Current Server Time: 12:48 (Central Europe)

#apertus IRC Channel Logs

2013/11/22

Timezone: UTC


23:15
mikkael
left the channel
23:21
jucar
joined the channel
23:27
jucar
left the channel
23:34
jucar
joined the channel
23:41
jucar
left the channel
23:41
jucar
joined the channel
23:46
jucar
left the channel
00:19
jucar
joined the channel
00:26
jucar
left the channel
00:51
jucar
joined the channel
01:15
jucar
left the channel
01:44
jucar
joined the channel
02:36
jucar1
joined the channel
02:37
jucar
left the channel
05:20
Bertl
off to bed now ... have a good one everyone!
05:31
Sasha_C
left the channel
06:08
leo_m
left the channel
06:26
Sasha_C
joined the channel
07:00
philippe
joined the channel
07:00
philippe
changed nick to: Guest77590
07:01
Guest77590
changed nick to: philippej
08:00
philippej
left the channel
08:54
Sasha_C
left the channel
09:05
Sasha_C
joined the channel
09:06
Sasha_C
left the channel
11:07
rexbron
left the channel
11:08
rexbron
joined the channel
12:12
intracube
joined the channel
12:23
S3bastian
joined the channel
12:23
S3bastian
good evening
12:28
S3bastian
afternoon?
12:37
mars_
left the channel
12:37
mars_
joined the channel
12:39
FergusL
hi here
12:40
FergusL
afternoon I'd say
12:49
leo_m
joined the channel
12:52
leo_m
hi everyone
12:57
FergusL
hi leo_m
13:10
S3bastian
hello!
13:46
Bertl
morning everyone!
13:56
S3bastian
hellO!
15:17
S3bastian
left the channel
15:39
leo_m
left the channel
17:09
FergusL
Hey, is tiff a float format ? Cc troy_s
17:09
FergusL
Cc rexbron
17:11
Bertl
The inclusion of the SampleFormat tag in TIFF 6.0 allows TIFF files to handle advanced pixel data types, including integer images with more than 8 bits per channel and floating point images.
17:16
philippe
joined the channel
17:16
philippe
changed nick to: Guest22341
17:18
Guest22341
changed nick to: philippej
17:31
philippej
Hi everyone !
17:31
Bertl
hey philippej!
17:32
FergusL
thanks Bertl, I knew you were the kind of guy that print and pins all the relevant RFCs just above the computer screen on the wall
17:32
philippej
Bertl, I got a reply from avnet, basically, they will at some point make a new board, but curently they are not yet in the design phase...
17:32
Bertl
FergusL: virtually, yes
17:33
Bertl
philippej: any ideas what the usual cycle for that would be?
17:33
Bertl
i.e. one month design, one month testing? or one year design, no testing?
17:33
philippej
"The only board we have today that would meet your requirements is the ZynQ MMP module AES-MMP-7Z045-G but itâs cost could be prohibitive. We have talked about building a 7015 board but donât have it planned yet so it would be several months before it would be available for sale."
17:34
philippej
"Does the CMV12000 allow you to double the data rate to 600Mbps? This would allow you to use a 32 bit interface to the MicroZed and leave enough pairs available to add an external high speed serializer for the SDI interface. This is probably the best solution we have at this time.
17:34
philippej
I will keep your requirements in mind as we develop the spec for our 7015 boards. "
17:34
Bertl
also the 7045 is a no-go for us, because the WebPack does not cover it
17:34
philippej
I thought already abot it :-(
17:34
philippej
this board is +- $1200
17:35
philippej
is it an option to add an external serdes to the microzed and use half io of the sensor for now? does it make sense technically and economically ?
17:36
Bertl
won't work, we wouldn't even have enough I/O ports port this
17:36
Bertl
i.e. we could use even less lanes from the sensor, but that kind of fails the point
17:37
Bertl
besides, I haven't had a chance to check out the microzed, but that will probably change soon
17:37
philippej
do yo understand like me they are planing a new board in months (not years)?
17:38
Bertl
well, my interpretation is, that it will take 'several' months, which is the usual vague hand-waving for 'some unknown time'
17:39
Bertl
all it means is that it won't be available in the next few months
17:39
philippej
yep
17:41
philippej
I really wanted to find a solution that allows to develop now the software side and provide a lower end (cheaper, less risky) model, while allowing future upgrades ... :-/
17:42
Bertl
as I said, I'm all for it, but I doubt it exists atm
17:42
philippej
yeps
17:43
FergusL
how does one deal with the glossiness of the IT8 paper ?
17:46
Bertl
good question, one suggestion I've seen several times is to use really diffuse lighting and keep objects away from the paper
17:46
FergusL
yes, that's the idea I guess
17:47
philippej
and position the light such as the paper doesn't reflect :-)
17:48
Bertl
well, with diffuse lighting there is no 'light' per se
17:48
FergusL
I wonder if right now it could be a slight bummer for the profiling
17:50
philippej
you just put it enough out of the axis of the camera
17:51
Bertl
which would cause a problem with the uniformity
17:52
Bertl
so you need at least two light sources to get somewhat reasonable results
17:53
Bertl
it helps that the chart is relatively small, so with two larger light sources you have a good chance to illuminate it evenly
17:55
FergusL
I've seen a chart shooting setup once, it had two diffuse light on the sides
17:55
FergusL
camera right in front, more than 2m away for the chart
17:57
Bertl
yep, sounds like a reasonable setup
17:57
Bertl
light sources probably at 45° to the chart normal
17:58
FergusL
yes
17:59
FergusL
from the top of your head, do you remember if the IT8 shot is flipped ?
17:59
Bertl
yes, it is flipped, so that it 'looks' normal
18:00
Bertl
(i.e. the letters are readable)
18:01
FergusL
I see
18:01
FergusL
well it's still important for the debayering
18:01
Bertl
register dump is at the end of the raw data
18:02
FergusL
I know ! it follows the addresses in the cmv doc I guess ?
18:03
Bertl
yes, precisely, 16bit each register from address 0 to 127
18:03
FergusL
ok
18:18
philippej
left the channel
19:44
gcolburn
joined the channel
20:31
dmj_nova
left the channel
21:36
dmj_nova
joined the channel
22:27
gcolburn
left the channel
22:34
gcolburn
joined the channel
22:45
troy_s
FergusL: There may be a float version, but mostly int as I know.