Sunday, 2014-07-13

*** tpb has joined #gst-switch00:00
mithrohey derpeter11:59
mithroderpeter: what gstreamer pipe are you trying?11:59
derpetermithro: hey. I have now moved with the testing over to snowmix so sadly my pipe ist not in the bach histroy anymore but it was very basic pipe thats has ended in the tcpclient source. Do you have somewhere a working pipe for file -> gst-swich?12:20
derpeterthe testing weekend is nearly over but i would live to give gst-swtich a second shot12:20
mithroderpeter: only what is in the documentation12:20
mithroderpeter: how's snowmix going?12:20
mithrogst-switch probably needs a lot more love then snowmix, but is probably more flexible12:25
derpetermithro: the pipes from the googledoc document didn't work in my tests so i tried to reduce them but i'm still in the process of understanding the pipes.12:36
mithroderpeter: which google doc? It's been quite a while since I've done anything with gst-switch12:37
derpeterat least snowmix examples work in our test. We know need to build pipes that work with our hardware (decklink to gestreamer)12:37
derpetermithro: https://docs.google.com/a/mithis.com/document/d/1_blj8u6ToyaZwdikCmUSGFwdpoAYNCyDYEHgXtPL8mI/edit?pli=1#heading=h.bhhck1xjratt12:38
tpbTitle: gst-switch - GStreamer based DVSwitch replacement - Google Docs (at docs.google.com)12:38
derpetermayb there is more that i didnt found?12:38
derpeteras far as i understand it dst-switch und snowmix have very similar design concepts12:39
mithroderpeter: yeah, they came to being at roughly similar timelines too12:39
mithroderpeter: did you take a look at https://github.com/timvideos/gst-switch/blob/master/tests/feed-test-video.sh ? and related files?12:41
tpbTitle: gst-switch/tests/feed-test-video.sh at master · timvideos/gst-switch · GitHub (at github.com)12:41
derpetermithro: no but looks interesting12:42
mithroderpeter: the pipelines in that document where theoretical ones for when designing the system12:42
derpeteri wont be able to test it today but we will have a confernce in august where i have one week of preparation where i hope i can do more testing12:42
derpetergoal is to have a fullhd pipline tested for 31C312:43
mithroderpeter: are you planning on running the fullhd pipeline in parallel to your non-hd pipeline?12:44
derpetermithro: the final goal would be to drop SD in the pipline. Last year we build a fullHD pipeline that worked very well but used a hardware mixer12:45
mithroderpeter: you want to replace the hardware mixer?12:45
derpeterso in theory we only need to replace the mixing software. everthing else is ready to move from rawdv to mpeg ts12:46
derpetermithro: the hardware mixer was rented12:46
mithroderpeter: and expensive I assume? :P12:46
derpetermithro: yes but price it not the only argument. One other point is that we have have framgrabber that only speak IP so we cant use them with an SDI mixer12:47
derpeterwith a software mixer we are much more flexible on the sources12:47
mithroyes, I understand the reason for a software mixer :) - I did pay for one to be developed12:48
mithrothat IP framegrabber are you using?12:48
derpeteri.e. we have pause animation that we put as input in dv-swtich. Doing this on the SDI mixer requires SDI hardware and an extra PC12:48
mithroderpeter: yeah12:48
derpeterhttp://www.epiphan.com/products/product.php?pid=10512:49
tpbTitle: VGA2Ethernet | Frame grabbers (at www.epiphan.com)12:49
derpeterthe only real alternative to this devices (exept the your grabber we are very interested in) are professional scanline converter wich are very expensive12:50
mithro$US 1,719.95 - ouch!12:51
derpetermithro: scanline convert are bejond 4k € :-(12:51
mithroderpeter: If only I could quit my job and do this full time :P12:51
derpeterthe epiphan device have the nice feature that they gernerate on the output side a new video signal. This is very nice because we can have digital video in and give than VGA to the beamer (which is in most locations the only way to speak to the beamer)12:52
derpetermithro: hehe same here12:53
mithroderpeter: yeah12:53
mithroderpeter: The Epiphan devices all work in very similar ways to the things I'm designing12:53
derpetermithro: :-912:53
derpeter:-)12:53
mithroderpeter: but generally our principle is to try and not touch the video signal12:53
mithroderpeter: as if everything goes wrong, we don't want to stop the presentation12:54
derpetermithro: the problem of converting to vga should go away in the future so that no show stopper12:54
mithroderpeter: do you know any C coders who know gstreamer and might be looking for 2-3 days work a week and would be willing to do it for $450 USD (maybe somewhere in eastern europe?)12:55
derpetermithro: are you involed with a conference or whats your motivation12:55
mithroderpeter: my "end game" is to get every presentation about FOSS in the world recorded (and live streamed)12:56
mithroderpeter: I started trying to put "dvswitch+recording set up in a box"12:57
mithroderpeter: s/box/suitcase/12:57
derpetermithro: in fact i know someone who should have the skills. i will ask him next week12:57
mithroderpeter: I'd be very interested to see how you go with snowmix, can you report back your experiances?12:58
derpetermithro: sure.12:59
derpetermithro: i stared with recording the froscon some years ago. now we are a team that tries to record all ccc events and the froscon13:00
mithroderpeter: cool13:01
derpetermedia.ccc.de show most of our recordings. (frontend is still wip)13:01
mithroderpeter: I've been involved with Linux.conf.au and PyCon US (plus PyCon AU which I started)13:01
mithroderpeter: if you know VHDL / Verilog hackers, that would also be useful :)13:01
derpeterah we will record the europython next week13:02
derpeterbut there we will user dv-switch13:03
mithroderpeter: cool13:05
mithroderpeter: have you guys got an Atlys board yet?13:05
derpeternot yet13:06
derpeteras we have 4 of the epihan devices a new grabber is not high prio. new mixer is needed first13:07

Generated by irclog2html.py 2.12.1 by Marius Gedminas - find it at mg.pov.lt!