Thursday, 2016-01-28

*** tpb has joined #timvideos00:00
*** tpb has joined #timvideos01:39
micolousbeep beep02:28
tpbTitle: wkoszek/fxload at multios: fxload from Linu... - GitHub (at
cr1901_modernmithro: I'm finishing sb0's scanwidget. The freq counter works, but it's not as accurate as I'd expect it to be. So I suspect a minor bug. Do you still want a pull request anyway?04:55
*** rohitksingh has joined #timvideos05:29
thaytanxfxf, how'd you go with the netclock?05:37
xfxfapologies - have been solving other complications. the ntp sync didn't work but given your explanation I'm not surprised05:44
xfxfwill compile up that code and ping you soon05:44
CarlFK                                 $6.54       normal price:   $13.0905:48
thaytanFor a while now, Jaycar have had a bin here with 2m HDMI cables for $1006:00
thaytanand ALDI were selling nice 8m ones recently. I've been using one of those for running 1080p from the mediacenter at the back of the room to the TV06:01
xfxfthaytan: turns out we need amp'ed cables, proving fun to locate06:05
CarlFKdeeprave: ?06:08
deepraveCarlFK: just added carl/ to the video-scripts repo. No longer uses flterm but uses pyserial (module needs to be installed obviously)06:08
deepraveas opposed to the flterm method, it actually works. uses a .1 sec delay between chars sent06:09
thaytanxfxf: cables with an amplifier in? never seen those06:09
xfxfthaytan: neither had i06:14
xfxfCarlFK: deeprave's script also works with the Opsis's, the thing you had before doesn't06:14
xfxfherm, trying to compile that c file, missing gst.h, thought it'd be in libgstreamer1.0-dev06:15
CarlFKit is the $6 I linked above06:16
thaytanxfxf, gcc `pkg-config --cflags gstreamer-1.0` `pkg-config --libs gstreamer-1.0` -o netclock-client netclock-client.c oughtta do it06:17
thaytanwhat distances are you talking that you need an amp in the cable?06:19
thaytanthat one is 1m long, which seems too short to make a difference06:19
xfxfja, ~1m.  ask mithro, seems every time i try explaining what i think the issue is i'm wrong ;)06:20
mithrothaytan: I'll explain over drinks at LCA06:24
thaytanoh, you've started drinking? :)06:24
mithrothaytan: This week has been going so well I might just start06:25
thaytan(things *must* be bad)06:25
CarlFKthaytan: we keep hearing things like "This hdmi splitter makes things much better when it works.  but getting it to work requires way to much fiddling so it's out."06:27
thaytanCarlFK, we saw a bit of that sort of thing in Dublin for GStreamer conf06:27
thaytanthe rooms had a HDMI transmitter/splitter to the projector and the ubicast streaming guys06:28
xfxfthaytan: still having issues compiling it, i was getting help but that person has been diverted06:28
thaytanand it'd work for some people and not others06:28
xfxfthaytan: oh man we have that to deal with too, dw, the uni equipment is _fun_06:28
thaytanxfxf, pastebin some cmdline output06:28
xfxfthat said the value of the hdmi2usb is once _that_ works with the stupid uni equipment it abstracts away that problem for anybody plugging into it06:29
CarlFKtumbleweed and I had 95% success with our prototype Atlys boards at PyConZA.  had to use twinpacts for 5 out of 50 talks06:31
tpbTitle: Pastebin | IRCCloud (at
CarlFKsome people couldn't get hdmi to work at all cuz they had smoked their hdmi port with fire or something.06:31
xfxfCarlFK: that's 90% not 95% :P06:31
CarlFKxfxf: I don't count the burning laptop as our problem06:32
xfxfi'm still impressed by that guy at pycon a bunch of years back with 5 travel converters plugged into each other who took out an entire circuit06:32
xfxfthere were sparks if i recall06:33
tumbleweedxfxf: and we managed 100% success at nodevember06:36
CarlFKnode is conf-ready-scale06:37
tpbTitle: Pastebin | IRCCloud (at
