| *** tpb has joined #timvideos | 00:00 | |
| *** sb0_ has joined #timvideos | 00:13 | |
| *** sb0__ has quit IRC | 00:15 | |
| *** sb0_ has quit IRC | 00:28 | |
| *** sb0 has joined #timvideos | 00:43 | |
| *** sb0 has quit IRC | 00:49 | |
| *** sb0 has joined #timvideos | 00:52 | |
| *** sb0 has quit IRC | 02:09 | |
| *** CarlFK1 has joined #timvideos | 05:21 | |
| *** ChanServ sets mode: +v CarlFK1 | 05:21 | |
| *** CarlFK has quit IRC | 05:37 | |
| *** CarlFK1 has quit IRC | 06:08 | |
| *** tpb has joined #timvideos | 06:29 | |
| *** CarlFK has joined #timvideos | 06:43 | |
| *** ChanServ sets mode: +v CarlFK | 06:43 | |
| *** rohitksingh has joined #timvideos | 12:05 | |
| *** sb0 has joined #timvideos | 14:03 | |
| *** [d__d] has quit IRC | 14:04 | |
| *** [d__d] has joined #timvideos | 14:05 | |
| mithro | _florent_: morning? | 15:09 |
|---|---|---|
| CarlFK | mithro: idea: xfxf likes to put a scaler? in front of the recording gear. this will muck up tests, but should give us 100% coverage for recording. | 15:23 |
| mithro | CarlFK: Yeah - I want to replace xfxf's need for a scaler | 15:24 |
| CarlFK | of course, but till then, like when I do pycon.za in two weeks, it would be good to get recordings out | 15:24 |
| CarlFK | I plan on trying witout, if no luck, switch to scaler | 15:25 |
| _florent_ | hello | 15:37 |
| _florent_ | you are still here mithro? | 15:38 |
| mithro | _florent_: yes | 15:38 |
| mithro | _florent_: That diagram you send me is awesome! | 15:38 |
| mithro | _florent_: what did you create it in? | 15:38 |
| _florent_ | http://sourceforge.net/projects/dia-installer/ | 15:39 |
| tpb | Title: Dia Diagram Editor | SourceForge.net (at sourceforge.net) | 15:39 |
| _florent_ | I think I fixed EDID issue (works not on one of my computer that was not working) | 15:41 |
| _florent_ | it could be interesting to do more tests with others video sources | 15:41 |
| _florent_ | all the video sources I'm able to test are working | 15:42 |
| mithro | _florent_: ahh cool, can we get the diagram committed to the repository? | 15:42 |
| mithro | _florent_: Can you walk CarlFK through using the EDID capture stuff you created so he can get some captures of non-working hardware? | 15:42 |
| _florent_ | the diagram is already in the repo | 15:43 |
| _florent_ | https://github.com/timvideos/HDMI2USB-misoc-firmware/commit/e2950db77487208a38328e1b02cfc98bab9c9b92 | 15:43 |
| tpb | Title: doc: add gateware architecture · timvideos/HDMI2USB-misoc-firmware@e2950db · GitHub (at github.com) | 15:43 |
| mithro | _florent_: oh great, still going through email :) | 15:43 |
| _florent_ | I clearly saw something wrong on EDID, so first it would be interesting to test if it's better | 15:44 |
| CarlFK | woot! give me 30 min or so to get setup. | 15:44 |
| _florent_ | if there are still issues, I'll explain how to debug EDID lines | 15:45 |
| _florent_ | hello CarlFK | 15:45 |
| _florent_ | last firmware is here: http://enjoy-digital.fr/open_hardware/hdmi2usb/atlys_hdmi2usb_gateware_2015_09_09.tar.gz | 15:45 |
| CarlFK | hi _florent_ - thanks! | 15:45 |
| CarlFK | _florent_: where do you live? | 15:46 |
| _florent_ | in France | 15:46 |
| _florent_ | Brittany | 15:46 |
| CarlFK | neat | 15:46 |
| _florent_ | CarlFK: your dead laptop is the one that was not working correctly with the firmware right? | 15:49 |
