seaLne | oh i forgot about the submodule stuff. did my git pull not bring it down? | 05:55 |
| *** hyades_zombie has joined #timvideos | 05:55 | |
| mithro | seaLne: the Makefile should have noticed the .gitmodule file changed and done an update... | 05:56 |
| mithro | seaLne: hold on a sec | 05:57 |
| mithro | seaLne: opps | 06:17 |
| mithro | seaLne: Pull my branch and type "make fx2-firmware" without updating the module | 06:17 |
| seaLne | mithro: git status still says the same | 06:19 |
| mithro | seaLne: what about "make gateware" ? | 06:19 |
| mithro | seaLne: actually "make third_party/fx2lib/.git" | 06:19 |
| seaLne | make: 'third_party/fx2lib/.git' is up to date. | 06:21 |
| mithro | seaLne: and what does "git status" show? | 06:21 |
| seaLne | https://paste.kde.org/p1aoxxiph | 06:21 |
| tpb | Title: KDE Paste (at paste.kde.org) | 06:22 |
| mithro | seaLne: "touch .gitmodules; make third_party/fx2lib/.git" | 06:22 |
| seaLne | https://paste.kde.org/peliok1qq | 06:23 |
| tpb | Title: KDE Paste (at paste.kde.org) | 06:23 |
| mithro | seaLne: hrm... Try rebuilding the fx2 firmware | 06:25 |
| *** tvCommitBot has joined #timvideos | 06:25 | |
| tvCommitBot | [gst-switch] jhenstridge opened pull request #219: Publish the audio and video input ports via Avahi (master...avahi-support) http://git.io/vCNY8 | 06:25 |
| *** tvCommitBot has left #timvideos | 06:25 | |
| *** jamesh has joined #timvideos | 06:28 | |
| seaLne | mithro: still behaving the same unfortunately | 06:33 |
| seaLne | including memtest | 06:34 |
| mithro | seaLne: can you explain the procedure you are following? | 06:34 |
| seaLne | i didn't rebuild gateware just fx2 | 06:34 |
| mithro | seaLne: that is fine | 06:36 |
| seaLne | make load-gateware; python hdmi2usb-mode-switch.py --mode=serial; make connect-lm32; encoder on; video_matrix connect pattern encoder; make load-fx2; vlc; make connect-lm32 | 06:36 |
| mithro | seaLne: don't do the mode-switch | 06:36 |
| seaLne | connect-lm32 needs a better way to cope with increasing ttyACMX at some point | 06:37 |
| mithro | seaLne: why is your ttyACMX increasing? | 06:37 |
| seaLne | disconnecting and reconecting? usb serial often does that | 06:38 |
| seaLne | without the hdmi2usb-mode-switch.py i don't have a tty this time | 06:38 |
| seaLne | dmesg hasn't said anything since jtag | 06:39 |
| seaLne | tty doesn't increase every time btw | 06:39 |
| seaLne | the tty increase would be solved by a udev rule creating /dev/ttyHDMI2USB or similar | 06:41 |
| jamesh | mithro: as a bit of background on my gst-switch pull requests, I'm a member of PLUG. I was poking around to see how the code works, with the eventual aim of getting it working with our AV gear (we currently use dvswitch for recording/streaming) | 06:42 |
| mithro | jamesh: You looked very familiar! :) | 06:42 |
| jamesh | the Avahi bug looked interesting because I'd like to get to a stage where we could just plug everything in and it would configure itself | 06:42 |
| mithro | jamesh: yeah | 06:44 |
| mithro | jamesh: so, Nick in Perth has one of my Atlys boards | 06:46 |
| jamesh | mithro: yep. That's the other reason I wanted to see if I could get gst-switch working. | 06:47 |
| mithro | jamesh: so if you want to play with what everyone else is playing with at the moment, feel free to grab it off him | 06:47 |
