Friday, 2016-03-25

*** tpb has joined #timvideos00:00
*** seaLne has quit IRC00:08
*** seaLne has joined #timvideos00:10
*** Neuron1k has quit IRC00:21
*** panther has joined #timvideos02:01
panthermithro, hi02:04
*** panther is now known as sab_12302:04
CarlFKhi sab_12302:55
*** panther has joined #timvideos03:03
*** sab_123 has quit IRC03:03
mithroI'll be around in about an hour03:18
mithroI'm around now04:05
panthermithro, cool04:06
*** panther is now known as sab_12304:06
sab_123i still dont understand the MPSSE04:06
sab_123how is the data packet actually sent04:06
sab_123where is the doc for this04:06
sab_123or is it not documented at all?04:06
sab_123For example04:08
sab_123section 3.304:08
sab_123will USB FIFO have04:08
sab_1230x10 0xlengthl 0xlength 0xdata04:09
sab_123is this how it is?04:09
sab_123I bit bang 0xdataa out04:09
sab_123and what is no read in these sections04:10
sab_123mithro, I meant the data in the buffer04:18
mithrosab_123: I believe so04:18
sab_123because without that04:19
sab_123we really dont know how ot handle this data04:19
sab_123JTAG seems to have byte mode, mithro04:19
sab_123dont understand that04:19
sab_123they say it is only single pin, mithro04:19
mithrosab_123: what do you mean? The commands are like "clock this data out", "clock this data in", "clock this data out and in at the same time"04:21
sab_123i mean that JTAG is 4 pins04:22
sab_123but they clock data byte wise04:22
sab_123and they have a bit mode04:22
sab_123so I dont understand04:22
sab_123and this is just the device side04:22
sab_123not the usb side, mithro04:23
mithroByte just means you have to clock it out 8 bits04:23
sab_123but there arent enough pins04:23
sab_123for JTAG, mithro04:23
mithrosab_123: what do you mean there aren't enough pins?04:23
sab_123mithro, JTAG is a 4 pin interface04:23
sab_123TDO, TMS,TCK and something else04:24
sab_123THis data sent from libftdi is now in FX2's buffer04:24
mithroThere is TDO, TDI, TCK and TMS/CS mentioned in that document04:24
sab_123those are single pins04:24
sab_123not a bus04:24
sab_123so how do you clock a byte out04:24
sab_123you can only clokc bits04:24
mithroSame way you would clock out 8 bits04:24
sab_123mithro, there seems to be something I am missing here04:25
sab_123libftdi on computer sends a command saying 0xcommand 0xlength 0xdata04:25
sab_123so now lets say 0xcommand is section 3.304:25
sab_123clock data out byte wise04:25
mithroIt means that the data is just whole bytes, so you are always going to end up clocking out multiples of 8 bits04:27
*** hyades_zombie has quit IRC04:29
sab_123confusing nomenclature04:29
sab_123byte wise means send 8 bit data on a bus04:29
sab_123anyway, mithro04:30
sab_123I have submitted the proposal04:30
sab_123mithro, if you have any suggestions, please let me know04:30
*** hyades_zombie has joined #timvideos04:30
mithroxfxf: ping?04:46
mithroxfxf: Have you seen a video example or something which shows a mixing desk controlling the fades using one of the fade sticks?04:47
xfxfi've used proprietary gear?04:47
xfxfso, directly, yes04:48
mithroxfxf: Just want an example which shows the wipes control by a hardware slider04:48
xfxfooh sorry04:48
xfxfyou mean do i know of a video04:48
xfxfi thought it was a wierd question :)04:48
xfxfit's a DIY thing but shows the fader04:51
mithroxfxf: thanks, that is perfect!04:52
mithroxfxf: so, you going to do hacking then? :P05:04
*** Neuron1k has joined #timvideos05:10
*** tija has joined #timvideos05:48
*** Bertl_zZ is now known as Bertl06:04
*** Neuron1k has quit IRC06:04
*** se6astian|away is now known as se6astian06:48
*** se6astian is now known as se6astian|away06:53
*** panther has joined #timvideos06:57
*** sab_123 has quit IRC06:57
xfxfmithro: prob gonna poke at the setup and vocto at some point over the weekend, ya06:58
xfxfand test that fix you wanted me to06:58
*** panther_ has joined #timvideos07:01
*** panther_ has quit IRC07:03
*** panther_ has joined #timvideos07:04
*** panther has quit IRC07:04
*** ssk1328 has joined #timvideos07:26
*** aom has joined #timvideos07:40
mithrohello aom07:54
mithroaom: So, just wanted to say I like your proposal quite a bit even though I've put quite a lot of comments on it08:09
aomAbout developing environment, I must choose migen ? or better migen?08:09
aomthanks for many comments. now i modify them.08:10
mithroaom: Unless there is a significant reason not to use migen, we should be using migen08:10
mithroaom: We can use Verilog, but only if there is a large project that we want to reuse - like the lm32 or the or1k cores08:10
mithrowe shouldn't use VHDL, as we don't have good FOSS tools for supporting them08:11
aomOkay! i will learn migen developing skill!08:11
mithromigen should be pretty easy to pick up if you know either VHDL or Verilog08:11
mithroYou will also probably find it much quicker to do development in once you are use to it08:11
mithroaom: Did you understand the idea of reusing the existing LiteEth stuff and the RTP headers instead of doing the Ethernet MIO stuff?08:12