| mithro | _florent_: It would be good to have the instructions even if CarlFK's stuff starts working so when we do run into weird stuff we can debug it without needing to bother you again :) | 15:51 |
| CarlFK | _florent_: one of them. and I replaced it with the same model. but the one I have been testing with is still alive. and my wife's doesn't work, and the pi/distro is werid. | 15:51 |
| _florent_ | ok, please share your results, if it's not working I'll prepare something to investigate | 15:53 |
| mithro | _florent_: I'm pretty sure we'll find more devices which don't work (although, hopefully, it will be increasingly be bad devices rather than us) | 16:12 |
| _florent_ | yes, but there are two type of issues with EDID: 1) device is not able to read the port informations (I2C is not working correctly in the gateware) | 16:18 |
| _florent_ | 2) device is able to read the port informations, but does not support the resolution | 16:19 |
| _florent_ | For 1) we need to get I2C working correctly for each situation, I fix a point on that | 16:19 |
| _florent_ | but not sure we have to keep a permanent core to analyse that in the gateware | 16:20 |
| _florent_ | fix/fixed | 16:20 |
| *** TobiX has left #timvideos | 16:37 | |
| *** sb0_ has joined #timvideos | 16:48 | |
| CarlFK | _florent_: can you make the encoder on by default? | 16:48 |
| CarlFK | _florent_: mithro: \o/ woot! negk input1: 1280x720 | 16:50 |
| *** sb0 has quit IRC | 16:51 | |
| CarlFK | and pi! \o/ | 16:53 |
| _florent_ | cool! | 16:54 |
| CarlFK | wow, even pi's initial color splash (video hardware initialization I think) and char mode boot scroll | 16:54 |
| _florent_ | https://github.com/timvideos/HDMI2USB-misoc-firmware/commit/70f8e1e9761ac7204dd1efc2a6f39c2598823eb8 | 16:55 |
| tpb | Title: firmware: enable encoder at startup (also move encoder service to pro… · timvideos/HDMI2USB-misoc-firmware@70f8e1e · GitHub (at github.com) | 16:55 |
| CarlFK | using the enocoder to monitor because I don't have a monitor handy | 16:55 |
| CarlFK | new laptop works \\o// | 16:56 |
| _florent_ | you have others video sources to test? :) | 16:57 |
| CarlFK | working with bbb! (Beagle Bone Black ) (not sure if this one had problems) | 17:01 |
| mithro | _florent_: at some point we should look at redoing the USB interface and make the CDC-ACM stuff work again | 17:02 |
| CarlFK | works with Sheila's! ~\o/~ (I am very excited!) | 17:03 |
| CarlFK | I think that is everything I have in here | 17:04 |
| CarlFK | has anyone looped the Altys out into in? | 17:04 |
| _florent_ | no, can you try it? (it should work, but it will ignore EDID since EDID is not implemented on outputs) | 17:05 |
| CarlFK | Sheila will yell at me if I keep testing things. she wants to go to lunch :p | 17:06 |
| _florent_ | but on the atlys you will probably have to desactivate an hdmi_out or the encoder to get it work (not enough DDR bandwidth) | 17:06 |
| CarlFK | I have seen 1 in and 3 out working fine (a week ago I think) | 17:07 |
| CarlFK | oh oh.. someting not working :( | 17:10 |
| _florent_ | what? | 17:10 |
| CarlFK | negk sees a display, encoder is showing mostly black with a vertical line of pixels about 11 blotches ... random stuff basically | 17:11 |
| CarlFK | lunc - bb in about an hour | 17:11 |
| _florent_ | mithro: sure, not sure I'll have to do it in the next weeks, but I can help if someone want to do it | 17:18 |
| _florent_ | I'll have/I'll have time | 17:19 |
| mithro | _florent_: it's not a high priority right now - getting the EDID stuff working / debugged is more important | 17:19 |
