#apertus IRC Channel Logs

hello...

you are very early :)

should we start right away?

and hi.

have pushed new update with direct register config for CMV sensor, but haven't tried on real camera yet, this update would allow to port scripts to daemon managed config

sound good.

nevertheless, ideas are there and will start to implement it as next step, after the test of latest update

made a simple setup and it works.

elements have view and edit mode.

is it intuitive from point of view of a user?

yes, please make a video, as i don't have a setup for the test at the moment

same here, haven't started to check possible daemon crashes more thoroughly yet

you mean daemon or your test server?

test server or a daemon that can serve the data

i will run it before vienna, have SSH access and live HDMI feed

i don't know ninja when it's better the make it's surly a good alternativ

will investigate the build problem with ninja a bit later

*your

are there other task that we need to discuss?

make the dev server more mature for testing the commuication / error handling websocket / error handling / syncing database / some small CSS / redo searchbar to solve some bugs / add favorite list / redo the view for all command list

first we needs stable base

no i haven't, because for me it would be hard to debug when something goes wrong. i'm not a C/C++ coder

just need someone else to test it besides me and maybe suggesting improvements for output, as it was done for debugging on the camera first

at the bottom is the command for deactivating camera-specific functions -> https://wiki.apertus.org/index.php/Control_Daemon

ok, by the way, my test was directly on the camera with lightppd

okey

just ask if you have questions or problems there

but on the camera you could test if you can set gains correctly

you can also connect to the HDMI feed, but first se6astian or Bertl_zZ have to add your SSH key also on capture PC

btw what are the targets when you are in vienna?

but also walking around in vienna, it's still my vacation

nope, second, first time was in june last year

nice, on my side i'm still open for coming, need to wait till a get the plan for next week for some short term work.

yes me too i like the after work drinking :-)

hi parimal, please check the logs at irc.apertus.org, i've tried to answer your question while you were off

Yes, I meant the version of Qt

version is not that relevant, have done latest tests with 5.10 or 5.11

okay

Hello

mrohit[m], what do you want to know?

the order is described there, first row is RGRGRGRG, second row GBGBGBGB

Why am I seeing garbage values when I print out the loaded buffer?

I mean is it the way it is supposed to be?

or you can shift them right by 4 to get 8-bit values

have you tried to open the RAW12 data with gimp?

I can try that if it helps

Actually I need to install a RAW loader to my gimp, so I will configure that and check

you don't really need a raw loader, it is included often

Okay, I didn't knew this.

i would try to decode 12bit to 16bit and then try to open, also swapping bytes

of course, maybe imgur

second note, you have to select "gray unsigned 16bit little-endian" as format

if you convert to 8bit or 16bit from raw12, then you should also be able to open the data in gimp in the same way or if it still gives problem, then big-endian

tried with IT8-chart-15ms.raw12 from here -> http://files.apertus.org/AXIOM-Beta/snapshots/

the image is a bit recognizable in the preview window, you can see the outlines of a chart

Bertl_oO, as the task is to process it by C++

thought you were trying to open it with gimp ...

why should i use terminal and search for the commands, if i have some UI?

outlines of the image are recognizable, it should show a guy in front of the shelf with different stuff on it

like the image in the wiki shows, first 3 bytes contain 2 sensels (pixels)

you have to split this 3 bytes into 2 values, either 16bit or 8bit, at least to see if the content is correct

vup2, in this case it's okay, as grayscale is not showing some meaningful image, but RGB allows to see at least some approximation

should be a lot clearer what is actually contained in the image

not really, as gimp relies on 8bit grayscale or 16bit

then that makes sense

yes, I tried with grayscale as well as RGB, RGB seems less noisy

try rgb565

but it's just a bare preview and not correct

the image should look like this at the end (thanks to anuejn) -> https://github.com/anuejn/batic

Thank you, this discussion helped me to clear out the clutter

good night