mithrodeeprave:  see that06:37
mithrodeeprave: "debug input0"06:38
mithrodeeprave: clone my mode switch repo -
tpbTitle: mithro/HDMI2USB-mode-switch at opsis-prod: Tool for switching boards supported by HDMI2USB firmware between multiple different modes (programming, webcam, etc). (at
mithrodeeprave: switch to the opsis-prod branch06:38
mithrodeeprave: run the ./ script as root06:39
tumbleweedxfxf: oh, another problem we had at pyconza was that linux desktop environments don't make it easy to mirror at resolutions that aren't native on the laptop panel06:39
tumbleweedI had to do xrandr magic06:39
*** tvCommitBot has joined #timvideos06:41
tvCommitBot[streaming-system] micolous opened pull request #111: Allow YouTube-only streaming. (
*** tvCommitBot has left #timvideos06:41
*** tvCommitBot has joined #timvideos06:42
tvCommitBot[streaming-system] mithro closed pull request #111: Allow YouTube-only streaming. (
*** tvCommitBot has left #timvideos06:42
xfxfthaytan: ping, sorry, not sure if you saw my last pastebin above (
tpbTitle: Pastebin | IRCCloud (at
tpbTitle: streaming-system/ at master · timvideos/streaming-system · GitHub (at
tpbTitle: Pastebin | IRCCloud (at
xfxfthaytan: changed the order of params, now getting this:07:07
tpbTitle: Pastebin | IRCCloud (at
xfxfwhich is odd because that appears to be part of gstreamer core07:07
xfxfthis is ubuntu 15.10 with libgstreamer1.0-dev installed07:07
tpbTitle: HDMI2USB-firmware-prebuilt/archive at master · timvideos/HDMI2USB-firmware-prebuilt · GitHub (at
mithrodeeprave: I think
tpbTitle: HDMI2USB-firmware-prebuilt/opsis_hdmi2usb-hdmi2usbsoc-opsis.bin at master · timvideos/HDMI2USB-firmware-prebuilt · GitHub (at
*** Bertl_oO is now known as Bertl07:24
deepravemithro: could you provide the complete link to the image?07:28
mithroI think it's the one just before that07:31
mithro_florent_: it's xapp495 apparently -
thaytanxfxf, we're getting into dinnertime here now - I'll bbl07:37
xfxfno worries07:38
xfxfMaZderMind is the developer, MaZderMind we may as well continue any discussion about this here so it's not confusing (thaytan is the gstreamer dev who has been helping me)07:38
xfxfthis might be relevant: <@MaZderMind> i don't know how to set the base-time correct for remote sources. as i said earlier we have never tested remote sources07:38
thaytanxfxf, add 'gstreamer-net-1.0' as a pkg-config pkg to pull in07:39
thaytandoes voctomix depend on the base time?07:39
thaytanhow does it normally distribute base time?07:39
xfxfMaZderMind: ^^07:39
thaytanI guess it would have to to get precisely synched mixing07:40
thaytanindependent of capture and transmission delays07:40
xfxfoh cool, the compile worked, thanks07:40
thaytanbut if you're sending matroska, don't the timestamps of any incoming feed always start at 0?07:40
mithrothaytan: it's suppose to be mostly independent of time07:41
thaytanoh, matroska preserves the input timestamps07:42
thaytanso that'd work07:42
xfxfthis is pretty much the only remaining issue preventing us from using vocto, btw07:42
xfxfso appreciate++ the help07:42
_florent_mithro: "TMDS_33 signals require 50Ω termination to 3.3V at the receiver."07:45
_florent_do we have it, or should it be provided by the FPGA?07:45
xfxfthaytan: is there any special way i should be calling netclock-client btw?07:47
tpbTitle: Pastebin | IRCCloud (at
mithro_florent_: they are in the schematic and loaded as far as I understand07:54
_florent_yes I just checked that07:54
mithro_florent_: actually, hold on - they are only on the input or output - can't remeber which one07:54
xfxfthaytan: nm, works07:57
xfxfthaytan: not sure what you're after, but:07:59
tpbTitle: Pastebin | IRCCloud (at
mithro_florent_: looking at that xapp, the UCF only has TMDS_33 specified08:02
