Current Server Time: 09:58 (Central Europe)

#apertus IRC Channel Logs

2017/01/19

Timezone: UTC


23:56
Spirit532
left the channel
01:37
alexML_
joined the channel
01:37
rexbron_
joined the channel
01:41
rexbron
left the channel
01:41
TD-Linux
left the channel
01:41
alexML
left the channel
02:03
TD-Linux
joined the channel
07:10
Bertl_zZ
changed nick to: Bertl
07:10
Bertl
morning folks!
07:37
pusle
left the channel
08:41
ItsMeLenny
left the channel
08:41
ItsMeLenny
joined the channel
08:51
dimaursu16
joined the channel
09:00
Spirit532
joined the channel
09:09
aombk3
left the channel
09:09
aombk3
joined the channel
09:22
dimaursu16
left the channel
09:40
aombk2
joined the channel
09:42
aombk
joined the channel
09:43
aombk3
left the channel
09:45
aombk2
left the channel
09:45
dimaursu16
joined the channel
10:15
jucar
joined the channel
10:43
se6astian|away
changed nick to: se6astian
10:54
jucar
left the channel
10:56
BogdanXor
joined the channel
10:57
BogdanXor
on the AWS Xilinx FPGA dev instance, I got access to the program and to their documentation + infrastructure and the documentation states clearly, as expected, that the Vivado installation is device-locked to the Virtex Ultrascale+ device they are using
10:58
BogdanXor
btw, have you guys considered using Slack for collaboration? :D
10:59
Bertl
we haven't considered any proprietary vendor-lock-in solution yet
11:00
Bertl
@AWS yeah, that makes sense
11:19
BogdanXor
left the channel
11:23
dimaursu16
left the channel
11:23
dimaursu16
joined the channel
11:23
dimaursu16
left the channel
11:23
dimaursu16
joined the channel
11:41
dimaursu16
left the channel
11:53
Spirit532_
joined the channel
11:57
Spirit532
left the channel
12:20
dimaursu16
joined the channel
12:20
dimaursu16
left the channel
12:20
dimaursu16
joined the channel
12:32
danieel
Spirit532_: to me it seems like a sensor defect on the 2 columns - does the camera appear as an UVC device, or you use any ximea drivers? They might be processing the image - anyway the DNGs are RGB not bayer thus that is definitely not raw data
12:33
Spirit532_
changed nick to: Spirit532
12:33
Spirit532
ximea drivers
12:33
danieel
it might be just one broken sensor column and it show as two after debayer in their hw/sw
12:34
Spirit532
I don't have control over the camera's functions atm
12:34
Spirit532
onky esposure and gain
12:34
Spirit532
only*
12:34
danieel
what else would you want?
12:34
Spirit532
you can turn on sensor correction & stuff
12:34
Spirit532
and other functions
12:35
danieel
those are model dependant and not everything is everywhere if you refer to their xi api
12:35
Spirit532
yes, but it's on mine.
12:36
danieel
Bertl: what hw corrections have the CMV* parts? I thought there are none.. as we were happy seeing the KAC did had some
12:36
Spirit532
none, basically.
12:36
danieel
so just sw postprocess from the reference row/cols
12:37
Bertl
danieel: haven't seen any hardware corrections in the CMV* parts yet
12:37
Spirit532
it's in the fpga
12:37
danieel
that would be black level correction only probably :) the fpga is fairly small
12:38
Spirit532
not sure what it does, but it does something.
12:38
danieel
Bertl: i know they are doing a (temperature,freq) based corrections - have you done any frequency sweep to see noise response?
12:40
Bertl
you mean sweep the LVDS clock? no haven't tried that yet
12:40
danieel
yes, that one
12:40
Bertl
with temperature, we haven't seen much change in our tests
12:41
Spirit532
I'm not sure what is wrong with the camera running over usb2
12:41
Spirit532
but when I decrease exposure, it goes all funny and noisy
12:41
Bertl
well, USB2 has probably too little bandwidth
12:41
danieel
the freq as a factor is there probably because line time and termal generation - slower you scan, higher the difference from top to bottom
12:41
Spirit532
but that's what happens when you run something on something you're not supposed to run it on
12:41
Spirit532
lol
12:42
Bertl
danieel: are you sure the sensor does any correction for that?
12:42
danieel
not the sensor, the sw or fpga
12:42
Bertl
ah, okay, yes that is correct, the FPGA is supposed to do any corrections
12:43
danieel
we were discussing something with them + got their head here.. and asked what they really do
12:43
Bertl
got it! thanks for clarifying
12:43
danieel
anyway it unusable for the situation
12:57
arpu
joined the channel
14:06
ItsMeLenny
left the channel
14:58
arpu
left the channel
14:58
davidak
joined the channel
15:01
se6astian
changed nick to: se6astian|away
15:10
arpu
joined the channel
15:58
Spirit532
I am very impressed with the cmv2k sensor so far
15:58
Spirit532
not perfect in extreme low light
15:58
Spirit532
but I wouldn't expect a 2/3" sensor to be, haha
16:00
Spirit532
but the camera tool is a bit wonky in exposure settings.
16:47
se6astian|away
changed nick to: se6astian
17:11
slikdigit
joined the channel
17:34
davidak
left the channel
18:12
davidak
joined the channel
18:14
slikdigit
left the channel
20:48
se6astian
changed nick to: se6astian|away
21:18
dimaursu16
left the channel
22:11
slikdigit
joined the channel
22:49
dimaursu16
joined the channel