Current Server Time: 19:10 (Central Europe)

#apertus IRC Channel Logs

2013/11/29

Timezone: UTC


01:25
Bertl
back now ...
02:35
FergusL
wb Bertl !
03:51
aombk
joined the channel
05:09
aombk
left the channel
07:51
Sasha_C
joined the channel
07:54
se6astian
joined the channel
07:54
se6astian
good morning
07:54
Sasha_C
Good morning Se6astian
07:54
Sasha_C
Did you receive my email response?
07:55
Sasha_C
(re: the IRC archive)
08:00
se6astian
just arrived in my office, yes
08:01
Sasha_C
Sorry about the mix-up
08:04
se6astian
our archives date back to November 2013
08:04
se6astian
now is november
08:04
Sasha_C
I know, but I'm thinking about what this will look like when people come to read it a month or so from now
08:05
se6astian
our archive starts in June 2013
08:05
Sasha_C
Well, when exactly in November do the logs begin? Early, mid, late?
08:05
se6astian
http://irc.apertus.org/
08:05
Sasha_C
Really?
08:05
Sasha_C
Ok, I'll change it then
08:05
Sasha_C
I only assumed that IRC was only being logged after Herbert activated aPinky?
08:06
Sasha_C
And I see you've updated the article...
08:06
se6astian
look at http://irc.apertus.org/
08:07
Bertl
news about the irclog? :)
08:07
se6astian
yes, the article is about the be released: https://www.apertus.org/irc-archives-available-article
08:07
Sasha_C
Yep
08:07
se6astian
*to be
08:08
Bertl
November 2014 is even better than November 2013 :)
08:09
Sasha_C
Yeah, that was my mistake, editing text at 1am :(
08:09
Bertl
to err is human, no problem ...
08:17
se6astian
Sasha_C, did you take a look at my changes now?
08:17
Sasha_C
Yep, looks good
08:23
se6astian
great, thanks
08:23
se6astian
then I will release now
08:25
se6astian
is it strange to say "IRC Chat"?
08:25
Bertl
yes
08:25
se6astian
as that actually means Chat Chat :)
08:25
Sasha_C
agreed
08:25
se6astian
but we cant say IR Chat
08:26
se6astian
that nobody will recognize
08:26
Bertl
let's say Internet Relay Chat (IRC)
08:26
Bertl
this way folks also learn the acronym
08:27
Sasha_C
great. I was going to suggest 'IRC room/channel', but Bertl's suggestion steals the prize
08:27
se6astian
fixed
08:28
se6astian
posted
08:28
se6astian
btw our last prototype update is almost a month old
08:28
se6astian
is there anything new we can report?
08:29
se6astian
pictures of the second unit being assembled?
08:29
se6astian
the black lens mount looks very stylish for example IMO
08:30
Sasha_C
Yes, people will definitely want to see that
08:34
Bertl
I did take a few pictures of the new unit, but it looks almost like the old one, so nothing really new there
08:34
Bertl
but the black lens mount with the silver bayonet ring looks sexy :)
08:38
se6astian
what else can we report?
08:38
Bertl
a lot of software news
08:38
Bertl
hdmi out from memory for example
08:39
Bertl
full throughput reader/writer test (now working :)
08:39
Bertl
the first IT8 chart?
08:40
Bertl
http://vserver.13thfloor.at/Stuff/AXIOM/RAW/IT8_incand.fix.png
08:42
Bertl
http://vserver.13thfloor.at/Stuff/AXIOM/almost2.png (this is from the hdmi out tests)
08:42
Bertl
http://vserver.13thfloor.at/Stuff/AXIOM/almost.png (one bug earlier :)
08:44
Bertl
http://vserver.13thfloor.at/Stuff/AXIOM/buggyglass.png (two bugs earlier :)
08:45
Sasha_C
This is great stuff. We should combine a couple of these pictures with an image or two of the camera and black lens mount (suggested by Se6astian above)
08:46
Bertl
yeah, maybe make it on the hardware side/ on the software side
08:47
Sasha_C
is "almost2.png" an example of the latest progress you've made with debugging the software?
08:47
Bertl
would have been material for two news entries, but I guess it is kind of moot to have two news entries at the same day (all happened over the last week or so)
08:47
Bertl
actually we fixed it completely, so it's working now
08:48
Bertl
but that was the last iteration before the final issue was fixed
08:48
Sasha_C
Fantastic! And raw capture and debayering progress?
