Please put your stories through a spell checker and include a custom dictionary if need be. The last two stories have been good examplws of why this is a good idea.
Among these is that since the luminance information is only being sampled from the green cells, there is actually less resolution available than one might think. About 2/3rds in fact
there is no possibility of either colour moire or related artifacting.Moire is a form of aliasing and aliasing significant on this camera - it is visually more noticeable than in the regular M9 due to no color information fuzzying it to the brain.
Engineering a monochrome sensor equipped camera isn't simply a matter of removing the Bayer array. Though based on the M9 sensor, a significant amount of reengineering at the chip level was required.This is utter nonsense. There is zero engineering required for the sensor to enable removal of the Bayer CFA. I am sure the Leica marketing disagrees, but it really is as simply as not installing it in the first place.
niche product within a nicheThis is very true.
I think this is the wrong way to face the discussion.Quotesince the luminance information is only being sampled from the green cells...With all due respect, all I can say is that if I have to choose between the technical information I get from the scientists and engineers working at the top digital imaging companies in Europe, Japan, and America, or you, somehow I think you'll end up being the loser.
Y = 0.2126 R + 0.7152 G + 0.0722 B(Only offered as a rough guide as to how the standard red/green/blue channels contribute to luminance - any given camera will typically have somewhat different response)
Graeme,
Thanks for your professional input. Your figure of the CFA reducing resolution by about 20%, plus the exta loss due to the AA filter that the MM also avoids, matches up very well with Michael's eye-ball estimate of 26-32MP: just removing that factor of 80% linear resolution would be comparable to increasing linear pixel count by 100/80, and so MP count going from 18 to 18*(100/80)^2 = 28MP.
That 80% is only for the Green filtered sensels.
Where did the straw man argument that green is only responsible for luminance come from? That's not what I wrote or have ever written. They are predominantly responsible in a Bayer array, but not totally.
It would be interesting to compare monochrome images processed from the Nikon D800E to those from the Leica MM.
Aku,
Thank you for your comments. With all due respect, all I can say is that if I have to choose between the technical information I get from the scientists and engineers working at the top digital imaging companies in Europe, Japan, and America, or you, somehow I think you'll end up being the loser.
Engineering a monochrome sensor equipped camera isn't simply a matter of removing the Bayer array. Though based on the M9 sensor, a significant amount of reengineering at the chip level was required
Foveon technology, where the colour filter layers are stacked verticallyYour lack of knowledge is clear here as well - Foveon does not have any color filter layers. Instead it has a triple photodiode construction where modest color separation is achieved due photons of different energies penetrating silicon to different depths with different probabilities.
Or if you do not trust theory, just look at the resolution measurement for the X3 style pixels of Foveon/Sigma sensors, which roughly match a Bayer CFA with about twice the total pixel count
In particular, the Kodak sensors for M cameras have a special design with microlenses offset towards the optical axis
elements amd exit pupil very close to the focal plane (no SLR lens does this). It could well be that even if a sensor supplier like Sony could do this, none except Kodak and Dalsa have developed suitable microlens technologies
I think this thing is a mistake, if only because it should have been a lot less expensive than the cooking version of M9.
Also with all due respect, you may well talk to all kinds of experts, but that does not mean that you are an expert in this field (digital imaging, including sensor operation). You are not. I just wish you could understand it and write your articles accordingly. Now they are filled with mistakes and no offense meant, you're quite arrogant in defending your writing by ignoring the criticism and praising your connections. This behavior is not too dissimilar to so called "fanboy" behavior which pollutes discussions.
Let me quote you again:
This is absolutely false. If you knew what you are talking about, instead of just repeating what someone at Leica has told you, you'd know that. Of course, I would be delighted if you would tell of one single thing that would need to be reengineered (at the "chip" level - I assume you mean the sensor, but maybe something else).
And something else from your article:Your lack of knowledge is clear here as well - Foveon does not have any color filter layers. Instead it has a triple photodiode construction where modest color separation is achieved due photons of different energies penetrating silicon to different depths with different probabilities.
And when it comes to resolution, BartvanderWolf has demonstrated in this forum that your estimates are quite false. I have also done my share of measurements and the results have been significantly different than your guestimates. I prefer measurements and science anytime over eyeballing.
