brianv Posted June 8, 2012 Share #301 Posted June 8, 2012 Advertisement (gone after registration) Do you use a full-spectrum digital camera? I have for over 20 years. I like to use various Wratten series filters, a nice one is a magenta filter. Basically removing the IR cut filter from a digital camera makes it like Infrared Ektachrome. I used that in the 1970s. Google "Infrared Ektachrome". Removing the IR filter from a camera that has a Bayer pattern layer will just make the camera almost useless for ordinary photography because the firmware will still try to interpret neighboring pixels, and if you do use the near-infrared spectrum you must be prepared for some pretty ugly outcomes, especially if a bodily unattractive woman in a sun dress walks into the scene on a bright sunny day. Google these words: IR photography x-ray Link to post Share on other sites More sharing options...
Advertisement Posted June 8, 2012 Posted June 8, 2012 Hi brianv, Take a look here Survey: Your opinion about the new LEICA M MONOCHROM. I'm sure you'll find what you were looking for!
wattsy Posted June 8, 2012 Share #302 Posted June 8, 2012 Another camera talked to death before it's even out. That's a bit rich. Link to post Share on other sites More sharing options...
mwilliamsphotography Posted June 9, 2012 Share #303 Posted June 9, 2012 That's a bit rich. Sorry, I don't know what that means. -Marc Link to post Share on other sites More sharing options...
mwilliamsphotography Posted June 9, 2012 Share #304 Posted June 9, 2012 Oh wait, I get it now ... because I'm talking about it also. I meant discounting the MM B&W performance before they even finish tweaking the camera up to launch date. Sorry, I should have been clearer. -Marc Link to post Share on other sites More sharing options...
brianv Posted June 9, 2012 Share #305 Posted June 9, 2012 Every review that I've read by someone that has actually used the camera has been positive. Reading forums, quite a few people have pre-ordered one. My Paypal is past 60% for ordering one. Link to post Share on other sites More sharing options...
AlanG Posted June 9, 2012 Share #306 Posted June 9, 2012 Oh wait, I get it now ... because I'm talking about it also. I meant discounting the MM B&W performance before they even finish tweaking the camera up to launch date. Sorry, I should have been clearer. -Marc I'm not following you. Why would the performance be anything but superb? That ability certainly has been demonstrated. While some posts are off the mark, I think what is germane are the reason for having one vs. using a color camera and converting or shooting b/w film. Clearly each method has its pluses and minuses and that is what is being brought up. I think these factors will probably be considered even by those who can easily afford it. My posts about the removal of the color filters mostly has to do with considering the possibility that the M9 might be boosting its signal to get 160 ISO from the sensor thus helping to explain why the MM at a base ISO of 320 has less noise. Link to post Share on other sites More sharing options...
mjh Posted June 9, 2012 Share #307 Posted June 9, 2012 Advertisement (gone after registration) considering the possibility that the M9 might be boosting its signal to get 160 ISO from the sensor The base sensitivity is the ISO setting maximizing dynamic range and for the M9 that is ISO 160. So there’s no boost. Link to post Share on other sites More sharing options...
AlanG Posted June 9, 2012 Share #308 Posted June 9, 2012 The base sensitivity is the ISO setting maximizing dynamic range and for the M9 that is ISO 160. So there’s no boost. We seem to be going around in circles here. It is my contention that the Bayer filter will cut out more than one stop if everything else stays the same. You don't think so. Therefore I think there is some boost and you don't. I do not have any information about the neutral density of these filters and would like to know that for sure. This man posted tests with and without a Bayer filter and figures about a 2 stop difference. (Assuming his procedure was accurate.) He had two versions of the same sensor. http://forums.dpreview.com/forums/readflat.asp?forum=1018&message=36488951&changemode=1 And consider that this talks about a 1 to 2 stop gain simply by leaving half the pixels clear not all of them in a full color array not a monochrome one. So it seems to me that the colored pixels must be boosted to match the light in the clear ones hence the range of "between two and four times." "The TRUESENSE Color Filter Pattern represents a new paradigm for obtaining full-color images from a single image sensor. By adding panchromatic (or “clear”) pixels to the standard red, green, and blue pixels traditionally used for color capture, image sensors can realize a significant increase in light sensitivity – between two and four times compared to standard designs – while retaining overall image quality and color fidelity. " http://www.truesenseimaging.com/technologies/truesense-sparse-color-filter-pattern Link to post Share on other sites More sharing options...