08:49
Bertl
debayering is not an issue atm, we do that either in post processing or not at all (for the HDMI out)
08:49
Bertl
raw capture is working fine for some time now, what we have to do is combine both, clean up the code and optimize the design
08:50
Sasha_C
Well, let's report this then! People will find it great to read / hear
08:51
Bertl
take your time to prepare something, I'll be off to bed shortly
08:51
Sasha_C
Will do, and have a good one :)
08:51
Bertl
but I can read/check anything in the evening
08:52
Bertl
(i.e. when I get up again :)
08:52
yahel
joined the channel
08:52
Bertl
welcome yahel!
08:55
yahel
Hi Bertl
08:58
yahel
left the channel
08:58
Bertl
off to bed now ... have a good one everyone!
08:58
Sasha_C
Take care man
09:09
se6astian
night Bertl
09:10
se6astian
sasha do you have time to draft an article together?
09:10
se6astian
I will be away over the weekend
09:10
Sasha_C
I'll get started on it tonight and present a draft early next week
09:10
se6astian
great, thanks!
11:59
ThatCantBe
joined the channel
11:59
Sasha_C
left the channel
12:10
notanerd
joined the channel
12:44
notanerd
left the channel
13:39
[1]se6astian
joined the channel
13:41
se6astian
left the channel
13:42
se6astian
joined the channel
13:42
FergusL
hello here
13:42
jucar1
joined the channel
13:44
jucar
left the channel
13:44
[1]se6astian
left the channel
15:05
[1]se6astian
joined the channel
15:05
se6astian
left the channel
15:05
[1]se6astian
changed nick to: se6astian
15:46
Bertl
morning everyone!
15:47
FergusL
'ning Bertl
15:48
se6astian
morning :)
15:52
se6astian
before I turn off the lights here at the office
15:52
se6astian
quick and dirty rendering of what we discussed yesterday
15:52
se6astian
https://cloud.gerade.org/public.php?service=files&t=88644302ffde4f6e5659627a3c8bbf32&path=/Axiom/open-modules-concept-inside01.jpg
15:53
FergusL
that was also discussed on the mailing list, right ?
15:53
Bertl
se6astian: yeah, but I'd go one step further and drop the module interconnects as well
15:54
se6astian
FergusL, not yet
15:54
FergusL
just because of the hassle it represents ?
15:54
FergusL
Bertl ^
15:54
se6astian
Bertl, how do you mean?
15:54
FergusL
and just go with external cables ? like power straight from battery/power block to frontend ?
15:55
Bertl
what if we can simply replace the backplane when changing the number of modules?
15:55
Bertl
(or extend it to some degree)
15:56
Bertl
judging from the feedback here on the channel, the number of used modules will not change very often
15:56
se6astian
that would mean every time you add/remove a module you would have to completely disassemble the entire camera, even the PCB with the sensor
15:56
Bertl
maybe, but not necessarily
15:57
Bertl
the backplane could use 90° sockets and a similar approach to extend between modules
15:57
FergusL
I agree with Bertl that setup wouldn't change that often
15:58
FergusL
except maybe for hypothetical sound and remote control modules which is the most subject to changes depending on the type of project the camera is used on
15:59
Bertl
so, if the change can be done 'easily' I don't think it's a big deal for anybody to disassemble the whole thing (when you change the first 3-4 modules)
15:59
Bertl
for addon/low bandwidth modules we might simply switch to header connectors
16:02
Bertl
have to grab some groceries, bbl
16:08
se6astian
lets continue this discussion soon
16:08
se6astian
gotta go
16:08
se6astian
left the channel
16:33
jucar1
left the channel
16:47
jucar
joined the channel
17:58
Bertl
back now ...
18:02
VBenjFR
joined the channel
18:03
Bertl
wb VBenjFR!
18:05
VBenjFR
Good evening!
18:25
se6astian
joined the channel
18:30
se6astian
back
18:30
Bertl
wb
18:31
se6astian
I want to evaluate what the team thinks about the requirements of "how easy" the modules should be to assemble/disassemble
18:31
se6astian
http://strawpoll.me/ is a nice way to gather simple opinions
18:32
se6astian
I am just struggling to formulate the questions in a way that contains enough information but does not give any bias towards a particular answer
18:33
Bertl
how about a questionaire? with questions like: how often do you plan to remove a module?