Anyhow, to me this new Leica is kind of like Pentax K-01 - a minimum risk new product with minum amount of reengineering required to make it. Pentax took a K-5, crippled it and put a new body around it and that's K-01, Leica took M9, used a CFA-less sensor and that's the new camera.
The camera is what it is. If you can afford it and like the way it works, go buy one.
If not, again, does it matter?
I do not understand why one person would choose to challenge the statements of an easily identifiable person while choosing to hide anonymously behind a fake name. What do you have to hide? Or, are you afraid of something? You would gain a lot of credibility by being brave and come out of hiding. Be brave young duck! 8)
... that does not mean that you are an expert in this field (digital imaging, including sensor operation). You are not.
I am here in Berlin with some of the leading journalists from the photographic industry and top engineering people from several of the major European digital imaging firms (not just Leica. Many of them read this site.Because you're a reasonable guy, not an authoritarian.
Several have come up to me within the past 12 hours and asked – why do you put up with that crap – referring to this thread.
Why should we trust a theory if it confirms your false assumption of Foveon resolving power? No sensor resolves beyond Nyquist, and Bayer CFA equipped sensors resolve over 90% of Nyquist.Aku, relax a bit; we are not in violent disagreement. My 2:1 pixel count ratio was a rough rule of thumb, not a precise measurement: it would be absurd to propose a single precise number for all CFA cameras with their varying demosaicing algorithms and OLP filters. In terms of linear resolution, my 2:1 corresponds to the Bayer CFA approach having about 70% of the linear resolution or a non CFA sensor. I am happy to refine that 70% up to about the 80% figure suggested by Graeme Nattress (whose job at RED includes things like designing demosaicing algorithms, or so I have been told). And as I noted elsewhere, that 80% figure reduces my 2:1 rule of thumb to about 1.6:1. Even express in the more reasonable unit of linear resolution, the gain is 25%, which seems worthwhile.
This camera would be very interesting with a color wheel.I have a dream: astronomers and some other scientists also have uses for monochrome sensors, often used with various filters, and indeed most models of sensor in the Kodak product list are monochrome. So might Canon or Nikon or Sony or Panasonic or (microscope maker) Olympus produce a "scientific/astronomical camera" with a monochrome CMOS sensor and Live View? This could also be of interest for monochrome artistic photography. (Though I personally would sacrifice a bit of resolution for the option of changing the filtering after the fact.)
But note: when I talk about resolution, I am not referring to the extinction level at which things are barely distinguished (and beloved of film zealots trying to prove the inferiority of digital to film) or measures with high contrast black-and-white test targets. I am instead talking about useful levels of retained contrast, like a MTF of 50% or better, and by that standard, 90% of Nyquist seems over-optimistic.
Quote from: dreedPlease put your stories through a spell checker and include a custom dictionary if need be. The last two stories have been good examplws of why this is a good idea.
I couldn't agree more.
Michael
Are B/W images from the M Monocrom significantly "better" or just different in a pleasing way - Than the same scene and subjects taken with an M9 and the same lens under the same light, with the image then converted to B/W from the colour file using the 'V' key in Lightroom? (no after-the-fact colour filtration allowed).
Also with all due respect, you may well talk to all kinds of experts, but that does not mean that you are an expert in this field (digital imaging, including sensor operation). You are not. I just wish you could understand it and write your articles accordingly. Now they are filled with mistakes and no offense meant, you're quite arrogant in defending your writing by ignoring the criticism and praising your connections. This behavior is not too dissimilar to so called "fanboy" behavior which pollutes discussions.
Let me quote you again:
This is absolutely false. If you knew what you are talking about, instead of just repeating what someone at Leica has told you, you'd know that. Of course, I would be delighted if you would tell of one single thing that would need to be reengineered (at the "chip" level - I assume you mean the sensor, but maybe something else).
And something else from your article:Your lack of knowledge is clear here as well - Foveon does not have any color filter layers. Instead it has a triple photodiode construction where modest color separation is achieved due photons of different energies penetrating silicon to different depths with different probabilities.
And when it comes to resolution, BartvanderWolf has demonstrated in this forum that your estimates are quite false. I have also done my share of measurements and the results have been significantly different than your guestimates. I prefer measurements and science anytime over eyeballing.
