23:11 | darthrak_ | joined the channel | |
23:25 | darthrak_ | left the channel | |
00:26 | darthrak_ | joined the channel | |
00:30 | darthrak_ | left the channel | |
00:51 | slikdigit | joined the channel | |
00:52 | slikdigit | left the channel | |
00:52 | slikdigit | joined the channel | |
02:27 | darthrak_ | joined the channel | |
02:32 | darthrak_ | left the channel | |
04:35 | hozer | left the channel | |
06:28 | darthrak_ | joined the channel | |
06:33 | darthrak_ | left the channel | |
06:42 | jucar | joined the channel | |
06:48 | se6astian|away | changed nick to: se6astian
| |
06:49 | se6astian | good morning
| |
07:07 | jucar | left the channel | |
07:39 | darthrak_ | joined the channel | |
07:46 | intracube | so that bottom right discolouration here: http://i.imgur.com/j23B4rt.png
| |
07:47 | intracube | is visible in the nonlinearity images after a contrast stretch: http://i.imgur.com/S8g32Eo.jpg
| |
07:47 | intracube | only in the green channel, though
| |
07:48 | intracube | a diff of green with either red or blue channel + further contrast stretch: http://i.imgur.com/3Qvbd1x.jpg
| |
07:48 | intracube | ^ alexML
| |
07:54 | alexML | yeah, I need a better clip frame (I'm experimenting with something that accounts for nonuniform illumination of the wall, but no meaningful results yet)
| |
07:55 | intracube | isn't that green a sensor issue rather than scene illumination?
| |
07:56 | alexML | I'm also trying to recover the response curve with other methods; unfortunately, with existing programs like mkhdr or pfshdrcalibrate, I've got only bogus results
| |
07:56 | alexML | so I'm reinventing the wheel :P
| |
07:56 | intracube | heh
| |
07:56 | alexML | it's not a sensor issue, it's just the calibration frame
| |
07:56 | intracube | oh ok
| |
07:57 | alexML | (I'm working on existing samples, remember)
| |
07:57 | alexML | (I don't have a beta yet)
| |
07:57 | intracube | ^ soon though :)
| |
07:58 | alexML | I also thought at something that would flip the exposure back and forth near the clipping point - if I let that run for a while, I guess it should give pretty exact per-pixel curves
| |
07:59 | alexML | (there is plenty of noise in that region)
| |
07:59 | intracube | you mean the granularity in the highlights (particularly saturated areas)?: http://i.imgur.com/qVfdkGs.jpg
| |
08:00 | alexML | yeah
| |
08:00 | alexML | that throws off the calibration, so if I want to get per-pixel curves from the existing dataset, there's a ton of noise in the data
| |
08:01 | alexML | but if you repeat that experiment 16 times, that noise would be reduced by 2 stops
| |
08:01 | alexML | if you repeat it 64 times => noise reduced by 3 stops
| |
08:01 | alexML | and so on
| |
08:01 | intracube | so in theory, it could be reduced a fair bit
| |
08:02 | alexML | well, I want to get the exact (unbiased) curves first
| |
08:02 | alexML | so the noise would have zero mean (that is, so it won't introduce color casts)
| |
08:02 | alexML | once you have good average colors, you can just denoise those highlights a bit
| |
08:03 | alexML | and then there's the repeatability issue
| |
08:03 | alexML | that needs to be checked, how it varies with exposure, temperature, whatever
| |
08:04 | intracube | and this profiling would have to be done for each individual sensor
| |
08:04 | alexML | and if it's just a simple offset, maybe it can be detected on the fly
| |
08:04 | alexML | yeah
| |
08:04 | intracube | and also possible changes as the sensors age
| |
08:04 | alexML | same for dark frames
| |
08:04 | Bertl_oO | changed nick to: Bertl
| |
08:04 | alexML | yeah, I want to have these calibration routines in the camera
| |
08:05 | intracube | is this extreme non-linearity in the highlights typical for cmos?