*** se6astian has joined #timvideos08:07
thaytanxfxf, I want you to check if you get clock synch (stats output + synchronised=(boolean)true) when talking to the vocto clock IP + port08:15
xfxfthat is the vocto clock IP + port, and I can see the above08:16
xfxfin that pastebin08:16
xfxfsynchronised starts off as false, but then is true08:16
xfxfoh, actually, you said in the python script before trying to sync clocks to wait a bit?  i'll try that too, i forgot you mentioned that08:17
thaytanxfxf, I want to hear what MaZderMind has to say about Vocto's basetime requirements08:18
xfxfok, np08:19
thaytanone easy thing vocto can do there is to always use a basetime of 0 everywhere08:20
thaytanthat's easy to transmit a priori08:20
tpbTitle: In 4:3 PAL, Y41B confuses ffmpeg2theora, and dvswitch PinP by stefanor · Pull Request #15 · timvideos/dvsource-v4l2-other · GitHub (at
xfxfi rearranged the script so the sync_clock() is under the set_state(PLAYING), with a time.sleep(3) under it08:25
xfxfzero idea if this does what i want, but i'm getting video08:25
xfxfas you say i'm interested to see your discussion w/ MaZderMind08:25
xfxfi'm just guessing08:25
thaytanno, you have to set the clock before playing08:28
thaytanor else it has no effect and you're using the default clock, which should be the system clock08:29
xfxfokay, will revert that then08:29
xfxfa time.sleep() didn't make any difference alone btw08:29
thaytanonce the pipeline gets to playing state (confirmed by the state-changed bus mesage), you can see which clock it chose with pipeline.get_clock()08:29
*** tvCommitBot has joined #timvideos08:31
tvCommitBot[dvsource-v4l2-other] stefanor tagged v0.1.1 at master:
*** tvCommitBot has left #timvideos08:31
mithro_florent_: doesn't seem to have made any difference I can see08:35
MaZderMindI'm just now reading backlog here08:36
MaZderMindthaytan: Up until now we used voctomix just on one system, so all pipeliens running on the same systemclock08:37
MaZderMindthaytan: In the nettime-branch (see I added a SystemClock to the core and really forced all pipelines to use this. I then added a NetTimeProvider which provides this clock onto the network. I then added code to the GUI to obtain this clock from the server and apply it to all pipeliens.08:43
tpbTitle: Use GstNetworkTime to sync core, gui and sources by MaZderMind · Pull Request #51 · voc/voctomix · GitHub (at
thaytanand a basetime of 0?08:44
mithro_florent_: was a nice try :(08:44
mithro_florent_: I'm compiling again locally just to make sure08:45
MaZderMindthaytan: hm i was unsure about this.
tpbTitle: Use GstNetworkTime to sync core, gui and sources by MaZderMind · Pull Request #51 · voc/voctomix · GitHub (at
MaZderMindthaytan: i'Ve set it to the current sysrtem clock time. Haven't had much time to read the docs or to hack on voctomix atm08:46
CarlFK this is all we need to touch, right?  byte 2:  bits 0-2: frame aspect ratio = 0 (4:3), 1 (16:9 letter-boxed in 4:3) or 208:48
CarlFK  VAUX source control (VSC)  Appears as pack 10 in VAUX block 2 (sequence block 5) in even  sequences and pack 1 in VAUX block 0 (sequence block 3) in odd  sequences.08:49
tpbTitle: DV format (at
MaZderMindthaytan: would you mind hopping to #voctomix so we can split the hardware-hacking and the software-hacking?08:55
MaZderMind#voctomix is currently the LCA-Voctomix-HelpChannel of choice ;)08:55
CarlFKxfxf: If you are stopping by a store, can you get me some cough drops - I'm  starting to cough just enough. but I have 2 with me, so don't delay  dinner with a special trip.09:04
*** tvCommitBot has joined #timvideos09:05
tvCommitBot[streaming-system] micolous opened pull request #112: Port monitor.html to bootstrap, fix minor issues in group.html (master...monitor-strap)