Anyhow, to me this new Leica is kind of like Pentax K-01 - a minimum risk new product with minum amount of reengineering required to make it. Pentax took a K-5, crippled it and put a new body around it and that's K-01, Leica took M9, used a CFA-less sensor and that's the new camera.
Aku, relax a bit; we are not in violent disagreement. My 2:1 pixel count ratio was a rough rule of thumb, not a precise measurement: it would be absurd to propose a single precise number for all CFA cameras with their varying demosaicing algorithms and OLP filters. In terms of linear resolution, my 2:1 corresponds to the Bayer CFA approach having about 70% of the linear resolution or a non CFA sensor. I am happy to refine that 70% up to about the 80% figure suggested by Graeme Nattress (whose job at RED includes things like designing demosaicing algorithms, or so I have been told). And as I noted elsewhere, that 80% figure reduces my 2:1 rule of thumb to about 1.6:1. Even express in the more reasonable unit of linear resolution, the gain is 25%, which seems worthwhile.
I have a dream: astronomers and some other scientists also have uses for monochrome sensors, often used with various filters, and indeed most models of sensor in the Kodak product list are monochrome.
This again means that the true resolution of the sensor in the monochrom(e) Leica is 18*3/2=27MP.Agreed on the numbers as an adequate indication of sensor resolution, but I would avoid the often abused adjective "true". What we have here is yet another rough equivalency that I am sure will be much debated: the "Bayer CFA sensor equivalent pixel count". Add this to the "36x24mm format equivalent focal length" (as a FOV measure) and the even more controversial "36x24mm format equivalent aperture" (as a measure of DOF, and maybe of light gathering speed).
To me the 80% of Nyquist would mean that 80% of the linear resolution of the sensor is the true resolution. This means that in terms of megapixels 80%*80%=64% of the sensors megapixels correspond to the true resolution of the sensor.
So what is the true resolution of a Bayer filter camera compared to a non Bayer filter camera? ;)
So what is the true resolution of a Bayer filter camera compared to a non Bayer filter camera? ;)Taking numbers from posts in this forum, the answer is that the use of a Bayer CFA reduces resolution to between 70% and 96.5% of what you would get without, depending on how you define and measure "resolution". (Unless you are a hard-core Foveonist using red-blue resolution test patterns, in which case the one true answer is 50%).
So what is the true resolution of a Bayer filter camera compared to a non Bayer filter camera? ;)
Mr. Reichmann has the eloquence to communicate technical details in an intuitive manner, and especially this makes these articles stand out. This is what good teaching is about. Absolute details are not neccessary in this context: informative, insightful articles about photography.As a photographer I completely agree and thank you for stating it so well.
There seems to be a lot of confusion in this thread about what's true or not, and what's exaggerated or not, and who's an expert or not.
Let's get down to basic principles. A red pixel in the Bayer type system has to block both green and blue light. The green pixel has to block both blue and red light, and so on.
If there were no overlap between the 3 primary colors, as much as 2/3rds of the light impinging upon the sensor would be blocked. However, there is an overlap, so the net effect of removing the color filter array is probably about one stop more light reaching the sensor with any given exposure.
If one were to remove the CFA from any Bayer type sensor, I therefore presume the manufacturer could raise base ISO by one full stop, yet still retain the same SNR and DR.
If the D800 were not only offered with an option of no AA filter, but also the option of no CFA filter, then the CFA-less version (D800BW) would have 14 stops of DR at ISO 200, instead of ISO 100, and would have greater resolution by a degree which exceeds the difference between the D800 and D800E.
Anyone care to dispute that? ;D
I think what people are trying to say is that...
For bayer, "effective MP" = 2/3(bayerMP)
For B&W, "effecitve MP" = "MP"
And the other way is if...
if a camera has x MP in a bayer array and another camera has x MP in a B&W sensor, then the B&W sensor is equivalent to 3/2(x) in a bayer matrix.
So a 36MP bayer is more realisticly a 24MP sensor, or at least produces the same amount of detail as would a 24MP B&W sensor.