| |
08:05 | intracube | I thought it was generally linear
| |
08:05 | Bertl | it isn't a problem of the photo sites, it is a problem of the ADC and gain
| |
08:06 | alexML | on Canon, it seems to clip quite harshly
| |
08:06 | intracube | Bertl: ah
| |
08:06 | alexML | on 5D3, I turned down all the gains I could identify, so it no longer clips to white
| |
08:06 | Bertl | or to be precise there are several effects going on
| |
08:06 | alexML | and got some fixed-pattern noise in highlights
| |
08:06 | alexML | but after correcting that, there was no useful data to recover, like on the CMV
| |
08:07 | Bertl | first, the CMOS charges a capacitor to a certain level
| |
08:07 | Bertl | then the photo diode discharges it when light is received
| |
08:07 | Bertl | this process is not really linear per se
| |
08:08 | intracube | ah ok
| |
08:08 | Bertl | also, it has a roll-off when the capacitor reaches zero voltage
| |
08:08 | Bertl | because the diode adds a certain offset there as well
| |
08:08 | Bertl | assuming that the actual sensel is operated in the linear range
| |
08:09 | Bertl | (or to be precise, in the range where it acts as expected)
| |
08:09 | Bertl | you'll get a nice and proper voltage to digitize
| |
08:09 | Bertl | this is where the second stage brings in non linearities
| |
08:10 | Bertl | i.e. you apply analog gain, which has nonlinearities at the rails
| |
08:10 | Bertl | and you also run the ADCs, which again have (usually quite small) nonlinearities at the limits
| |
08:11 | Bertl | given that everything is optimal, you will only see a proper linear range
| |
08:11 | Bertl | the CMV is a little odd in this regard, as it allows for a number of analog gains, which are well below the ADC range
| |
08:11 | Bertl | this
| |
08:11 | Bertl | per se is not a bad thing
| |
08:12 | Bertl | it just means that the full 12bit range will not map to a linear range in all gain settings
| |
08:16 | intracube | interesting, thanks for the explanation :)
| |
08:17 | Bertl | you're welcome!
| |
08:23 | darthrak_ | left the channel | |
12:31 | darthrak_ | joined the channel | |
12:33 | darthrak_ | left the channel | |
12:34 | darthrak_ | joined the channel | |
12:35 | darthrak_ | left the channel | |
12:39 | ar | left the channel | |
13:33 | jucar | joined the channel | |
14:46 | jucar | left the channel | |
15:17 | jucar | joined the channel | |
15:22 | se6astian | changed nick to: se6astian|away
| |
15:33 | jucar | left the channel | |
15:42 | Bertl | off for now ... bbl
| |
15:42 | Bertl | changed nick to: Bertl_oO
| |
16:22 | se6astian|away | changed nick to: se6astian
| |
17:07 | jucar | joined the channel | |
18:49 | se6astian | anyone currently here speak spanish?
| |
18:51 | Bertl_oO | Lo sentimos, no hablo español! :)
| |
18:52 | Bertl_oO | or maybe better? ¡Lo siento, no hablo español!
| |
18:55 | intracube | dos cervezas, por favor
| |
18:56 | Bertl_oO | LOL
| |
19:08 | troy_s | Bertl_oO: Yo
| |
19:09 | troy_s | Bertl_oO: intracube noticed that there's some odd glow in the lower right corner. Any clue as to what this might be? Light leak / splash off of an LED? Heat? Electronics?
| |
19:09 | davidak | joined the channel | |
19:09 | troy_s | Bertl_oO: http://i.imgur.com/S8g32Eo.jpg
| |
19:10 | jucar | left the channel | |
19:20 | alexML | troy_s: that corner was not completely overexposed, but the rest of the image was (obvious if check the other images in the set)
| |
19:21 | troy_s | alexML: I don't mind. I'm more interested in what causes it.
| |
19:21 | alexML | once you go past some overexposure level, the sensor output starts to decrease
| |
19:21 | troy_s | I'm more interested in the fact that it is there at all.
| |
19:21 | troy_s | And why.
| |
19:21 | troy_s | Could be an LED leak like was had in the early Alpha.
| |
19:22 | troy_s | Or heat, or magnetic magic fubar or who knows.
| |
19:22 | alexML | check the nearby samples (render them in the same way and play the movie)
| |
19:22 | alexML | you will no longer blame cosmic radiation after you do that :P
| |
19:24 | intracube | http://i.imgur.com/F0oLpFD.png
| |
19:25 | alexML | check from 30 to 90ms for example
| |
19:25 | alexML | or, even better from 1ms
| |
19:26 | troy_s | alexML: I don't blame anything. I'm just leaving everything open. Those shots are hooped.
| |
19:26 | troy_s | It is upper right too.
| |
19:26 | troy_s | So could very well be an LED leak off of one of the boards.
| |
19:26 | troy_s | Or any other thing.
| |
19:27 | alexML | it's not visible in dark frames, so I doubt it's a LED leak
| |
19:27 | troy_s | Dunno.
| |
19:27 | troy_s | That's my point.
| |
19:27 | troy_s | Bertl will have good insight.
| |
19:27 | alexML | but please - see the entire sequence before judging
| |
19:27 | troy_s | How come?
| |
19:28 | troy_s | Does it flicker?
| |
19:28 | intracube | I thought the 'nonlinearity-gain1x-offset' images were dark frame?
| |
19:28 | alexML | no, they are not
| |
19:28 | troy_s | Anyways... off to pick up damn Cintiq number THREE for the love of god.
| |
19:28 | intracube | ooh
| |
19:28 | alexML | they are a blank out of focus wall
| |
19:29 | troy_s | Dark frames are portal capped images then?