18:33
Bertl
how often do you expect to add a new module?
18:34
Bertl
how long can it take to assemble 5 modules?
18:34
Bertl
and have some choices there like
18:35
se6astian
for me those questions are not clear, because in a system where chaning modules just takes a few seconds I would do it everyday before I start shooting, but if its more complicated I would leave those I use most of the time installed
18:35
Bertl
every hour, every day, twice a week, every fortnight, asoasf
18:36
Bertl
fair enough, question: why would you do it every day?
18:36
Bertl
i.e. what would you change on a daily basis?
18:37
se6astian
the project requirements, for example I know I will shoot with an external audio guy one day and therefore reqire the meta module to give out a genlock signal
18:38
se6astian
on the next day I will record audio on the camera but dont need the meta module
18:38
se6astian
or was it the "time" module with genlock...
18:38
Bertl
but the question is, why would you want to remove it in this case?
18:39
Bertl
for sure, I do not need all the apps on my smart phone, doesn't mean I will uninstall them evey day, despite the fact that it only takes a few seconds :)
18:43
Bertl
IMHO it will be similar to the normal way folks go about camera setups, i.e. have a number of components, assemble them, balance the entire thing and that's it, if requirements change, like for example the output or storage type, then the camera will be modified/adapted, but otherwise it will remain untouched
18:44
Bertl
the only thing which will be permanently changed is the battery :)
18:44
Bertl
(and if we find a smart way to avoid that, like remote inductive charging, even that will be avoided I guess ...
18:44
Bertl
)
18:45
se6astian
some motivation might be weight and size: when you shoot on a crane, sidemounted to a car, helicopter, underwater you will want to have a camera as small and lightweight as possible
18:45
se6astian
but those shots might be just in the middle of a normal shooting day
18:46
Bertl
yes, but this will require special handling anyway
18:46
Bertl
so it won't matter if it takes 30 seconds or 5 minutes
18:46
se6astian
currently yes, because there is no camera that could do it all without being modified
18:47
Bertl
for example, we could do the double side backplane concept, but stop with one backplane after the main processing module
18:47
se6astian
agreed, but a major factor is if it can be done without a workbench, esd protection, certain screwdrivers and other equipment without risking damage to the camera due to moisture, dust
18:48
Bertl
that would allow the modules to be added/removed from the side, assuming that the module body stays the same
18:49
Bertl
so for example, the 'main' body could already provide space for 3-4 'modules'
18:49
Bertl
where you can, like with PCI cards, plug in your favorite module
18:49
Bertl
or remove it, if it needs to be changed or adapted, or whatever
18:50
Bertl
very similar to PCs, you will not change it very often, once you've found your favorite combination
18:50
se6astian
interesting ide
18:50
se6astian
a
18:50
Bertl
still provides an upgrade path, an open interface and all the flexibility you get if you really need to
18:51
se6astian
most modules that you want to swap will be low bandwidth anyway (audio, tiemcode, etc.)
18:51
Bertl
precisely
18:51
Bertl
also allows for the I/O part to be modular
18:52
Bertl
i.e. easily switch between 4x SDI out or SATA out or whatever
18:52
se6astian
but we would need at least 1-2 modules with double backplane for storage/raw/whatever else
18:52
Bertl
we might/should get away with single side for the high speed serial tranceivers
18:52
se6astian
or would the one sided interfaces be enough if we use high speed serial interfaces for sata anyway
18:53
Bertl
that's the idea, might require higher quality slots though
18:53
se6astian
with gold plating etc.?
18:54
Bertl
more importantly with proper impedance and shielding
18:54
Bertl
but e.g. PCIe slots should be more than fine
18:54
Bertl
(or the laptop equivalents if size is an issue)
18:55
se6astian
are pcie slots better in terms of impedance and shielding than SODIMM slots?