I think I got that right :*)
I do not understand why one person would choose to challenge the statements of an easily identifiable person while choosing to hide anonymously behind a fake name. What do you have to hide? Or, are you afraid of something? You would gain a lot of credibility by being brave and come out of hiding. Be brave young duck! 8)
Besides, since when the name or person of the author has been relevant to the factuality of the content? Or is is only content one doesn't like that can be dismissed unless the person reveals his real identity? I could just as well use an alias which would sound like a real persons name - you'd not be able to critizice the content because of the anonimity, but would that be more honest from me?
If one were to remove the CFA from any Bayer type sensor, I therefore presume the manufacturer could raise base ISO by one full stop, yet still retain the same SNR and DR.Basically yes, except that it is a matter of definition not choice when it comes to doubling the ISO saturation based sensitivity (what is so often refered to as base ISO, but I prefer to use the proper wording as in the ISO's documents, like ISO12232).
Indeed this is what is seen with the Leica MM: its base sensitivity is 320, compared to 160 for the M9, with exposure index on the MM being a "push", overexposing and losing one stop of highlight headroom in exchange for better shadow handling.
It seems, from looking at he published DR graphs of digital cameras (have not seen one from this B&W sensor yet), that there really is only one "real ISO speed" and high ISOs are manufactured by underexposure and sacrificing DR.There is a lot of ambiguity in the phrase "ISO speed", which is why I try to refer to the multiple different measures of sensitivity and exposure index in the actual ISO standard 12232.
Michael, with all respect, since no colour RAW histogram has been possible till now according to the article, I must then be the first human being plotting a real colour RAW histogram (Canon 350D):
Where did the straw man argument that green is only responsible for luminance come from? That's not what I wrote or have ever written. They are predominantly responsible in a Bayer array, but not totally.I see now that you have corrected that part of the review. Excellent.
But you are unusual and special, Guillermo. ;D Most of us are not concerned with such finer technical points, and what may be technically possible if we take a lot of trouble, especially when such achievements have certain disadvantages, perhaps in respect of the attractiveness of the review image on the camera's LCD screen, for example.
I would love to have RAW histogram in camera. Even if its available only in review mode.Agreed
RAW developers can show RAW histograms and it should be really easy to code, especially if you leave 4 colors with two different greens.I tend to think that the gain of the two green channels is a complexity that I dont want to think about when using my camera. If there is some slight deviation, I'd rather see them bundled together in one histogram.
I don't see reason why it's harder than making histogram from jpeg file. WB is not important, because I want to see RAW data histogram, and this are just numbers.They are selling in-camera JPEGs to the most customers, for low-end to medium-end products, that functionality is probably the most important one, and re-using this functionality for high-end products is basically free. Canon and Nikon seem to be putting a lot of R&D and marketing effort on producing the "best" jpeg files, meaning that they are unlikely to jeopardize their investement.
It's just producers are lazy and they don't want to see how users struggle with UniWB, etc. to get something closer to RAW histogram.
AgreedI tend to think that the gain of the two green channels is a complexity that I dont want to think about when using my camera. If there is some slight deviation, I'd rather see them bundled together in one histogram.They are selling in-camera JPEGs to the most customers, for low-end to medium-end products, that functionality is probably the most important one, and re-using this functionality for high-end products is basically free. Canon and Nikon seem to be putting a lot of R&D and marketing effort on producing the "best" jpeg files, meaning that they are unlikely to jeopardize their investement.
Any raw feedback would (probably) have to be developed especially for niche cameras, and for a niche audience. Soccer-mums might think that the additional flexibility clutters the user interface. So while the requested functionality might be real simple, it might not make sense for the major players to include it - especially as long as the lu-la readers & friends continue to purchase products despite the lack of raw histograms.
Are photographers at the Super Bowl or Olympics going to be interested in raw histograms or fiddling around with raw conversion?Most of them: no. That was what I was trying to say in my post.
Do you count landscape and street photograph as "niche"?Not necessarily, but I count those who care about ETTR as a minute niche compared to all of the Japanese and Germans purchasing expensive cameras to take snaps of their pets and family and occasional holliday trips using AE and jpeg exclusively.
Most of them: no. That was what I was trying to say in my post. Not necessarily, but I count those who care about ETTR as a minute niche compared to all of the Japanese and Germans purchasing expensive cameras to take snaps of their pets and family and occasional holliday trips using AE and jpeg exclusively.