Pedro Posted June 9, 2012 Share #309 Posted June 9, 2012 (...) I need a camera that produces. I would love to have a Leica digital with the quality and USEABLE low light sensor range of the two Nikon D3s cameras I currently use for weddings. This ain't it. (...) Agree. Link to post Share on other sites More sharing options...
mjh Posted June 9, 2012 Share #310 Posted June 9, 2012 We seem to be going around in circles here. It is my contention that the Bayer filter will cut out more than one stop if everything else stays the same. And I wonder why you insist on this. It’s not like there was any evidence in favour of your position. This man posted tests with and without a Bayer filter and figures about a 2 stop difference. (Assuming his procedure was accurate.) He had two versions of the same sensor. Not the sensor in the M9 or M Monochrom, obviously. The methodology wasn’t particlarly accurate anyway. If there was a two stop difference between the M9 and the M Monochrom then either the base sensitivity of the M9 is actually ISO 80 (which it isn’t since we get the maximum dynamic range at ISO 160) or the base sensitivity of the M Monochrom is ISO 640 (which it isn’t since its maximum dynamic range is achieved at ISO 320). You are entitled to believe otherwise but then you are at odds with the facts. Link to post Share on other sites More sharing options...
AlanG Posted June 9, 2012 Share #311 Posted June 9, 2012 If there was a two stop difference between the M9 and the M Monochrom then either the base sensitivity of the M9 is actually ISO 80... Yes, that is the assumption. But we can't shoot at ISO 80 with an M9 to see how that would look. Besides the ISO value is only computed after doing a lot of image processing to equalize the light through each color filter and then applying a tone curve. A lot can happen to noise and dynamic range in all of this. Of course another possibility is that the MM uses a slight magenta filter to cut down the green response. Thus lowering that camera's overall sensitivity to get that 320 ISO. Or some combination of factors. And I agree that I do not have the data to know what is going on as it is a complex system that we can't break down into individual components. But I do know something about filters so I wonder, none the less, until someone can show measurements of the filters themselves. Truesense does claim a 1-2 stop improvement just by removing filters over half the pixels. Does that not mean anything about their light absorbing qualities? So where does the M9's 1 stop filter array come from? I can't see how one can shortcut this. In my world, unless you are making instant grits, they take 25-40 minutes to cook. Please keep in mind that I am going through a thought process here. I'm not trying to argue with you or convince you of anything. Link to post Share on other sites More sharing options...
jaapv Posted June 10, 2012 Share #312 Posted June 10, 2012 :confused: The M9 has an ISO 80 pull setting because the base ISO is 160....So you can shoot it at ISO 80. And lose dynamic range. Link to post Share on other sites More sharing options...
AlanG Posted June 10, 2012 Share #313 Posted June 10, 2012 :confused: The M9 has an ISO 80 pull setting because the base ISO is 160....So you can shoot it at ISO 80. And lose dynamic range. Sorry, I am simply trying to get at that something else is going on and I know this is a very wonky point. There is a conundrum here because all of the information is not available and clear. As I said, maybe the base ISO of the sensor is actually higher than 320 and they cut that down with a filter on the MM to smooth out the color response. (This wouldn't be needed if there are color sep filters over the sensor.) Maybe the IR filtering is also a bit stronger and adds more neutral density. It is hard to know if there is a combination of processing, filters, or other factors. But let's simply concentrate on the idea of filtering light. I come to this as a very simple observation about light... leaving the sensor aside. (That I agree is not that big a deal in the greater scheme of things but I don't see any explanation for.) In order for color separation filters to only lose 1 stop, each color would have to be able to take in half of the "white" light in the scene and have no neutral density at all. (I am not aware of any kind of filters that are this pure.) These would not be red, green, and blue filters either. (More like orange, greenish and cyan.) Standard Kodak color separation filters that are used with film absorb approximately 3 stops of light. So it seems very obvious to me... even if each filter somehow can miraculously pass half the light in the visible spectrum with absolutely no neutral density, how can orange, pale green, and cyan filters work as color separation filters? And from looking at other Truesense filter information, that is not how filters work on those sensors. Link to post Share on other sites More sharing options...
pico Posted June 10, 2012 Share #314 Posted June 10, 2012 As I said, maybe the base ISO of the sensor is actually higher than 320 and they cut that down with a filter on the MM to smooth out the color response. (This wouldn't be needed if there are color sep filters over the sensor. Maybe the IR filtering is also a bit stronger and adds more neutral density. It is hard to know if there is a combination of processing, filters, or other factors. But let's simply concentrate on the idea of filtering light. Therein is the heart of our problem, Alan. We do not know (at least I do not know) what IR filtering technology they are using. There are at least three types used over digital sensors that I know of. They are very different, however their goal is to have insignificant optical density (that's the term, not 'neutral density'). So it seems very obvious to me... even if each filter somehow can miraculously pass half the light in the visible spectrum with absolutely no neutral density, how can orange, pale green, and cyan filters work as color separation filters? And from looking at other Truesense filter information, that is not how filters work on those sensors. I would not be concerned with orange and cyan filters, nor RGB filters. Consider other filtering technology such as 'cut' filtering, and of vertical and horizontal low-pass filtering. Link to post Share on other sites More sharing options...
