#apertus IRC Channel Logs

I am working on the qualification task T1162, and have a doubt - The icon image being 1-bit depth, is it necessary to store it in an 8-bit array? Or can we use STL and store it in bitset?

hello preetimenghwani[!

and BAndiT1983|away will try to answer that one once he gets back ...

don't forget that the task is meant for a pic32, so the things have to be slim and fast

meeting in 20 minutes

meeting time!

who is here?

great, maxime promised to attend today as well

yay!

hello everyone!

hi

good timing

by 18:30 iirc?

yep

so, we resume the AXIOM clubs with Bertrand

and, as we discussed with Sebastian in feb, we are going to use the Beta for time-lapse recordings and stop-motion animation movie shooting

and, with the help of a little script we crote during previous AXIOM Club, the students manage to make 92 frames of a simple animation shot!

!

nice and some hands are there too ;)

(and I intend to do the same with my own students, which are studying stop motion at La Cambre, a fine art school for future art professionnals!)

so, the script was converting the raw12 in PNG images (for the ease of a studentproof postproduction workflow on KDENlive)

and we had issues with the color because we were using the line in the wiki using IM convert

thx!

how far did you get with ufraw/dcraw?

btw, adding -define sample:offset=0 should make it work again

(the image00044.TIFF, and/or the jpg I converted from that)

it's command line and super efficient, but hard to master since dcraw doesn't do color temperature/tint based correction, but something else

-r mul0 mul1 mul2 mul3 Specify your own raw white balance. These multipliers can be cut and pasted from the output of dcraw -v.

will investigate it !

well if you wait 0.5 or 0.7 seconds is probably not a big deal :)

which considerably easier, but take more time to batch process (yet)

I was wondering who worked on color calibration so far, in order to get in touch with him/her to go further without re-inventing the wheel!

intracube: ?

that's for me !

we are going to re-do AXIOM club every two weeks or so, and invite Oscar asap!

any news on the promo material printing front yet?

btw max started this tutorial a long time ago, maybe that would be a good way to document all the things around capturing and processing still images with the beta: https://docs.google.com/document/d/1-kFjebwvUg3e8jLKwXSemBWsNIL6_P-NPFUXbv7-jmk/edit#heading=h.p0drc1axbzga

in the future

otherwise we are right on time for you to leave by 6.30

BAndiT1983: any news to report?

yes, a bit

started to straighten out the painter code, as the origin of drawing was moved to top-left, as some questions arose lately, it was done for performance reasons and hardware structure

awaiting challenge code from students, before the evaluation starts, also proposals should trickle in soon

that's it from me

oh i am sorry to interrupt

Bertl: anything you want to share?

I've spent most of my time on GSoC prepartions in various ways

then, as there is plenty of interest in 'remote' tasks, we decided to do another round of AXIOM Remote prototypes

so hopefully we will soon have a remote remote :)

https://644db4de3505c40a0444-327723bce298e3ff5813fb42baeefbaa.ssl.cf1.rackcdn.com/8bec86edbcde6d5c021249a977852f75.png (pads on on the borders)

https://644db4de3505c40a0444-327723bce298e3ff5813fb42baeefbaa.ssl.cf1.rackcdn.com/14f6c1cd7544f7783f6c796cabbf7fdd.png

https://644db4de3505c40a0444-327723bce298e3ff5813fb42baeefbaa.ssl.cf1.rackcdn.com/1ff5e963a8839691c0ddb598e35687d9.png

have fun! cya!

bye maxime

PCBs have been received (last one got in right before the weekend) and the parts are here as well, so assembly will be started pretty soon

just to remind you, the design is for two rails with 5V/3A each, for input voltages between 6 and 17V

but this is not really advised, as the regulators reach temperates above 100 degree celsius easily

with moderate air cooling and/or heat transport via skeleton, we can reach the full 3A per rail

so I think we should be good there.

Next step here is a fully assembled power board with the changes which I plan to do this week after preliminary testing the received PCBs

I think that's it from my side.

vup/anuejn anything from your side?

last week I asked about ways and places to adverize gsoc (universties, mailing lists, etc.) did anyone have any news ideas already?

great, I asked more people about a stereo3d beamsplitter rig in vienna but so far nothing useful

se6ast1an: yep, here for a bit.

That's all!

intracube do you want to share a summary of the color calibration you were working on ?

ok, just a brief overview as I haven't prepared;

for now the adjustments are just by eye and using video levels/scopes for greyscale tracking

the second part is to take the adjustments (currently being made in Fusion/Resolve) and create a LUT + matrix which could be loaded onto the camera

is the "basic camera matrix" embedded in a DNG ?

or could be embedded (not sure we do it currently)?

ok

I think that's about it for now

quick update from my side:

maker faire vienna is still scheduled

video done, reviews pending, article to be created

thats from todays glas filter cutting visit to the glazier

his cut was not perfectly straight but the glass didnt shatter

none the less it worked out fine for a test

will need to investigate, anyone with experience with such a glue?

https://cloud.apertus.org/index.php/apps/gallery/s/LqkCfTLRJYd9wFB#shim%20design

isn't that an UV filter?

you mean we wouldnt be able to cure it? :)

good point :)

https://cloud.apertus.org/index.php/apps/gallery/s/jycWfbwpWYAwfJ6

and the NATO Rail system

if you need that why not ;)

anyone else, otherweise meeting would be concluded

thanks for having us!

quick, to the supermarket!

and tons of stuff from the supermarket of course!

BAndiT1983: I read about bitset in STL. If there aren't any bit level operations involved (flip, set,reset), using bitset will be inefficient. Also, in a 32-bit processor, it allocates memory in multiples of 4 bytes. So from memory usage point of view, an array of 1 byte data-type will be more efficient in this case

not everything which is comfortable for development will be efficient, but for image data plain array is the better solution, STL data storages can be also good but this also depends on the compiler and the environment it would be used

pratyush: some kind of CRC is fine, even just summing up the bytes would be okay

off for now ... bbl