Current Server Time: 05:53 (Central Europe)

#apertus IRC Channel Logs

2015/02/18

Timezone: UTC


23:53
mithro
well, if people know of anyone who might be interested, please get them to email me at mithro@mithis.com
23:57
intracube
left the channel
00:16
fsteinel_
joined the channel
00:19
fsteinel
left the channel
00:27
g3gg0
left the channel
00:30
fsteinel_
changed nick to: fsteinel
02:51
Bertl_zZ
changed nick to: Bertl
02:51
Bertl
back now ..
02:51
Bertl
was a longer nap this time :0
02:53
Bertl
mithro: sound great! do you have anybody in mind to do the implementation?
02:56
mithro
Bertl: yes, _florent_ from enjoy-digital -> https://github.com/enjoy-digital/ -- he came up with the $25k estimate and has done a bunch of great FOSS cores already (by getting companies to agree to open then up when he does it for them)
02:58
Bertl
nice! please keep me updated on the progress in this regard, as apertus would be definitely interested
03:00
Bertl
we might even be able to pitch in either with development or money, if there is a problem getting the five companies
03:00
mithro
Bertl: I have a specification I originally wrote to try and guide a PhD student to writing such a core
03:00
mithro
https://docs.google.com/a/mithis.com/document/d/1xSo71wu_t8SUuUZehakvqTL_png5YN-bOwD47PRXGIM/edit
03:01
mithro
Bertl: If you have any contacts at companies / people who might be interested, I'd love introductions
03:03
Bertl
yes, of course, I will drop you a note if I find somebody interested
03:04
mithro
how is apertus going?
03:05
Bertl
fine as far as I can tell, we are making great progress with the Beta design and testing
03:05
Bertl
both on the electronics as well as case and ergonomics side
03:06
mithro
my sister is an budding film maker here in Australia, I'm sure if you guys are successful it would be super useful for her
03:07
mithro
my project is also kind of interested in your stuff from a conference recording perspective - having everything be foss - including the camera gear is kinda of our crazy dream :)
03:07
Bertl
yeah, would be cool I guess :)
03:08
mithro
is anyone working on a PTZ style mount at all?
03:08
Bertl
you have to check with se6astian on that
03:12
troy_s
What is a PTZ?
03:12
troy_s
(Stoopid question I am sure.)
03:13
mithro
Pan-Tilt-Zoom
03:14
troy_s
Ok. Even stoopider question, what the hell is a mount then?
03:14
troy_s
Or what is a PTZ-style.
03:15
Bertl
I'd say a thingy where you can screw the camera on and it will be panned and tilted by some kind of servo system
03:15
mithro
Bertl is correct
03:16
Bertl
the zoom part sounds strange though, as it is more likely a feature of the lens system
03:16
Bertl
http://www.mcmelectronics.com/product/DEFENDER-SECURITY-82-12435-/82-12435
03:16
Bertl
(just as an example)
03:16
mithro
Bertl: yeah - that is often a sticking point with doing PTZ mounts rather than a full PTZ camera
03:17
mithro
https://www.kickstarter.com/projects/1608311032/lenzhound-wireless-lens-motor-control-system/posts <- something like that can work
03:18
Bertl
ah, yes
04:38
troy_s
Remote head.
04:38
troy_s
Hence why I asked. Generally the term is a remote head.
04:39
troy_s
Sorry, looks like a FIZ unit in the link.
04:40
troy_s
Sounded like a remote head, but is actually a FIZ.
04:41
troy_s
Focus, Iris, Zoom. Pan, Tilt, Zoom confused the hell out of me.
04:41
troy_s
Seems like a confusion of gear. :)
06:12
Bertl
yeah, interestingly it seems to be quite common
06:12
Bertl
but note that the "PTZ" gear linked above is Pan, Tilt (no zoom though)
06:13
Bertl
I also think that the confusion comes from the dome cameras, which usually combine pan, tilt and zoom into one "remote controlled" unit
06:14
Bertl
(typical surveillance cameras)
06:14
kgugala
joined the channel
06:14
Bertl
morning kgugala!