18:56
Bertl
probably because they are designed for higher frequencies
18:56
Bertl
I'd suggest the longer backplane to do near and far connects
18:56
se6astian
they also have lower pin counts
18:57
Bertl
like this:
18:57
Bertl
---------, ,-----, ,------
18:57
Bertl
---------+-------+--------
18:57
se6astian
then I would suggest to not do the backplane on the sides as that would block the space to add connectors but rather at top/bottom (as in my concept image)
18:58
Bertl
fine as well, but requires to open the top (if it is on the bottom) and even more problematic to open the bottom if it is on the top
18:59
Bertl
OTOH, I don't see a problem with connectors if one of the backplanes is shorter, i.e. they will be on the same side as the short backplane
18:59
se6astian
you would need custom enclosures per module anyway as each module has different connectors (right side) and button(left side)
19:00
se6astian
true, but inside the head where we still have two backplanes there could not be any connectors to the sides
19:00
Bertl
and there won't be any connectors on the baseboard/sensor frontend
19:00
Bertl
no need for the second backplane there I guess
19:01
Bertl
we already have high speed serial out from the baseboard
19:01
se6astian
unless there is a dedicated 3rd pcb for the head that has the connectors and just uses a single backplane side
19:01
Bertl
so any I/O or storage module will use that
19:01
Bertl
that's the idea
19:01
Bertl
so double side only applies to the sensor frontend, and the baseboard module(s)
19:02
se6astian
sounds like a plan
19:02
se6astian
but what if a module would have buttons on the left side and connectors on the right side?
19:03
Bertl
then that would have to be a wireless module :)
19:04
Bertl
from what I've gathered, buttons on the left side do not make much sense anyway
19:05
Bertl
professionals do not want that, because the CA is on the other side, and I'm not convinced that the indie folks will love buttons on their right ear or so :)
19:06
Bertl
(same goes for connectors :)
19:06
Bertl
but it is just an idea, and if we do the backplane at the bottom, you can have ear buttons and even BNC connectors which will hurt a lot :)
19:08
Bertl
so, SPI cable finished as well, which means that I will visit the lab this WE
19:08
Bertl
(to finally get the shoebox done :)
19:17
VBenjFR
You explain that you will make a module with buttons on the left side and connectors on the right side [20:02:50], and would like to open by top and bottom [19:58:2]. Wouldn't That be problematic, if you want install it on a rail?
19:18
se6astian
CA on other side agreed but if there is no CA the DOP requires controls on the left side
19:19
se6astian
so it might be best to have the dictator or whatever way is used to control the camera deal with all module controls as well
19:20
se6astian
second prototype ready <- hurray!
19:20
VBenjFR
greetings!
19:20
se6astian
Hi Benjamin!
19:21
Bertl
I don't see a problem with an external device controlling all the camera modules
19:21
VBenjFR
Hi Se6astian
19:21
se6astian
I made this concept earlier today: https://cloud.gerade.org/public.php?service=files&t=88644302ffde4f6e5659627a3c8bbf32&path=/Axiom/open-modules-concept-inside01.jpg
19:21
Bertl
we need to have some communication anyway, and I guess all will be available via ethernet or similar
19:21
se6astian
agreed
19:22
se6astian
ok gotta go afk for a bit now
19:22
Bertl
unless you add your proprietary interface module :)
19:22
se6astian
will adapt the backplane design to visualize the single/double backplane concept
19:23
se6astian
see you
19:23
se6astian
left the channel
19:40
VBenjFR
Good night
19:42
VBenjFR
left the channel
20:29
se6astian
joined the channel
21:42
Sasha_C
joined the channel
21:50
Sasha_C
left the channel
22:02
intracube
joined the channel
22:13
se6astian
left the channel
22:57
theverant
joined the channel
23:23
Bertl
wb intracube! welcome theverant!
23:24
theverant
hello, love the project
23:24
Bertl
glad to hear!
23:26
theverant
thought I'd pop in and see if there was any chatter. :D
23:27
Bertl
atm it is quiet, but you can use the pause to introduce yourself and how you found the project or what interests you most
23:28
theverant
I'm a tech at a small film dept in a fine arts university
23:28
theverant
we have a couple of Red cameras, etc
23:28
theverant
so I'm always scanning the horizon
23:28
theverant
in private life I pursue open source as much as possible
23:28
theverant
especially hardware
23:29
theverant
so an open source cinema camera is basically Nirvana for me :D
23:29
Bertl
yeah, sounds like quite a match
23:30
theverant
I'm not sure there's a lot I can do right now, but perhaps I might be some small help as things progress
23:30
Bertl
well, if you don't mind, I'd suggest you hang around here as often as possible, because we always need input and information
23:30
theverant
yeah, for sure, that's why I'm here :D
23:30
Bertl
for example, a few months ago, I had no idea whatsoever about cameras
23:31
theverant
I haunt #reprap and #arduino regularly, so I'll make this another stop.
