Bad Pixels: Unusual Behaviour

SkepticaLee

50

Re: Bad Pixels: Unusual Behaviour

« Reply #30 on: 22 / October / 2013, 05:07:16 »

Advertisements

One last word on dead pixels (the pixels in badpixel.bin):

As I document at http://www.skeptic.de/CHDK/Bad%20Pixels/index.htm#Index_5, as many as 3 different badpixel.bin files can be produced by any one camera, depending on the Bv and ISO. The composition of these files is relatively stable over time, but the exact conditions under which any particular set of dead pixels is detected is difficult to nail down.

Because of this I have formulated the following recommendation for generating and using badpixel.bin:

Quote

Therefore I recommend applying the Precautionary Principle, and using the largest possible dead pixel set that can be generated on an individual camera. In relative terms, the number of dead pixels is typically < 0.1% at a maximum, which should hardly affect picture quality at normal resolutions. In this way, the worst case is covered.

To generate such a dead pixel file, call the Create badpixel.bin function at the highest ISO rating possible for a shot in the low-to-medium light range.

For the rationale behind this, please read the document in the link given above.

Logged

reyalp

14082

Re: Bad Pixels: Unusual Behaviour

« Reply #31 on: 22 / October / 2013, 13:23:08 »

Quote from: SkepticaLee on 22 / October / 2013, 04:57:42

Since the JPGs that the camera produces do not seem to have any of the hot pixels that are quite visible on the DNGs, my question now is: Where does the camera store this information, and can it be retrieved?

In my experience, this is not correct. "Hot" pixels show up in the jpeg. Bad pixels known to the camera show up in the DNG as a saturated pixel, because one color of the bayer pattern is 0. These do not show up in the jpeg.

On my cameras, there are relatively few true "hot" pixels. On my D10, a 1 minute dark frame gave me ~70 pixels near the max possible value.

It's possible that this varies on some cameras or sensor types I guess, but I suspect you still getting tripped up by one bayer element being dead.

The bug Phil fixed in http://chdk.setepontos.com/index.php?topic=10828.0 may have added confusion to your analysis.

edit:
In http://www.skeptic.de/CHDK/Bad%20Pixels/Awaroa_4.png the pixels are clearly "dead" not "hot"

« Last Edit: 22 / October / 2013, 13:24:42 by reyalp »

Logged

Don't forget what the H stands for.

SkepticaLee

50

Re: Bad Pixels: Unusual Behaviour

« Reply #32 on: 22 / October / 2013, 15:50:47 »

Quote from: reyalp on 22 / October / 2013, 13:23:08

In http://www.skeptic.de/CHDK/Bad%20Pixels/Awaroa_4.png the pixels are clearly "dead" not "hot"

Really? I don't care whether they are dead or hot, but they certainly don't show up as bad pixels in badpixel.bin. And they are not dead all the time, so perhaps we can call them "Jesus pixels", or just "Zombie pixels".

What kind of pixels does badpixel.bin contain? I was under the impression, that it contained pixels where R=G=B=0 under all lighting conditions. If we don't clarify such simple matters as definitions, then we will just be going round and round the mulberry bush.

Whatever is in the badpixel.bin file is not useful to me when trying to fix the multi-coloured pixels that turn up in both over- and underexposed images, but do not turn up in the corresponding JPGs. Just compare http://www.skeptic/chdk/bad%20pixels/Totaranui_Pan_Left_Max.png (the DNG image) to http://www.skeptic/chdk/bad%20pixels/Totaranui_Pan_Left_Max_JPG.png. Same selection, same image, and yet the JPG shows none of the problematical pixels. How come?

Logged

reyalp

14082

Re: Bad Pixels: Unusual Behaviour

« Reply #33 on: 22 / October / 2013, 16:19:10 »

Quote from: SkepticaLee on 22 / October / 2013, 15:50:47

What kind of pixels does badpixel.bin contain?

Dead (or rather by the canon firmware to zero) pixels.

Quote

I was under the impression, that it contained pixels where R=G=B=0 under all lighting conditions.

NO. The sensor has a bayer filter. Pixels on the sensor are individual bayer elements. One of R, G or B.

The Canon badixel list (and thus badpixel.bin) operates on individual sensor elements, since that's what goes bad.

When only one of these is zero, the resulting pixel in the de-bayered image contains the colors of the neighboring pixels, usually resulting in a yellow, cyan or magenta pixel (but sometimes you have several next to each other, which changes things. This seems common on newer cameras). This is exactly what we see in the image I linked.

Logged

Don't forget what the H stands for.