Monday, 2015-09-28

*** tpb has joined #timvideos00:00
CarlFKmake gateware ... make[1]: *** No rule to make target `../../software/include/generated/cpu.mak'.  Stop.00:35
*** raron has quit IRC01:09
*** raron has joined #timvideos01:09
*** springermac_ has joined #timvideos02:14
*** springermac has quit IRC02:15
*** springermac_ is now known as springermac02:15
*** Oswald has joined #timvideos02:19
*** Oswald has quit IRC03:06
CarlFKmithro: why did you remove gateware-$(TARGET) from
tpbTitle: Makefile: fix build on Windows · timvideos/[email protected] · GitHub (at
*** raron has quit IRC03:27
mithroCarlFK: I think that was a mistake03:40
CarlFKmithro: not sure how travis passed it, cuz it broke here.  I had to git checkout HEAD~ two times, now it builds04:16
mithroCarlFK: I think because the travis script does a "make firmware" before it does a "make gateware"04:17
mithroCarlFK: guess it might need to do a "make clean" between the two or something04:18
CarlFKmithro: is there a way for me to tell (visually, dmesg, whatever) that this loaded?  PROG=fpgalink make load-gateware04:23
mithroCarlFK: not at this moment04:24
mithroCarlFK: I think you can check an LED on the FPGA - but I don't know which one04:24
CarlFKmithro: remember you had me add some extra parameters to dvsource-v4l2-other  ?04:38
mithroCarlFK: nope!04:38
CarlFKit was before I recorded the 1 talk I have recorded.  something about color space?04:39
mithroNope! Check the logs?04:40
tpbTitle: Search IRC logs (at
CarlFKneat - i didn't know there was a search04:44
mithroCarlFK: also why we should try and keep everything to the public channel as possible04:53
mithroCarlFK: looking at from what I can see in the logs the colorspace stuff got merged04:54
tpbTitle: Making the color space work on all modes. · timvideos/[email protected] · GitHub (at
CarlFKmithro: thanks - suddenly got super busy here, having trouble focusing on anything :p04:55
*** CarlFK has quit IRC07:31
*** CarlFK has joined #timvideos07:56
*** ChanServ sets mode: +v CarlFK07:56
*** rohitksingh has joined #timvideos08:53
*** hamster_ has joined #timvideos09:59
hamster_@mithro - borrowed the platform cable from work, and finally got designs up and running. :-)10:01
*** hamster_ has quit IRC10:07
*** seb_ has joined #timvideos11:14
seb_i would like to understand what is realy done by the numato opsis card11:15
seb_is it only mix and switch different input and output or is it able to encode stream too?11:16
mithroseb_: What you mean by "encode" the stream?11:24
seb_i mean give us a compressed video file from uncompressed stream11:25
seb_but i think i've my answer, it's more a passthru system11:26
mithroseb_: The USB / Ethernet output is mjpeg as you can't fit a raw 720p60 video through either of those interfaces11:27
*** seb_ has left #timvideos11:27
mithro_florent_: I'm around now if you have some time to chat11:27
*** seb_ has joined #timvideos11:28
seb_ok, so the card encode in mjpeg the hdmi input11:29
seb_can it deal with 4k videos?11:29
seaLnethe hdmi inputs can't but the dp ones could11:37
seaLne(from what i've seen discussed before)11:38
seb_and what about the outputs?11:38
seaLnesame i think11:38
seb_i'm a backer of the apertus axiom beta project (
seb_and in their goal there is a 4k raw pc recording option using hdmi interface11:42
mithroseb_: We have not encoded mjpeg at 4k rates and even with mjpeg compression it would be impossible to ship 4k video out the USB / Ethernet interface - there isn't enough bandwidth11:42
mithroseb_: I'm actually hoping to finish a backer update which explains stuff around resolution a bit more11:43
mithroIt's a complicated problem11:43
mithroseb_: You have to think about things like memory bandwidth too11:43
seb_you both work around opensource hardware with FPGAs, maybe there can be interesting interaction between you ;)11:44
seaLnecould theoretically it spit out webm instead of mpeg (not solution to resolution, would just save transcoding to a less restricted format)11:45
seaLneor in the future other formats11:45
seaLnenot sure how much mjpeg is tied to hdmi11:46
seb_and for direct ssd save through the expansion port?11:46
mithroseb_: I'm in the apertus IRC channel too :)11:48
seb_mithro: you're working on the timvideo project?11:49
mithroseb_: I started the project :)11:50
seb_ho, ok :) respect11:50
seb_did you know the apertus project?11:51
seb_do you already have contact with them?11:51
mithroseb_: yes, as I said I'm in the apertus IRC channel right now :)11:57
seb_ok, cool :)12:01
seb_did you participate to their crowdfunding campain?12:02
mithroseb_: No, sadly didn't have the funds for that12:05
mithroseb_: was very tempted though :)12:07
mithroseb_: I've already discussed with BlueCmd about an adapter board to make apertus expansion cards compatible with the Opsis12:08
mithroseb_: sadly, they didn't choose compatible pins - I guess in their case space was a premium12:09
seb_if you can be compatible with their 4K RAW output with experimental HDMI formats that could be very cool12:11
seb_for recording thought your expansion port maybe?12:12
mithroseb_: A purposed build expansion card which interfaces with the apertus would definitely be possible. As there are FPGAs on both sides of the interface you could do pretty much anyting.12:13
seb_if you can just use common FPGA libs it can be interesting for future development12:14
mithroseb_: yeah, I think they are concentrating on "getting things done" at the moment12:14
seb_yes :) they are a litle bit late on their planing :p12:15
*** seb_ is now known as troisdsman12:20
mithrotroisdsman: but we will try and collaborate as much as possible :)12:20
troisdsmancool :)12:21
mithrotroisdsman: of course we are a much smaller project then they are with substantially less funding12:22
mithrotroisdsman: no big Hollywood names giving us money :(12:23
mithrotroisdsman: The more successful the Opsis is, the more ability we'll have to do cool things in the future12:25
troisdsmanyes :)12:28
mithrotroisdsman: so please do get all your friends to back us :)12:28
mithrotroisdsman: IE we'd love to develop a well tested, performant webm encoding core - but we just don't have the resources currently12:29
troisdsmanfor our utilisation your project is a little bit small :/12:37
troisdsmanwe're working on a setup to record 4 flux in realtime (3 camera and one hdmi output)12:38
troisdsmanis the numato opsis able to get 2 hdmi flux in realtime through usb?12:44
troisdsmanstream sorry13:25
mithrotroisdsman: Fitting 2 hdmi streams through USB2.0 isn't possible without significant lossy compression. We can probably get two streams out of the Gigabit Ethernet or another option would be via a PCI-Express adapter card connected to the GTP transceivers13:32
troisdsmani'll talk of this option to my coworkers13:33
troisdsmanthanks for your help!13:44
*** troisdsman has left #timvideos13:54
*** rohitksingh has quit IRC14:20
*** rohitksingh has joined #timvideos14:36
*** CarlFK has quit IRC16:35
*** Sewar has quit IRC18:41
*** Sewar has joined #timvideos18:41
*** miselin has quit IRC21:20
*** miselin has joined #timvideos21:24
*** springermac has quit IRC22:13
*** springermac has joined #timvideos22:15
*** raron has joined #timvideos23:12

Generated by 2.13.1 by Marius Gedminas - find it at!