23:31
theverant
so what kind of ideas do you have now? :D
23:32
Bertl
well, I finished the second prototype today (axiom alpha)
23:33
theverant
awesome! I did read on the first prototype and the image processing you were doing
23:33
Bertl
and I'm preparing the software so that we can actually make some movies with the thing
23:33
theverant
lovely
23:34
theverant
I have students if you need lackeys and want to send a prototype :D
23:34
Bertl
yeah, I guess we could send out 50 prototypes :)
23:34
Bertl
the problem is not that we do not want to, it is more the cost involved
23:35
theverant
of course, it's everyone's problem
23:35
Bertl
and for folks basically working for free a 1000 bucks or more are quite a lot
23:35
theverant
try running a film school on a severely limited budget :D
23:35
theverant
hmm
23:35
theverant
only $1k for a prototype?
23:36
Bertl
well, a little more actually, but the sensor was cheaper because we got a defective one (i.e. with known defects) for half the price
23:37
theverant
ah
23:37
theverant
I would be tempted to buy one for myself @ $1k
23:38
Bertl
you would have to build it yourself as well :)
23:38
Bertl
the PCB for example was ordered via OSHpark (you might know it)
23:39
Bertl
that made it quite cheap
23:39
theverant
Yeah I haven't ordered from them, but I know them
23:39
theverant
I have no problem building it
23:39
theverant
as long as I don't have to design the PCB :D
23:39
Bertl
well then, the PCB is already shared IIRC
23:39
Bertl
but I'd suggest you get the updated version
23:40
Bertl
you can then directly order it and assemble the frontend
23:40
Bertl
if you solder on the sensor directly, you can also safe about 60 bucks for the socket
23:44
theverant
hmm. Is this the sensor you are sticking with to the production version?
23:44
theverant
Do you have a price point target for the final camera?
23:44
theverant
you seem to have lifted a page from the Red design manual, as well, eh? :D
23:45
theverant
with the modules and backpack system
23:45
Bertl
you won't believe it, but we mostly arrived at the various design concepts by not looking at other designs
23:46
theverant
well, it wasn't meant as an insult, it is a logical form for the camera to take
23:46
Bertl
I suggested the modular concept when 'Red' was just a color for me
23:46
Bertl
yes, similar problems lead to similar solutions
23:47
theverant
I find their solutions to really lack ergonomics though
23:47
theverant
not unlike shooting with a brick
23:47
Bertl
and we think we are making the right decisions when I see that there is some overlap with existing designs
23:48
Bertl
for the price target, well, as cheap as possible without sacrificing quality
23:48
Bertl
the modular concept will allow for quite some price range, depending on what you need/want in your camera
23:49
theverant
I realize it is still early days to really talk about price
23:49
Bertl
the most expensive parts are the sensor and the FPGA
23:49
theverant
but it's usually the deciding factor :D
23:50
Bertl
sure, but while I'm certain that we cannot match price dumping as done by some companies, we can definitely match quality and offer the additional bonus of being open
23:51
theverant
yeah that will certainly count
23:51
theverant
I think what really caused Red to rise so fast was the community they built around their product
23:51
theverant
as well as having groundbreaking tech
23:52
theverant
but sharing ideas early on, as you guys are doing
23:52
theverant
it seems to be something the grand old companies don't understand
23:54
Bertl
people gave up a lot of freedom already for vague concepts like intelectual property and security :)
23:54
Bertl
we believe it is time to get back some of those lost treasures
23:56
theverant
Yeah, no kidding
23:56
theverant
the 3D printing world is seeing big waves these days
23:56
theverant
patent trolling
23:57
theverant
it's all about control and money. It'll be sad if the reprap movement gets damaged by that
23:58
Bertl
so much for free world :)
23:58
theverant
Do you guys have access do cameras like the Red, BlackMagic, Canon C class, etc?
23:58
Bertl
anyway, we are doing our best to change a little bit here and there ...
23:58
Bertl
some folks do I guess
23:59
theverant
Well I was thinking for testing
23:59
theverant
comparing
23:59
theverant
as you do image tests
23:59
Bertl
probably too expensive atm, might be an option lateron
23:59
Bertl
maybe we also get some kind of hardware donations later