#apertus IRC Channel Logs

Bertl_oO: How goes?

it goes slowly but steady ...

didn't do much work on the sensor yet, mostly design and testing for assembly

I don't think so, but you'd have to check with se6astian (who is currently away, despite the nick)

Bertl_oO: Will be interesting to see where this project goes.

there is a tool to convert the raw images to DNG .. some folks compile it for the camera or use it on their laptop

Bertl_oO: troy_s has already told me his interest in the BM Ursa :P

Bertl_oO: Not sure what serious work is. If you and I and Gabe and the lot sit in a room, then I'm all up. Grab us a DSC and lets go. If it involves floating around the rubbish of flotsam and jetsam junk that no post production pipeline gives a rats ass about, then not so much.

Bertl_oO: The whole DNG thing is *such* a fools errand. But I know that in your head you already know that.

regarding 'sit in a room', we can arrange that, virtually and physically if you like to travel (or we find the budget to travel to you)

Bertl_oO: Have you heard from Gabe lately?

did some work with preview via Android and such, but I haven't seen anything yet

I forgot who else was hammering on the colorimetry... it was irieger?

Bertl_oO: Has he been active?

It was mostly summer season the last few months with everybody going on vacation

Bad luck.

intracube: Oops. Missed your messie. How you doing P?

Bertl_oO: What's quite interesting was the similar rubbish the BM cameras were doing with those skewed pixels in the higher end. Biasing the highlights. Were you ever able to get the sensor to be less irregular up there?

Bertl_oO: Or were fool's errands at work? Should have just chopped and moved on to a completely stable, chromaticity correct LUT. (Via DSC chart)

i.e. you don't want to 'throw away' data unless you really have to

Not data at all, as you said ages ago.

If the BM folks, with a solid bit of capital behind them and some decent imaging engineers can't get it to behave and instead are clinging to the marketing rubbish of stops, I'm pretty convinced that a small team of renegades that are under-funded would probably be prudent to simply chop and focus on post production stability.

Guessing binned photosites is junk.

Bertl_oO: do you mean the sharp roll-off at the very top end?

:)

intracube: yes, it is a bend/knee, but it keeps going steady

Bertl_oO: Bottom line, did anyone get stable xy values?

Bertl_oO: Without artificially mangling and massaging the data off the sensor?

(I'm skeptical given the Ursa has the same gibbled up sensor and the transform yields irregularity there too.)

Bertl_oO: It's not the nonlinearity that is a factor.

Bertl_oO: Getting a 3D LUT to properly get *stable* and *accurate* xy values is effectively a trivial thing.

(If that means chop, so be it.)

if you want to use the sensor to its full extent, then you have to deal with the few dozen values beyond that bend as well

s/zz/ss/

Bertl_oO: It's that stability factor that you saw *way* back when.

please refresh my weak memory ...

danieel: Greets. Hope you have been well!

morning folks!

hi

on my side good :-)

a little bit of 'home improvement' vacation yes :)

at least something that makes the brain fresheee :-)

the main reason is that it was not optimized for production, i.e. high number of PCB layers, expensive components and large board areas

hey, do you still work with FPGAs? :)

a friend of mine is a prof. dev with fpga.

that was his first project after school: https://labs.ti.bfh.ch/gecko/wiki/modules/gecko3main/start

very appreciated!

btw. where do you have at the moment most resource problem?

despite the fact that we got a lot of Early Betas into developer hands

neat! :))

heya

do you need FPGA designers?

glad to hear! and yes!!! ;)

yeah, you already knew ...

let me just wrap my head around the project a bit more

I mean, one question could be: how can I get a Axiom Beta now that the Indiegogo campaign is long over?

hehe, that is probably the simplest to answer

so there's like a waiting list?

how many currently waiting?

I have to check, but the number is small .. various reasons, some change their mind, some get into monetary problems and need every cent

what is the probability of doing an indiegogo campaign for the Axiom Gamma?

hard to say, but at the moment we focus on the AXIOM Beta

(and the eco system around it)

no repo yet, indeed, but the designs are FOSS/OH

my internet seems to have gone away for a few minutes, I probably missed if you guys said anything in the meantime

Open Hardware

stupid me :)) ofc

yes, they are

they always have been public

but not on any github repo?

I can't seem to find it on github anyways

(at least not yet)

so how could I help? what info do you require from me (if any)?

just checking looks nice :-)

yep, that's what I was thinking about

the HDL code we are using for the Beta is roughly the same we developed for the AXIOM Alpha, so a major rewrite is in order (sooner or later) to unleash the full power of the device

I've applied for an account, so I can contact you. Herbert Pötzl, I presume?

yeah, it's not the only weird thing ... complaints go to se6astian (when he's around or via e-mail :)

off for a nap ... bbl