Leica M11 Monochrom: Disturbing Number of Hot Pixels / Stuck Pixels and No Way to Map Them Out
Related: black and white, dark frame, digital sensor, exposure, fairy dust aka hot/stuck pixels, Leica, Leica M, Leica M11 Monochrom, long exposure, long exposure noise reduction, noise, pixel mapping, rangefinder
re: monochrome
re: hot pixels
re: long exposure noise reduction and dark frame
Shooting in the evening, I discovered “fairy dust” on some images—single bright pixels scattered randomly about*.
Looking closer, it was clear that these pixels did not change between exposures—“hot pixels” or “stuck pixels”. There are hundreds of them. They are popping up at realtively short exposure times(2 seconds, 4 seconds) but at base ISO 125.
The Leica M11 Monochrom lacks a pixel mapping function, and it does not invoke dark frame subtraction until 160 seconds (at ISO 125), so AFAIK, there is no way to map-out these pixels short of grossly tedious effort every image, which is ridiculous.
Yet the Leica Q2 Monochrom has a pixel mapping function as of two years ago. Maybe Leica will get around to it with the M11M, but in the meantime, every image taken potentially suffers.
Brand new camera and I was not trying to find imaging defects. Doh! The M11M appears to have fewer “pimple” defects, but with these stuck pixels, call me unimpressed as to Leica’s seriousness about image quality.
...
Roy P writes:
Wow, there are a lot of them, and all over the place. Got tired of marking them up, so I stopped after some obvious ones. Maybe you can hold a contest for people to find all the hot pixels!
Seriously, this seems excessive. I don’t know what the QA/QC threshold is to flag a sensor as defective and reject it, but it doesn’t feel right to ship a camera with a sensor like this. At some print size, maybe 16x24 or even smaller, they will likely start showing.
DIGLLOYD: if one examines the image on a 220 ppi Retina display, many of these are likely to go unnoticed barring exceptional close-range vision. On the 110 ppi NEC PA302W, they are easier to spot. But yes, with large prints you are f*cked.
I expect exceptional image quality from a monochrome sensor. Its only raison d^etre is image quality above and beyond what a color sensor can capture—because a color sensor can be converted to B&W with far more control and far better tonal mapping in 'post'. So the damned sensor had better offer something really great. So while overall image quality is clearly a step up in many senses with the the M11M, this sort of imaging defect is a bad joke IMO. And... there are also pimple defects.
I suspect it is not this particular sensor per se; all have similar issues. Indeed, I saw this with the Fujifilm GFX100S right away when I first tested it at 60 seconds (same core technology genetics). The difference is that Leica (a) seemingly did not map them out before shipping, (b) offers no facility for the user to do so, and (c) restricts dark frame subtraction for long exposure noise reduction to 160 seconds or longer at ISO 125.
It is unacceptable that Leica has neither a dark frame option nor a pixel mapping feature to eliminate this image defect. Meanwhile, each and every image we take is contaminated with this imaging defect.
More hot pixels are here than are circled... they abound
Crop is only 0.232 megapixels eg 1 part in 258 of the entire image..
f2.8 @ 60.0 sec, ISO 125; 2023-04-27 20:14:37
LEICA M11 Monochrom + Zeiss ZM 35mm f/1.4 Distagon + filter B+W KR-12 RAW: +10 Whites, +10 Clarity
Full-size image below can be found in the examples.
-
LEICA M11 Monochrom + Zeiss ZM 35mm f/1.4 Distagon + filter B+W KR-12 RAW: +10 Whites, +10 Clarity
[low-res image for bot]
Fairy dust exists at 2s or 4s as well (and maybe faster speeds).
-
LEICA M11 Monochrom + Zeiss ZM 35mm f/1.4 Distagon + filter B+W KR-12 RAW: +10 Whites, +10 Clarity
[low-res image for bot]