| mithro | jamesh: currently CarlFK is feeding the HDMI2USB output into dvswitch - which is kind of crappy as you are downscaling a HD feed into SD - but it is a good stepping stone | 06:47 |
| jamesh | We've also been talking about buying a few of the new boards too | 06:47 |
| jamesh | (all our DV cameras have HDMI output, so it would be a first step towards going HD | 06:48 |
| jamesh | ) | 06:48 |
| mithro | jamesh: the HDMI2USB does "hard switching" so something xfxf is playing with is just putting the presenter and camera into the single HDMI2USB and just using that for switching | 06:48 |
| mithro | jamesh: then you don't need any software mixing at all | 06:49 |
| seaLne | mithro: unfortunately i've run out of time this morning as i need to leave the house shortly | 06:50 |
| mithro | seaLne: okay | 06:51 |
| mithro | seaLne: will you be back later? or pretty much done for the day? | 06:52 |
| seaLne | probably away for ~10h | 06:52 |
| mithro | seaLne: okay, might still be around when you get back but probably not | 06:54 |
| mithro | jamesh: the crowdfunding ends in a week - so you should get your act together ordering the Opsis boards! | 07:49 |
| mithro | anyone know a difftool which will only compare the first x characters of a line | 08:01 |
| *** travis-ci has joined #timvideos | 08:17 | |
| travis-ci | [mithro/HDMI2USB-misoc-firmware/opsis-fixes#26] (7494351): The build has errored. (https://travis-ci.org/mithro/HDMI2USB-misoc-firmware/builds/85996790) | 08:17 |
| *** travis-ci has left #timvideos | 08:17 | |
| *** se6astian is now known as se6astian|away | 08:58 | |
| *** se6astian|away is now known as se6astian | 10:26 | |
| *** Bertl_zZ is now known as Bertl | 10:30 | |
| *** se6astian is now known as se6astian|away | 10:38 | |
| jamesh | mithro: yep. I just sent an email to our committee list to see if I can get a positive decision. | 10:39 |
| *** rohitksingh has joined #timvideos | 11:47 | |
| *** rohitksingh has quit IRC | 12:45 | |
| *** CARAM__ has quit IRC | 14:07 | |
| *** CARAM__ has joined #timvideos | 14:10 | |
| *** rohitksingh has joined #timvideos | 14:48 | |
| CarlFK | mithro: looks like diff ... man diff ... --line-format=LFMT format all input lines with LFMT | 15:21 |
| CarlFK | %[-][WIDTH][.[PREC]]{doxX}LETTER | 15:22 |
| CarlFK | printf-style spec for LETTER | 15:22 |
| CarlFK | %L contents of line | 15:22 |
| *** Bertl is now known as Bertl_zZ | 15:42 | |
| *** se6astian|away is now known as se6astian | 15:44 | |
| CarlFK | jamesh: also look at https://github.com/voc/voctomix | 15:50 |
| tpb | Title: voc/voctomix · GitHub (at github.com) | 15:50 |
| CarlFK | it is another candidate for replacing dvswitch | 15:51 |
| CarlFK | which also hasn't been used in production, but it is close: (10:50:33 AM) derpeter: CarlFK: not yet. we wanted to try it last week but MaZderMind had to work late | 16:02 |
| *** se6astian is now known as se6astian|away | 16:08 | |
| *** thaytan has quit IRC | 17:45 | |
| *** thaytan has joined #timvideos | 17:45 | |
| *** ChanServ sets mode: +v thaytan | 17:45 | |
| *** se6astian|away is now known as se6astian | 18:35 | |
| *** CarlFK has quit IRC | 19:24 | |
| *** rohitksingh has quit IRC | 19:35 | |
| *** CarlFK has joined #timvideos | 19:53 | |
| *** ChanServ sets mode: +v CarlFK | 19:53 | |
| *** se6astian is now known as se6astian|away | 21:36 | |
| *** Bertl_zZ is now known as Bertl | 22:28 | |
| *** Bertl is now known as Bertl_oO | 23:00 | |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!