| |
19:29 | troy_s | (I take it)
| |
19:29 | intracube | ok, that does change things slightly. but this issue only seems to be on the green channel
| |
19:29 | intracube | nothing on red or blue
| |
19:29 | troy_s | That's odd.
| |
19:29 | troy_s | If that's true then it is a magical source.
| 19:29 | intracube | will have a look at the darkframes
|
19:30 | troy_s | To only impact one data channel would be pretty damn magical.
| |
19:30 | alexML | they all behave the same electrically, but green is usually receiving more light (hint: white balance)
| |
19:30 | troy_s | White balance?
| |
19:30 | troy_s | This is off of the raw data.
| |
19:30 | troy_s | So how would white balance even factor in pre-transform?
| |
19:30 | alexML | yes, if you render raw data without white balance, what color cast do you get?
| |
19:30 | troy_s | You don't get any colour cast
| |
19:30 | troy_s | There's no colour.
| |
19:31 | troy_s | You might as well be dumping a musical track to an sRGB screen.
| |
19:32 | intracube | just to confirm; I'm not seeing anything on a dark frame image
| |
19:32 | intracube | (darkframe-100ms-gainx4-10.raw12)
| |
19:33 | troy_s | intracube: Then it must be across all data channels if it is a leak
| |
19:34 | intracube | yep, unless it is a fairly pure green light from somewhere
| |
19:34 | intracube | but you'd expect to see something on the other channels, I guess
| |
19:34 | alexML | again, it's a blank wall
| |
19:35 | alexML | in the middle of the image, r=g=b (roughly, on raw channels, no colorimetry)
| |
19:35 | alexML | in that corner (not yet clipped), green starts to clip, red and blue are lower
| |
19:35 | troy_s | alexML: So it's a leak?
| |
19:35 | alexML | what color do you get when you render with wb multiplers 1,1,1
| |
19:35 | alexML | no
| |
19:36 | alexML | it's not a leak
| |
19:36 | Bertl_oO | well, I was wondering about the corner as well (on the image intracube "corrected")
| |
19:36 | troy_s | Bertl_oO: More like amplified.
| |
19:36 | alexML | that's how a white paper appears when it's close to overexposure, and uniwb
| |
19:36 | troy_s | It's got to be a leak.
| |
19:36 | alexML | try it on Canon, Nikon, whatever you want
| |
19:36 | troy_s | If it isn't on the dark frame, and it is on the other non-darks
| |
19:36 | troy_s | It is a leak of some sort.
| |
19:37 | alexML | it would look very similar
| |
19:37 | Bertl_oO | but then I saw the other image somebody visiting sebastian made from the setup
| |
19:37 | troy_s | Bertl_oO: And?
| |
19:37 | troy_s | Bertl_oO: Difference?
| |
19:37 | Bertl_oO | and in this image, there is a bright light source on the left, and a table lamp on the right, just illuminating that corner
| |
19:38 | troy_s | Flare leak.
| |
19:38 | troy_s | Or just a literal light
| |
19:38 | troy_s | I didn't think twice when I first saw it and just assumed it was a glow from a lamp.
| |
19:38 | troy_s | It's not shaped enough to be a flare, but it is also in the top right corner too.
| |
19:38 | troy_s | Slightly.
| |
19:39 | Bertl_oO | so, I would double check with sebastian, that the setup is not flawed here before discussing this effect further
| |
19:39 | troy_s | It has falloff like a lamp.
| |
19:39 | troy_s | I agree.
| |
19:39 | troy_s | I mean as far as I am concerned anyways, it's all moot until the next iteration comes in with proper latitude.
| |
19:39 | troy_s | Not much to be done until then.
| |
19:40 | troy_s | Bertl_oO: Would you be able to get a camera set up such that it could be remotely controlled?
| |
19:40 | troy_s | Just to tweak the PWLC etc?
| |
19:40 | troy_s | (God. Must go get Cintiq Blech.)
| |
19:40 | intracube | troy_s: I forgot to mention... I blurred the images to get rid of HF FPN too
| |
19:40 | intracube | so don't read too much into the shape of it
| |
19:42 | alexML | again - take an image of a blank wall, with any digital camera, overexpose it, leave just one corner a little unclipped, and render the raw file in the same way
| |
19:42 | alexML | upload the result here
| |
19:43 | Bertl_oO | troy_s: I'd say we already have part of that in place
| |
19:44 | Bertl_oO | with a bunch of remote controlled switches to change the lighting, it should probably work
| |
19:44 | Bertl_oO | the Beta itself is already remotely accessible
| |
19:46 | troy_s | Bertl_oO: should put a rotate data channels into the format so folks done erroneously call them colours
| |
19:47 | troy_s | Bertl_oO: with a calendar we could have a nice lab
| |
19:48 | troy_s | Very keen on PWLC
| |
19:49 | troy_s | Would be a terrific way to test calculated knee points
| |
19:49 | alexML | troy_s: we are looking at low-level sensor output, before the matrix gets applied
| |
19:50 | troy_s | (Reverse a log step per bit per stop optimization to knees)
| |
19:50 | troy_s | I know this
| |
19:51 | alexML | I doubt that :P
| |
19:51 | troy_s | Notice how I never say colour when looking at data?
