#apertus IRC Channel Logs

morning folks!

alright, have a nice meal

maybe it's something to bring up at the next irc meeting

it's telling that one of the first youtube comments is about moire/aliasing on the water sequences

other than rare speciality cameras (Nikon D800?) where you have to choose your shooting subject carefully

and the extra detail is essentially false detail

"lets ignore Nyquist and digitise the signal unfiltered"

BAndiT1983: it feels like it's being sidelined a bit

if it would be _that_ simple, we'd probably already have that option :)

for a start, let's find a suitable OLPF for the CMV12k which can be easily used with the current design without any changes

$300 seems to be a rough figure

and even if it was $600, if it's a necessary element (which it is), it needs to be factored in :)

red ones are 195$

well, what do you call it when you say "it is essential, add it!" without putting any work into it?

Bertl_oO: I'm just trying to get a debate on it going :)

yep, seems very reasonable, at a quick glance

well, we had one a long time ago, and it ended with ... nothing affordable which would easily fit ... we need to do some testing ... not really required ATM

I'd be happy to have a solution which can be provided to our backers, so go ahead and find one :)

and can be weighted into the design process

We can't make our own OPLF?

http://www.optics-online.com/lpf.asp

there was also a company called Mozaic

300-400$

yeah, significantly more

might be possible to organise a bulk order discount

but realtime AA sounds like digital processing

I don't think it's possible

Ok, normally run them over 24 hours, so...

or you might loose more detail than if you used an OLPF

the whole OLPF discussion on the internet has no real resukt, some like it, some like it when they remove it from camera

thats why I planned the OLPF to be removeable in the beta

as you still need IR/UV cut off filter without OLPF

but its unlikely you will find those two with matching thickness and density

and therefore require the lens mount to be changed

yes, that will change FFD even more

I wonder how Nikon D800E owners deal with it

does the D800E have an OLPF that you can just pop out/in?

hm, seems not :/

"Nikon started the mainstream AA-filterless trend with the D800E and its AA-filtered D800 twin. "

yup

although it seems like it's optional on some of the Red cameras

Poll will run for four days - https://twitter.com/apertususer/status/848583403989598209

Roopal08: i'm on it, just give me a moment

Yes.

RexOrCine[m]: maybe change real-time for physical?

or just remove the real-time, we have "optical" already there anyway

"How important is an Optical Low-Pass Filter (OLPF/AA-filter)"

That means we lost those two votes - https://twitter.com/apertususer/status/848584997388812289 settle down man.

Oh. Sorry .

no problem

RexOrCine[m]: ok, no worries

So i need not add it in my deliverables ?

i think i will add some feature which allows to create panels with parameters, ike rawtherapee and darktable do

ok. I will just integrate my modules with OCProcess

but you can still adjust the UI if there is something not fitting the needs

what were the two early votes for?

Nah you can't edit them mate. This is a long standing issue actually because its been being demanded for years. I suppose Twitter is about trapping someone's thoughts from a specific moment, and it wouldn't be as solid if people were running around editing what they've said.

erf, should've left the original then :)

nope, BAndiT1983 still here

no post-factic? ;)

<BAndiT1983|away>, Do i need to add implementation details in my proposal ?

Ahh

Thanks :) . Anything else I can improve upon ?

I will use the one in the CPU

possibly not, if something won't work as expected then i can counter-check it on my PC

<BAndiT1983|away> Ok.

what are implementation details for you?

reminds me of "The Lawnmower Man"

Green Edge Directed Demosaicing Algorithm : http://www.academia.edu/18351844/Green_Edge_Directed_Demosaicing_Algorithm Efficient, high-quality Bayer demosaic filtering on GPUs : http://graphics.cs.williams.edu/papers/BayerJGT09/

ah, that things, yes, definitely, my impression was something about internal classes

Will add them soon

this one is also important -> http://graphics.cs.williams.edu/papers/BayerJGT09/

Roopal08: one more suggestion: please put emphasis on the actual deliverables in the schedule

or similar

Okay

Hello!

Hey Bertl, I have updated my proposal a bit: https://github.com/usmanwardag/gsoc17_proposal/blob/master/proposal.pdf

I am still working on refining the language. If anyone can review my proposal for that, I would really appreciate it.

I will correct that

Ok great, I will see if I can further refine it till tomorrow's deadline

off for a nap ... bbl

<se6astian> I have updated the proposal with all the changes you suggested. If possible can you go through it once again.

color code, symbol

whatever you find suitable

good night