#apertus IRC Channel Logs

http://scien.stanford.edu/pages/labsite/1999/psych221/projects/99/tingchen/main.htm

yes it is

the explaination might be enough to write the code for it

you mean if I dump the data at full resolution?

can we call them photosites?

well, we can call them what we want, the manufacturer calls them pixels

but I'm finr with photosites as well, no problem there

I've even read sensel somewhere, which I consider funny :)

sebastian and I started using that term after I got confused as to the amount of photosites/pixels in the sensor initially

not the same for other technologies like X3 or else

bits

well, we do 4:2:2 RGB there :)

I usually consider a "pixel" to be a single full-color dot

each color in the HDMI output is a sub-pixel

I suppose one could also call each single-color dot on the sensor a sub-pixel, but photosite seems a better fit

sensels are SENSor ELements, pixels are PIX, picture ELements

blocks of 5x5 !

that's the root of my question about the area

I was wondering how the algo's accomodate this

it just picks red 'pixels', blue 'pixels' and 'green' pixels and combines them to an image, no interpolation

one step further, you get alhorithms looking way beyond one pixel to 'guess' the missing information

http://scien.stanford.edu/pages/labsite/1999/psych221/projects/99/tingchen/algodep/vargra.html

really ?

yes, it is all guesswork and a little voodoo :)

(well it is, but not by method)

i.e. if you take a sharp picture and run a blur over it, you throw away high frequency bands

that's why there are so many image sharpening algos out there, one works better for this case, the other better for the next

let me give you a simple but intuitive example:

consider an image (artificial test image), which has a red point on every blue photosite, a blue point on every green photosite and a green point at every red photosite

but the image itself will look white to the observer, and it defintely is neither

I think I get the idea of this example

and now I'm off to bed ... have fun!

sensel

FergusL: dcraw has a VNG algo you can try.

but tbh a clean explanation like the ~tingchen page provides seems better

I trust you, I've seen your tests with this in Blender

there are packages for oiio in ubuntu now

FergusL: The scaling algo is (it is a frequency domain scale) but there are some nuances

the image isn't a pure set of "identicals" but rather three sub images with different citinh

So to maximize it, the cubic b with pre would need some fractional adjustments based on the lower level sensel positions

(technically they all resolve to subpixel positions, so a "perfect" chroma scale would be slightly shifted to perfectly align the scaled B and R planes to match with the G)

(where G could be considered to be a half res perfect citing)

erf

Full Y

OIIO is for pixels

Make sense?

I had two questions : how about the borders of the picture ? simply drop and keep an images that is slightly smaller than max resolution ? or interpolate differently ?

explain?

well, if I'm at pixel 0,0

And regarding the second, a correct conversion to float would likely suffice. And I would do that pre-debayer as your data will be deeper.

it's an array, an x,y array

I know for some scales you offset and assume black

or do linear interps (as in assume identical pixels near borders)

ok

that's what I will be doing I guess, floor() and ceil() the X and Y values so that it "repeats" the borders

I think that eliminates shadow darky fringes actually - just assume where < min or > max = min ? max

that properly weights the edges

okay, got it

as in > max y AND > max x eg

at which point an unsophisticated lerp should suffice

(effectively lerp stretching the corner pixel with the adjacent row or column pixel)

that has yet to be tried anywhere

I'll consider this

the cbp ?

yes, it seems obvious

haha, yes

(where interpolated accuracy is desireable)

can you elaborate on implementing cpb for debayering ?

FergusL: Very simple

Take each plane of R, G, and B, convert to float, and scale accordingly.

Scale R to 2X and 2Y

then merge for a "phase one"

Which may be tricky... as assume we are looking at the upper most Bayer row of say, RGRGRG

the first R and G actually have different parts of the image - ever so slightly offset (different citing)

To perfectly align the red channel, we actually want the interpolated value between the first R and the second R.

And it gets equally ugly going down

Well think of it this way

if we took a full sensor as a mono (no filters)

we have a perfect 4k image or whatever

each sensel represents a perfect greyscale value

so RGRGRGRGRG then GBGBGBGBGB etc.

the first R and G represent _different_ parts of the image

very close together, but different parts

not just three different planes of RGB

(and in a semi complex checkerboard fashio

ok, this time I get it

so to marry the RGB, you have to be acutely aware of the sensel layout on the sensor

(hence citing)

We would actually want to stretch the green a sensel to the left on the top row

to perfectly align with the green on the second row

but we don't have the data for it!

(because you could then at least reconstruct 50% of the edge pixel based on the diagonal of greens for example)

yes

so it could very well already have a complex solution buried in it

(and given that it is in the frequency domain, who knows the result. for certain the x positions should ideally be shifted a sensel over on alternates... or generate two proper aspect images and scale each then merge into a single channel)

Which boggles my mind.

Anyways... night.

FergusL: yeah, i'm interested working on that code

good morning

oh

all I did was sleeping since we last talked about it

and to add confusion, the sensor can do color binning, which will give overlaping bins per color channel :)

hey

not some shady math unknown operator ?

looks like, as there is nothing else specified, in 'normal' math those are simply absolute values

good evening

evening!

they do not have to know everything about the project and the technical details, they just have to know whom to ask for details

questions and statements in general, otherwise there will be a lot of speculation and drawing (mostly wrong) conclusions

but so far there is nothing to comment to

"they’ve got images in 4K. However, just like me first thing in the morning before I put on my makeup, they’re looking a little rough. "

don't you wear makeup every day? :P

aah... good ol time when Koo was writing every single article himself

Bertl, I just checked the lens mount files konstantin sent me and they are indeed intact NURBS curves when openend in Rhino3D, and Rhino can make edits and save as STEP file again

I just never used Rhino before so I have no idea how to edit the curves yet, but everything is there to make these changes

I am just writing back to John Burt

ATM, there is not much to be tested/helped with, i.e. the analog devices chip works as expected and is what we will use for the prototype

actually we never wanted to have HDMI output but it was what the zedboard had already ;)

at some point maybe....

the reply-to options now work by default in my gmail as well btw

we wont hit our record from the open modules concept

more coming on monday when office hours continue most likely

yes, this time we didnt give the writers any headstart

lower resolution and smaller sensor diameter than the cmv12000

LCC carrier might be a little problematic for testing

(but at least it has the global shutter option :)

the package the sensor sits on?

less than half the number of photosites, 1/3rd the max framerate

maybe contact BAE and get a detailed data sheet?

yup

cmv12000: Dark noise 13 e- (RMS)

CIS: <30 e-/pixel/sec dark current @ 20°C

no the dark noise and the dark current can be compared

so at roughly equivalent temperatures (CMV12k needs to be cooled for that)

the cmv12k has roughly 4 times the pixel nois of the CIS

the read-out noise is something different

but it is hard to tell for both what precisely is meant

(in the analog domain)

the range is interesting, with 30ke- full well charge and 86dB

but for now, time for bed ;)