| |
19:51 | alexML | yeah, how should we call it?
| |
19:51 | troy_s | Data
| |
19:51 | intracube | red: http://i.imgur.com/lFH9rOq.png green: http://i.imgur.com/qAhCrX1.png blue: http://i.imgur.com/aun2ttc.png
| |
19:52 | intracube | troy_s: this is the data off the r, g, b sensels though. so you can legitimately call them that
| |
19:52 | intracube | those images contrast all contrast stretched the same amount
| |
19:53 | niemand | joined the channel | |
19:53 | troy_s | Which has no reference so you may equally call them radish, turnip, potato
| |
19:54 | intracube | no reference? the sensels have colour filters, no? :)
| |
19:54 | troy_s | Yes but no reference
| |
19:55 | troy_s | If you rotate those channels the data is equivalent
| |
19:55 | alexML | troy_s: to be able to rotate the channels, you first need linear response (not clipped stuff)
| |
19:55 | alexML | we don't have that here, so we can't rotate yet
| |
19:56 | alexML | we are just looking at the raw data before rotating
| |
19:56 | alexML | that's it
| |
19:58 | alexML | intracube: you have applied white balance to the data
| |
19:58 | alexML | can you show the raw (unmodified) one?
| |
19:58 | alexML | maybe r,g1,g2,b
| |
19:59 | troy_s | Say what?
| |
19:59 | intracube | alexML: this was with Bertl's script. I didn't think it did any white balance
| |
20:00 | intracube | and just debayered
| |
20:00 | troy_s | If you rotate those channels and profile it will deliver guess what
| |
20:02 | intracube | convert \( -size 4096x3072 -depth 12 gray:input_raw_image.raw12 \) \( -clone 0 -crop -1-1 \) \( -clone 0 -crop -1+0 \) \( -clone 0 -crop +0-1 \) -sample 2048x1536 \( -clone 2,3 -average \) -delete 2,3 -swap 0,1 +swap -combine output.png
| |
20:02 | intracube | ^ alexML
| 20:03 | intracube | doesn't fully understand how it works :)
|
20:04 | alexML | intracube: nevermind, can you do the same thing for the other images in the set?
| |
20:05 | alexML | let's say from 1 to 100ms
| |
20:05 | troy_s | Anyways had my fun in the asylum.
| |
20:05 | alexML | in particular, compare the 75ms one with the 88ms one (since I didn't download past that)
| |
20:05 | alexML | compare all 3 (or 4) channels
| |
20:05 | intracube | alexML: I'll have a go later, yep
| |
20:06 | intracube | need to get some food in a bit
| |
20:10 | troy_s | Did you rotate the channels and characterize yet? ;)
| |
20:11 | alexML | troy_s: this is an overexposed image, what do you expect?!
| |
20:12 | troy_s | I expect the exact same results actually.
| |
20:12 | troy_s | Rotated primaries ;)
| |
20:12 | troy_s | 12:55 <alexML> troy_s: to be able to rotate the channels, you first need linear response (not clipped stuff)
| |
20:12 | troy_s | That is crap
| |
20:12 | alexML | proof please
| |
20:13 | alexML | (as in, apply the matrix on a nonlinear image and show that you get correct results)
| |
20:13 | troy_s | Not the damn matrix
| |
20:13 | troy_s | To characterize.
| |
20:13 | troy_s | Hence why I keep repeating they ain't colors.
| |
20:14 | troy_s | Anyways... Fun time is over. This is depressing.
| |
20:34 | darthrak_ | joined the channel | |
20:44 | Bertl_oO | intracube: do you want to know how it works? :)
| |
20:47 | Bertl_oO | if so, please remind me when I'm back from my nap ...
| |
20:47 | Bertl_oO | changed nick to: Bertl_zZ
| |
21:00 | darthrak_ | left the channel | |
21:01 | se6astian | off to bed
| |
21:05 | davidak | left the channel | |
21:06 | se6astian | changed nick to: se6astian|away
| |
21:19 | davidak | joined the channel | |
21:27 | intracube | left the channel | |
21:32 | davidak | left the channel | |
21:42 | niemand | left the channel | |
22:00 | darthrak_ | joined the channel | |
22:05 | darthrak_ | left the channel | |
22:37 | SL2 | joined the channel |