05-12-2012, 07:18 PM
[quote name='popo' timestamp='1336840145' post='18150']
Two interesting points there, since I do try bandwidth reduced imaging with colour sensors, and it is quite annoying.
The possible highlight recovery from using information from separate channels is possible in theory, but what software that handles that well? I was wondering about using that as like a single shot "HDR" sensor since the colour channels are each differently sensitive.[/quote]
Well enough for me. The nice thing is, that one can indeed treat the different channels independently. Personally, I use PS plus a bunch of add-ons, set my own black point, etc. I am quite happy with the results, as they look like the results I got in the past from my own approach to the Zone System, but with significantly more detail <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/biggrin.gif' class='bbc_emoticon' alt='
' />.
I reckon that it doesn't really work like 2X linear resolution, based on the photographs I have seen so far. I think 50% more is being generous. Don't forget that there still is diffraction, and that just sets in earlier, relatively speaking, with higher medium (sensor) resolution, and thereby still limiting the resolution at an exponential rather than linear level. A camera like the Canon 5D II manages maximum theoretical resolution minus 5 % to 8 % due to the AA-filter, be it potentially looking slightly softer due to interpolation. However, how much is that really going to be? There is no additional interpolation for a monochrome sensor, it is what it is. Furthermore, amazing things are possible with the right demosaicing and sharpening algorithms for RGB. So, I doubt whether the M-Monochrome has really got a 2X advantage. Its advantage potentially lies in a better monochrome tonal scale, provided it is processed well enough. IOW, colour prints vs properly processed Zone System type prints. And that comes, especially with a digital sensor, with its own peculiarities, especially as the DR of the sensor is what it is - with film we can influence the tonal scale and tone mapping in the negative already: there are no such controls, other than iso (which is very limited in this regard) to do so with a sensor, unless someone implements such an option in hardware, a curve steepness control type of thing. Don't forget that B&W film in principle has about the same DR, or even more, as a good sensor. IOW, we can control the entire DR scale of a subject with film, whereas we can only work within the possibilities of the sensor DR for digital, unless we do proper HDR type photography. The latter will only work for what are essentially static subjects. The Zone System approach allows for compression of, e.g., 20 stops of DR to e.g. 10 into a single image in one go, where a sensor "only" has a DR of 10, or maybe 12 stops in RAW. With a sensor we cannot simply capture those 20 stops.
<img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/biggrin.gif' class='bbc_emoticon' alt='' />
From a monetary POV, as a hobbyist, not mine either <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/biggrin.gif' class='bbc_emoticon' alt='' />.
Thinking about all this, my conclusion really is more and more going towards sticking with what I have already - no need for a separate monochrome body <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/biggrin.gif' class='bbc_emoticon' alt='' />.
Kind regards, Wim
Two interesting points there, since I do try bandwidth reduced imaging with colour sensors, and it is quite annoying.
The possible highlight recovery from using information from separate channels is possible in theory, but what software that handles that well? I was wondering about using that as like a single shot "HDR" sensor since the colour channels are each differently sensitive.[/quote]
Well enough for me. The nice thing is, that one can indeed treat the different channels independently. Personally, I use PS plus a bunch of add-ons, set my own black point, etc. I am quite happy with the results, as they look like the results I got in the past from my own approach to the Zone System, but with significantly more detail <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/biggrin.gif' class='bbc_emoticon' alt='
' />.Quote:The resolution increase is what I'd really be interested in from any mono sensor. If you move away from the green channel, kiss goodbye to resolution. A mono sensor would offer 2x linear resolution, or 4x area resolution in that case. No more is your 18MP sensor working as a sparsely sampled 4.5MP one.
I reckon that it doesn't really work like 2X linear resolution, based on the photographs I have seen so far. I think 50% more is being generous. Don't forget that there still is diffraction, and that just sets in earlier, relatively speaking, with higher medium (sensor) resolution, and thereby still limiting the resolution at an exponential rather than linear level. A camera like the Canon 5D II manages maximum theoretical resolution minus 5 % to 8 % due to the AA-filter, be it potentially looking slightly softer due to interpolation. However, how much is that really going to be? There is no additional interpolation for a monochrome sensor, it is what it is. Furthermore, amazing things are possible with the right demosaicing and sharpening algorithms for RGB. So, I doubt whether the M-Monochrome has really got a 2X advantage. Its advantage potentially lies in a better monochrome tonal scale, provided it is processed well enough. IOW, colour prints vs properly processed Zone System type prints. And that comes, especially with a digital sensor, with its own peculiarities, especially as the DR of the sensor is what it is - with film we can influence the tonal scale and tone mapping in the negative already: there are no such controls, other than iso (which is very limited in this regard) to do so with a sensor, unless someone implements such an option in hardware, a curve steepness control type of thing. Don't forget that B&W film in principle has about the same DR, or even more, as a good sensor. IOW, we can control the entire DR scale of a subject with film, whereas we can only work within the possibilities of the sensor DR for digital, unless we do proper HDR type photography. The latter will only work for what are essentially static subjects. The Zone System approach allows for compression of, e.g., 20 stops of DR to e.g. 10 into a single image in one go, where a sensor "only" has a DR of 10, or maybe 12 stops in RAW. With a sensor we cannot simply capture those 20 stops.
Quote:And as a very un-Leica thought, narrowband imaging potentially allows the use of lower quality optics, since CA effects are reduced.Only seemingly so. They show up in B&W rather than in full colour glory, hence are less noticeable. The big winner is really noise, not CA. Noise is really less disturbing in B&W, especially as luminosity pattern noise starts to look like what we see with human vision in very dark or very high contrast circumstances.
Quote:So I guess my point from all this is I wish someone did it a bit cheaper? Leica wouldn't be my first choice...
<img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/biggrin.gif' class='bbc_emoticon' alt='
' />From a monetary POV, as a hobbyist, not mine either <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/biggrin.gif' class='bbc_emoticon' alt='
' />.Thinking about all this, my conclusion really is more and more going towards sticking with what I have already - no need for a separate monochrome body <img src='http://forum.photozone.de/public/style_emoticons/<#EMO_DIR#>/biggrin.gif' class='bbc_emoticon' alt='
' />.Kind regards, Wim
Gear: Canon EOS R with 3 primes and 2 zooms, 4 EF-R adapters, Canon EOS 5 (analog), 9 Canon EF primes, a lone Canon EF zoom, 2 extenders, 2 converters, tubes; Olympus OM-D 1 Mk II & Pen F with 12 primes, 6 zooms, and 3 Metabones EF-MFT adapters ....