AlanG Posted June 11, 2012 Share #315 Posted June 11, 2012 I would not be concerned with orange and cyan filters, nor RGB filters. Consider other filtering technology such as 'cut' filtering, and of vertical and horizontal low-pass filtering. Except we know that this sensor uses standard RGB dye based filters and nothing so esoteric. I may be wrong using the term neutral density but it is easy to understand and should apply for dye based filters. I was taught that these kinds of filters are impure and thus have a neutral density component as well as their filtering properties. Regarding IR filtering. It is my understanding that most sensors use dichroic filters that reflect unwanted IR but that this wouldn't work with the M9 due to the requirement of the M's shorter lenses. So they used a dye based IR filter that has a side effect of absorbing more white light than a dichroic IR filter would lose via its "optical density." So maybe the MM needs a stronger IR filter than is used in the M9 as it does not get any benefit of IR filtering from the Bayer filter. All I'm saying is that when I first heard that the MM sensor was the same as the M9's but only gained 1 stop of light from the removal of the Bayer filter... I immediately thought that something is wrong in that statement. I'm not sure what is the cause and can only guess at it. This is just a thought exercise for me and I don't think it is significant in application and is not a knock on either camera. Link to post Share on other sites More sharing options...
pop Posted June 11, 2012 Share #316 Posted June 11, 2012 The arguments runs as if the M9 had filters in front of the sensor while the MM did not. Further, if such was the case, the density of the filter in the M9 is deemed too low at 1EV. However, we can assume that the spectral sensitivity of a bare sensor does not lend itself for an adequate rendering of colours in a monochrome photograph. Hence, I presume that the 1EV difference in sensitivity is the difference of the densities of the filters in those cameras. Link to post Share on other sites More sharing options...
wilfredo Posted June 11, 2012 Share #317 Posted June 11, 2012 Yadi Yadi Yada... Link to post Share on other sites More sharing options...
brianv Posted June 11, 2012 Share #318 Posted June 11, 2012 Regarding IR filtering. It is my understanding that most sensors use dichroic filters that reflect unwanted IR but that this wouldn't work with the M9 due to the requirement of the M's shorter lenses. That was true 20 years ago, up through the Kodak DCS400 series. Most digital cameras use IR absorbing glass. I removed it from a Nikon Coolpix 950 and replaced with clear. It is about the same color as that used in the M8 and M9. It is thicker. Low-end cameras, like the cheap digicams and cell phone cameras use dichroic filters. Easy to tell: look into the camera and you see the reddish reflections. As far as the Bayer pattern mosaic filters: you can read the KAF-10500 long sheet for the effect that the RGB dyes have on the quantum efficiency. Overall, they knock it to about 1/2 of the QE of a detector with IR absorbing glass over it. The Red-Green-Blue pixels have a lot of spectral overlap. They have long tails, and not sharp cutoffs as do the color filters used for black and white photography. I've owned monochrome and color versions of the DCS200. The answer is: the RGB bayer Pattern Mosaic Filter cuts the ISO in half. It is 1 F-Stop. Nothing else is done in the camera, it did not do any processing- just dumped the raw image including calibration rows. They look like Sprockets if you do your own raw processor and leave them in the image. This version of the detector is newer, would love to see the long sheet on it. If Kodak/Truesense improved the Dark Signal even further- it is possible that it improved the SN ratio. That is evident when comparing the M8 to the M9: Kodak improved the dark current performance of the KAF-18500 over the KAF-10500. reading other long sheets from sensors of the same generation as the KAF-18500, Kodak improved it enough to account for 1 full stop of improvement. IF Truesense managed to improve it further- that would make the ISO 10,000 on the M9 Monochrome better than just removing the RGB filter. Link to post Share on other sites More sharing options...
AlanG Posted June 11, 2012 Share #319 Posted June 11, 2012 The Red-Green-Blue pixels have a lot of spectral overlap. They have long tails, and not sharp cutoffs as do the color filters used for black and white photography. So the filters would have to be so unsaturated that they only absorb half the light and at minimum they would allow half of the visible spectrum through. (Of course these two aspects are probably mutually exclusive too in that I don't see how you can block only 1/2 the light and control the colors blocked very well.) Do you think satisfactory color separation could be accomplished with such filters? That is what it comes down to and it seems impossible to me. But I'm open to proof. Link to post Share on other sites More sharing options...
jaapv Posted June 11, 2012 Share #320 Posted June 11, 2012 It only needs enough separation to allow interpolation. Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.