| *** rohitksingh has quit IRC | 17:30 | |
| *** rohitksingh has joined #timvideos | 17:45 | |
| *** sb0_ has quit IRC | 18:12 | |
| CarlFK | back | 18:13 |
| CarlFK | _florent_: mithro http://img.flnet.org/a/shot0001.png | 19:04 |
| CarlFK | I see one pixel in the upper left quadrant | 19:04 |
| CarlFK | oh, I see a few more, all on the same X | 19:05 |
| _florent_ | what is it supposed to be? what's the configuration? | 19:05 |
| CarlFK | ah.. just noticed dvisampler1: IDELAY busy timeout | 19:06 |
| CarlFK | input0: 0x0 | 19:06 |
| CarlFK | input1: 0x0 | 19:06 |
| CarlFK | output0: 1280x720@60Hz from input0 | 19:06 |
| CarlFK | output1: 1280x720@60Hz from input0 | 19:06 |
| CarlFK | encoder: 1280x720 @ 30fps (1781Mbps) from input1 (q: 85) | 19:06 |
| CarlFK | ddr: read: 2205Mbps write: 0Mbps all: 2205Mbps | 19:06 |
| CarlFK | it should be a ubuntu desktop | 19:06 |
| CarlFK | _florent_: I am getting this about 1 per second: dvisampler1: IDELAY busy timeout | 19:10 |
| CarlFK | rebooted the source (negk), it is using the display device, now there are more dots on that X | 19:11 |
| CarlFK | http://img.flnet.org/a/shot0003.png | 19:12 |
| mithro | https://www.reddit.com/r/FPGA/comments/3kfx8m/remember_milkymist_and_netv_new_open_fpga_for/ | 19:12 |
| CarlFK | dvisampler1: IDELAY busy timeout | 19:12 |
| CarlFK | dvisampler1: IDELAY busy timeout | 19:12 |
| CarlFK | input0: 0x0 | 19:12 |
| CarlFK | input1: 0x0 | 19:12 |
| CarlFK | output0: 1280x720@60Hz from input0 | 19:12 |
| CarlFK | output1: 1280x720@60Hz from input0 | 19:12 |
| CarlFK | encoder: 1280x720 @ 30fps (4Mbps) from input1 (q: 85) | 19:12 |
| CarlFK | ddr: read: 2218Mbps write: 109Mbps all: 2327Mbps | 19:12 |
| tpb | Title: Remember Milkymist and NeTV? New open FPGA for HDMI and DisplayPort video signal announced : FPGA (at www.reddit.com) | 19:12 |
| mithro | Can everyone upvote -> https://www.reddit.com/r/FPGA/comments/3kfx8m/remember_milkymist_and_netv_new_open_fpga_for/ | 19:13 |
| tpb | Title: Remember Milkymist and NeTV? New open FPGA for HDMI and DisplayPort video signal announced : FPGA (at www.reddit.com) | 19:13 |
| _florent_ | CarlFK: can you reset/reload the FPGA? | 19:14 |
| CarlFK | http://paste.ubuntu.com/12330425/ issued reboot command, no help. hitting reset button now | 19:16 |
| tpb | Title: Ubuntu Pastebin (at paste.ubuntu.com) | 19:16 |
| mithro | https://www.reddit.com/r/videography/comments/3kfylb/timvideos_hdmi2usbtv_project_now_has_foss_hardware/ | 19:22 |
| CarlFK | _florent_: also mplayer is spewing [mjpeg @ 0x7f53ac04ab00]Found EOI before any SOF, ignoring | 19:27 |
| _florent_ | if input are not detected, that's the reason you are seeing a black screen | 19:29 |
| CarlFK | _florent_: reboot, reset button, un/plugged hdmi, no change. power cycle? | 19:32 |
| mithro | https://www.reddit.com/r/opensource/comments/3kg0rv/foss_hardware_for_recording_foss_conferences_and/ | 19:33 |
| _florent_ | yes you can try that | 19:34 |
| *** sb0 has joined #timvideos | 20:15 | |
| CarlFK | mithro: what did you want me to test? my laptop snap cleard out my short term memory | 20:33 |
| mithro | CarlFK: did the powercycle change anything? | 20:42 |
| CarlFK | mithro: _florent_ power cycle, reload everything, un/plug usb, all working again | 20:50 |
| _florent_ | ok good | 20:50 |
| *** sb0 has quit IRC | 21:04 | |