06:15
kgugala
morning
07:02
g3gg0
joined the channel
07:07
Bertl
morning g3gg0!
07:07
g3gg0
left the channel
07:48
Juicyfruit
joined the channel
08:27
se6astian|away
changed nick to: se6astian
08:29
se6astian
good morning
08:29
Bertl
morning se6astian!
08:36
se6astian
I created email forwarders for nikos in /etc/postpfix/alias and added them to the whitelist in /etc/postfix/local_recipient and did run make afterwards, stil nikos reports its not working, did I forget anything?
08:36
se6astian
does postfix need a reload/restart?
09:39
Bertl
should be fine :)
09:39
Bertl
off to bed now ... have a good one everyone!
09:39
Bertl
changed nick to: Bertl_zZ
09:55
Andrej74
joined the channel
10:07
Andrej74
left the channel
10:18
Andrej74
joined the channel
11:56
intracube
joined the channel
14:29
Yannick
joined the channel
14:37
se6astian
hey thre
14:38
yannick__
joined the channel
14:38
yannick__
left the channel
14:41
se6astian
Yannick: ?
14:42
Yannick
left the channel
14:43
yannick
joined the channel
14:43
yannick
Hey Sebastian, I'm up now
14:44
yannick
Thanks for getting back to me
14:45
se6astian
great
14:45
se6astian
my pleasure
14:45
se6astian
so unfortunately bertl is still sleeping
14:45
se6astian
and I will leave the office in 15 minutes
14:45
se6astian
but that still leaves us a bit of time to chat :)
14:46
yannick
sound good
14:46
se6astian
for the VR rig you emailed
14:47
yannick
As I said in email, my end goal is to use the Beta camera to build multi-camera rigs. First stereo then VR
14:47
se6astian
do I understand it right that you want to create a left eye and a right eye spherical panorama, right?
14:47
yannick
I came accross the VR open surce group you mentionned but their focus is really video games, not live action
14:48
yannick
Yes, the final rig will be 6 cams, 3 stereo pairs
14:48
yannick
this is the direction the business is going right now
14:49
yannick
most people build them out of 6 reds but they are too big to be optimal
14:51
yannick
the problem with panorama video rigs is parallax. To avoid it, we would need all 3 lenses to be in the same physical space which is not really possible but by reducing the diameter of the rig to a minimum, it minimizes the problem
14:51
se6astian
I see
14:52
se6astian
Beta electronics designs are currently more wide than tall but you can easily just rotate the entire camera hardware by 90°
14:52
se6astian
the sensor is 4:3 so you dont loose a lot of vertical resolution
14:52
yannick
this is why this project is perfect for me since I can really optimimize the design. I'm taken apart tens of thousands of dollars worth of cameras in the past but that doesn't make something solid
14:53
yannick
I was thinking about just redesigning the chip board, flip it 90 degrees
14:56
se6astian
that will be rathe tricky
14:56
se6astian
*rather
14:56
se6astian
as the image sensor pins are top/bottom of the sensor
14:57
se6astian
and that are 64x 600Mhz LVDS pairs that need to be length matches
14:57
se6astian
*matched
14:57
se6astian
I am not saying its impossible, but you need to know what you are doing when you redesign this board
15:00
se6astian
ok gotta go
15:00
se6astian
I am happy to chat more later
15:00
se6astian
or maybe herbert awakes in the meantime :)
15:00
se6astian
feel free to stick around
15:01
yannick
I'm not expecting it to be trivial but I'm sure I can find someone to make this happen
15:01
yannick
Thanks, great talking to you
15:04
se6astian
definitely, we added a board between the sensor PCB and main Beta board that actually does exactly this 90° rotation twice so it is definitely possible
15:04
se6astian
see you
15:04
se6astian
changed nick to: se6astian|away
15:21
wescotte
joined the channel
15:21
wescotte
left the channel
15:58
yannick
left the channel
16:06
yannick
joined the channel
16:12
Andrej74
left the channel
16:15
Andrej74
joined the channel
16:31
g3gg0
joined the channel
16:41
se6astian|away
changed nick to: se6astian
17:31
Bertl_zZ
changed nick to: Bertl
17:31
Bertl
morning folks!