mithroaom: Did that make sense?08:12
aomyes, I understand!08:13
mithroaom: It's cool that you have 4 Atlys boards :) - I'm down to 3 I think (I've lent mine out to people)08:14
aomyes! and I have other xilinx boards. and i have developing tools like ISE. so I can easily test the hardware!!08:15
mithroaom: while initially during development you can do things like hard code IP addresses and stuff, it would be good as an extension goal to allow dynamic discovery and configuration using something like Avahi / mDNS08:16
mithroaom: I'd also recommend giving building the HDMI2USB-misoc-firmware building a go (after you have finished your proposal) - See instructions here ->
tpbTitle: HDMI2USB-misoc-firmware/scripts at master · timvideos/HDMI2USB-misoc-firmware · GitHub (at
aomthat is greate ! i will check it.08:17
mithroaom: _florent_ in this channel is also the developer of LiteEth and most of the firmware, so if you have specific questions about how things work, he is a good person to ask08:18
*** seaLne has quit IRC08:19
*** seaLne has joined #timvideos08:19
mithroaom: You might also consider with this approach that you can develop one side at a time and might be able to use a normal computer on the otherside too08:19
mithroseaLne: ping?08:19
mithroaom: might make things a bit easier while developing stuff08:19
mithroaom: You probably want to start with something like 640x480 when debugging as well (before getting 720p working)08:20
aomsorry, what do you mean the approach?? i cannot get it.08:22
mithroaom: If you are using RTP, it is easy for you to go FPGA->Ethernet->Computer rather then FPGA->Ethernet->FPGA08:23
aomi understand it! exactly, its better to reuse existing LiteETH and RTPtimestamper.08:26
mithroaom: any questions before I disappear for a couple of hours?08:26
aomnothing now! anyway I complete my document.08:27
aomThank you for much feedback!08:27
mithroaom: feel free to ask in the channel if you do, I'm try and answer when I get back08:28
*** ssk1328 has quit IRC08:28
*** Bertl is now known as Bertl_oO08:37
*** sab has quit IRC08:53
*** thaytan has quit IRC09:18
*** Sewar has quit IRC09:41
*** aom has quit IRC09:44
*** Sewar has joined #timvideos09:49
*** Sewar has quit IRC09:53
*** hyades_zombie has quit IRC09:53
*** hyades_zombie has joined #timvideos09:54
*** Sewar has joined #timvideos10:00
*** hyades_z- has joined #timvideos10:01
*** kflux has joined #timvideos10:05
*** hyades_zombie has quit IRC10:06
*** kflux has quit IRC11:03
*** thaytan has joined #timvideos11:52
*** ChanServ sets mode: +v thaytan11:52
*** ssk1328 has joined #timvideos12:06
ssk1328mithro: The original github issue of my project <>, referred to issues 27 of misoc-firmware issue tracker12:09
tpbTitle: [HDMI2USB] Add "hardware mixing" support to HDMI2USB firmware · Issue #34 · timvideos/getting-started · GitHub (at
tpbTitle: Support mixing the two HDMI inputs together · Issue #27 · timvideos/HDMI2USB-misoc-firmware · GitHub (at
ssk1328This particular issue tracker referred to developing some test cases for verifying the working of board.
tpbTitle: Embed a wide range of test patterns · Issue #185 · timvideos/HDMI2USB-misoc-firmware · GitHub (at
ssk1328That's why I added a section for overall testing of board issues like overscan/underscan/aspect ratio in my timeline(week11)12:15
mithrossk1328: as you are only mixing and not transforming,overscan/understand/aspect ratio shouldn't be a problem for you12:27
mithrossk1328: you should be more concerned with color mixing12:27
ssk1328I will add joystick and scaling things instead of that.12:28
mithrossk1328: the joystick thing is more about having something cool to show off :)12:28
ssk1328Yeah. Pretty Cool :)12:28
mithrossk1328: eventually we'd like to support things like MIDI and DMX controllers for doing the fades12:30
mithrossk1328: if we get your and aom's stuff working together, that will also be super awesome12:30
*** tija has quit IRC13:38
*** Neuron1k has joined #timvideos14:25
mithro_florent_: ping?14:37
mithrowb Neuron1k14:37
Neuron1kHi mithro14:37
_florent_mithro: yes?14:45
mithro_florent_: any idea how much work it would be to get the HDMI2USB firmware onto master misoc/migen? I saw you have ported a lot of your stuff recently?14:45
_florent_most of the work will be around the DMA since some modules have been removed, but in a first we can copy these modules in HDMI2USB14:51
_florent_doing that it's probably a few days of work14:52
mithro_florent_: maybe that is worth trying to do before we have GSoC students hacking on things?14:53
*** rohitksingh has joined #timvideos15:24
*** ssk1328 has quit IRC15:50
*** jp111 has joined #timvideos16:07
*** jp111 has quit IRC16:11
*** panther has joined #timvideos16:36
*** panther_ has quit IRC16:36
*** panther has quit IRC16:47
*** rohitksingh has quit IRC17:38
*** rohitksingh has joined #timvideos18:11
*** Neuron1k has quit IRC18:57
*** vikram9866 has joined #timvideos20:06
*** vikram9866 has quit IRC20:09
*** rohitksingh has quit IRC20:12

Generated by 2.13.1 by Marius Gedminas - find it at!