| mithro | _florent_: any idea what was going wrong there? Never seen busy timeout on IDELAY setting before? | 21:11 |
| CarlFK | may be related to screen saver ... let me enable that again | 21:12 |
| _florent_ | difficult to say from here | 21:15 |
| CarlFK | close lid, input1: 0x0 (ok) encoder stream goes black. (1/2 surprised, but maybe it sends black frames before the signal drops?) | 21:18 |
| CarlFK | hmm.. pulling the hdmi cable also creates black. I thought it would keep the last frame in the buffer? | 21:19 |
| _florent_ | no, now the cpu is clearing the buffer | 21:19 |
| CarlFK | _florent_: can you flip to the test pattern? | 21:19 |
| _florent_ | (that's what was in mixxeo) but we can keep the last frame if needed | 21:20 |
| CarlFK | last frame is good too. I'll let mithro decide | 21:20 |
| mithro | CarlFK: the encoder / usb stream is still running when disconnected? | 21:21 |
| CarlFK | mithro: yes. but black. I would prefer something that gives me a clue what is going on | 21:21 |
| CarlFK | last frame is good for recording. | 21:22 |
| CarlFK | test pattern is good for debugging | 21:22 |
| mithro | last frame with some type of overlay seems like a good idea? | 21:22 |
| CarlFK | fantastic | 21:22 |
| _florent_ | what do you want in the overlay? | 21:23 |
| CarlFK | I am 1/2 thinking test pattern is easy enough to remove in post. | 21:23 |
| CarlFK | _florent_: to give a clue there is a problem | 21:23 |
| _florent_ | btw if you want to experiment things on that, only the firmware need to be modified and it take a few seconds to test changes | 21:24 |
| CarlFK | _florent_: often other bits of our system have issues, and we need all the help we can get trying to figure out what bit needs to be kicked | 21:24 |
| CarlFK | oh crap.. I bet there isn't a battery on board to maintain a clock? | 21:26 |
| mithro | _florent_: I've been trying to convince CarlFK he should be able to hack on the firmware himself - he knows enough C to be dangerous :P | 21:51 |
| CarlFK | mithro: what did you want me to test? my laptop snap cleard out my short term memory. tumbleweed | 21:51 |
| CarlFK | tumbleweed packaged someting I think | 21:52 |
| mithro | CarlFK: oh - the Atlys UART driver package | 21:52 |
| mithro | CarlFK: we changed it to use dkms rather than module-assistant - which means it shouldn't need to be manually built after install any more | 21:52 |
| CarlFK | mithro: whats the status of vga -> Atlys as an input ? I have one of the boards from LCA | 22:48 |
| mithro | CarlFK: its not supported on the new firmware yet | 22:48 |
| mithro | rohitksingh: ping? | 22:48 |
| CarlFK | mithro: do you have any hadware for VGA -HDMI? | 22:52 |
| mithro | CarlFK: not really | 22:52 |
| CarlFK | cuz that wold be hepful for pyconza in 20 days | 22:52 |
| CarlFK | mithro: I just realized the dvsource-other thing will let me use a web cam as a source. so a static pointed at the audience - any recommendation? | 22:55 |
| mithro | CarlFK: ? | 22:55 |
| CarlFK | mithro: instead of using cheep DV cams, I just realized I can use usb web cams | 22:56 |
| CarlFK | mithro: wondering if you have any recommendations | 22:56 |
| mithro | The logic tech HD UVC cameras are the ones I've used | 22:57 |
| *** rohitksingh has quit IRC | 23:30 | |
| *** rohitksingh has joined #timvideos | 23:46 | |
| *** sb0 has joined #timvideos | 23:49 | |
| *** sb0 has quit IRC | 23:56 | |
| *** sb0 has joined #timvideos | 23:57 | |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!