18:39
se6astian
good morning
18:39
se6astian
new PCBs just shipped from OSHpark to you
18:40
Bertl
great! we should be able to submit the next pack tonight or tomorrow morning :)
18:41
se6astian
great :)
18:41
se6astian
trouble with the layout?
18:43
Bertl
not really, but I had to rearrange a few things to make it fit
18:44
se6astian
sounds good
18:45
Bertl
yeah, it is rather packed for a test PCB now, but OTOH, I managed to add a few options which were not planned
18:47
se6astian
please share, I love good surprises :)
18:47
Bertl
I'll upload in a few minutes (the current state)
18:48
se6astian
btw I just ordered SD cards for another project and saw that there is a new really fast generation of SDHC cards that claim write speeds of 250MB/s - now I have no idea of the SD card interface but I guess it would be simpler to interface than say SATA...
18:49
danieel
se6astian: which model?
18:51
se6astian
eg. SanDisk Extreme PRO SDHC 16GB 280MB/s, UHS-II U3
18:51
danieel
UHS was not part of the free specs some time ago.. not sure if that changed by having class3 devices now
18:52
Bertl
UHS yes, but UHS-II?
18:52
danieel
pin assignment is there
18:53
danieel
even from uhs-I it misses some diagrams
18:53
Bertl
UHS-II requires 0.4-volt operation, but otherwise it should be similar
18:53
Bertl
not sure about the protocol though
18:53
danieel
says FD156 and HD312 (numbers in MB/s)
18:54
Bertl
yep, HD would be fine though
18:54
danieel
at 8/10 it is 1.56 Gbps
18:54
danieel
both are probably same freq, just the arrangement of lanes differ (1:1 vs 2:0)
18:55
Bertl
yeah, sounds plausible
18:55
danieel
the transfer rate spans from 0.39 to 1.56 G... if that is self-clocked, then it is a nice feature
18:56
danieel
you can achieve like 1.25G at ioserdes at 1:10?
18:57
Bertl
hmm, UHS uses 4bit mode, no?
18:57
danieel
uhs1, uhs2 has these 2 extra pairs
18:57
danieel
uhs is rated up to 104MB/s (208MHz x 4b)
18:58
Bertl
I'm just wondering how you arrive at the 1.56Gbps (per lane)
18:58
danieel
156MB x 10 (as for 10/8 coding)
18:58
danieel
8b10b
18:59
danieel
the gbps are written in the specs
18:59
Bertl
but not on a single pair though
18:59
danieel
on a single pair
19:00
danieel
ok, so there is a dedicated diff. clock as well
19:01
Bertl
and yes, 1.25G even 1.6G is not a problem with the ioserdes
19:02
danieel
what about bidirectoinal - on the fly change?
19:02
Bertl
although it is not completely within the spec over the full range
19:02
Bertl
possible but a little trickier
19:02
danieel
1.6 is usually only at higher speed grades / lower voltages no?
19:03
Bertl
the main problem is the output interface (capacitance, switching)
19:04
Bertl
ah, yes, RCLK is sent through DAT0/1
19:05
Bertl
se6astian: so yes, might be worth investigating
19:07
se6astian
great :)
19:07
danieel
the timing chart is missing
19:07
se6astian
I ll create a lab task as reminder
19:07
danieel
and about the voltage, only the 0.4v is mentioned, probably the vpp swing
19:08
lab-bot
sebastian created T292: Investigate SDHC UHS-II interface. http://lab.apertus.org/T292
19:12
danieel
the card seems to do 180MB/s write by the reviews
19:12
danieel
http://www.thessdreview.com/featured/sandisk-extreme-pro-uhs-ii-memory-card-review-16-gb/2/
19:13
Bertl
so interleaving 2-4 cards would work fine
20:46
lab-bot
sebastian created T293: Create Beta development status overview page. http://lab.apertus.org/T293
21:47
se6astian
time for bed!
21:47
se6astian
changed nick to: se6astian|away
21:47
Bertl
nn
21:52
Andrej74
left the channel