*** tvCommitBot has left #timvideos09:05
mithro_florent_: hrm, the memory error is back09:05
mithro_florent_: did you get anywere with the "quick target" ?09:14
_florent_I think we can simply use VideoMixerSoC09:15
_florent_last time I tried P&R was fast09:16
thaytanMaZderMind, you're not there :)09:41
MaZderMindthaytan: ah :( #voctomix on hackint09:46
thaytanyou've doomed us all!09:47
thaytanwe've already split the continuum09:47
thaytanalso, I've never heard of hackint09:47
MaZderMindok, nevermind09:48
MaZderMindthen we'll stay here ;)09:48
MaZderMindthaytan: <- I'm trying to get this working on the local host first09:49
tpbTitle: 0bin - encrypted pastebin (at
*** tvCommitBot has joined #timvideos10:09
tvCommitBot[streaming-system] mithro closed pull request #112: Port monitor.html to bootstrap, fix minor issues in group.html, purgo old assets (master...monitor-strap)
*** tvCommitBot has left #timvideos10:09
mithro_florent_: any idea what is up with the mem issue?10:13
*** tvCommitBot has joined #timvideos10:14
tvCommitBot[dvsource-v4l2-other] stefanor pushed 2 new commits to master:
tvCommitBotdvsource-v4l2-other/master afee7cb Stefano Rivera: Bump version in changelog10:14
tvCommitBotdvsource-v4l2-other/master 9a4d2ba Stefano Rivera: That looks like an accidental copy-pasta10:14
*** tvCommitBot has left #timvideos10:14
_florent_mithro: no, I've never been able to reproduce it10:17
*** tvCommitBot has joined #timvideos10:26
tvCommitBot[dvsource-v4l2-other] stefanor pushed 1 new commit to master:
tvCommitBotdvsource-v4l2-other/master 52aa647 Stefano Rivera: The package containing gst-inspect is not called gst-inspect10:26
*** tvCommitBot has left #timvideos10:26
mithroit seems to come and go10:29
CarlFK# export GST_DEBUG=410:57
deepravemithro: looks like I won't be getting back tonight due to my newly arrived wife's bidding; will make it an early morning instead. Going to look at tkinter, might be easy to do what is needed there11:07
mithrodeeprave: okay11:07
_florent_mithro: here I'm not able to get WER != 011:25
_florent_tested with different cables and sources11:25
mithro_florent_: :-/11:25
_florent_have you tried with another power plug?11:27
xfxfdeeprave: np11:32
mithro_florent_: btw I get the memtest failures with both Atlys and Opsis boards atm11:34
mithro_florent_: oh wait, this is my Atlys 50MHz branch11:38
*** CarlFK has quit IRC11:39
*** tvCommitBot has joined #timvideos11:40
tvCommitBot[streaming-system] micolous opened pull request #113: Merging @funkybob's 1.6+ port to master WIP (master...grand-funk-railroad)
*** tvCommitBot has left #timvideos11:40
_florent_ah yes, didn't think about that, but yes DDR won't be working with Atlys at 50MHz without modification11:40
_florent_it's probably better to stay with 75MHz11:41
*** CarlFK has joined #timvideos12:30
*** ChanServ sets mode: +v CarlFK12:30
mithrocr1901_modern: sb0__: yeah, I've been following that along - I notice that you don't13:36
mithrodidn't mean to do that13:37
tpbTitle: HDMI2USB-firmware-prebuilt/archive/lca2016 at master · timvideos/HDMI2USB-firmware-prebuilt · GitHub (at
*** se6astian is now known as se6astian|away16:12
*** rohitksingh has quit IRC17:45
*** rohitksingh has joined #timvideos17:59
*** se6astian|away is now known as se6astian18:16
*** hyades has joined #timvideos19:41
*** rohitksingh has quit IRC20:37
*** micolous has quit IRC21:08
*** micolous has joined #timvideos21:08
*** MaZderMind has quit IRC21:39
*** MaZderMind has joined #timvideos21:40
*** se6astian is now known as se6astian|away22:32
*** hyades has quit IRC22:36
*** Bertl is now known as Bertl_zZ22:38
CarlFKtumbleweed: can you pick up some cough drops for me? lemmon, honey, cherry, vegemite, don't care what flavor and stuff.22:44

Generated by 2.13.1 by Marius Gedminas - find it at!