Do you think that all landscape and street photographers work in raw format and obsess about the noise levels found in the darkest parts of their images? I dont. I am sure that you could find any number of photograpers producing more interesting images than myself who havent got the faintest idea what ETTR is or how many stops of usable DR her camera has.
So much of this discussion is centered around the resolution issues involved with the Bayer array. But in order to give quantifiable numbers for comparision, we really do need to be VERY specific about the demosiacing algorithm being used.
In a handful of algorithms, green is what is used to determine luminance. In others, each one is used for that pixel location. Yet in others three pixels are averaged together, while once in a while an algorithm actually used four.
Since we photographers here on Luminous Landscape are the smartest kids in the classroom, I personally would prefer that when we toss these numbers out that we identify just which conversion algorithm we are using. It's the little things like this which make a big difference in backing up our conclusions.
We ARE the smart kids, right?Ken N.
QuoteMichael wrote:This is absolutely false. If you knew what you are talking about, instead of just repeating what someone at Leica has told you, you'd know that. Of course, I would be delighted if you would tell of one single thing that would need to be reengineered (at the "chip" level - I assume you mean the sensor, but maybe something else).
Engineering a monochrome sensor equipped camera isn't simply a matter of removing the Bayer array. Though based on the M9 sensor, a significant amount of reengineering at the chip level was required
Your lack of knowledge is clear here as well ...
Also with all due respect, you may well talk to all kinds of experts, but that does not mean that you are an expert in this field (digital imaging, including sensor operation). You are not.
Several differences can be mentioned here...Aku claimed that there were no re-engineering changes needed on the sensor level. Your reply talked only about changes needed on firmware level. I don't see your post disputing any of Akus claims?
It is quite easy to pick on many (most??) articles for technical correctness. ...I think it is important that _if_ some photography reviewer wants to talk about sensels and other technical details, he/she should get it right. If they don't want to talk about nitty-gritty technical details, that is fine too.
Aku claimed that there were no re-engineering changes needed on the sensor level. Your reply talked only about changes needed on firmware level. I don't see your post disputing any of Akus claims?
You can try to be a smart alec, and find all sorts of ifs and buts.Not trying to be anything except discuss the matter at hand. I suggest you do the same. Michael said "something". Aku interpreted this as "A", and claimed that "A" was false. You claimed that "B" was true. B != A.
The point, as I see it, is that Micheal implied some differences in monochrome and color cameras can have an effect on the availability, timing, etc. You can try to narrow it down to a sensor and firmware difference and nit pick. For the cameras we have designed, firmware was mostly on the cameras. So, for me, in broader sense, the question is the difference between color and monochrome camera design and not necessarily between color and monochrome sensors.So if Michael had more generally stated that a monochrome version of a CFA camera may be more expensive and complicated to manufacture than what one might initially think, both me and you would agree to that. Claiming that "a significant amount of reengineering at the chip level" was required is more concrete.
Not trying to be anything except discuss the matter at hand. I suggest you do the same. Michael said "something". Aku interpreted this as "A", and claimed that "A" was false. You claimed that "B" was true. B != A.
No, there is no Boolean (or any other) algebra involved here.I am not a camera designer, and have never claimed to be one.
(2) Open almost any book, Internet article, etc and stare at the xyz chromaticity diagram. Do you see what is wrong? I shall leave that as a homework exercise ;D.
Insinuating that my motivation for participating in the discussion is any less honorable than yours does not change my view (lines such as: "smart alec", "go ahead and nitpick", "armchair sensor designer").
Since I'm very interested on this subject, could you please explain what is wrong?
Yes, the diagrams are usually drawn that give an incorrect notion of distance in the color space, IMHO.
Ok, Thanks, they are not perceptually uniform and one can draw wrong conclusion. Anyway this is too OT, so I'll leave it.
Why should we trust a theory if it confirms your false assumption of Foveon resolving power? No sensor resolves beyond Nyquist, and Bayer CFA equipped sensors resolve over 90% of Nyquist. Even if we use the conservative 90% figure, it leads to Bayer pixel being worth 81% of a Foveon pixel in this regard. Not 50%.Please go away!
Nothing "special" about that. All manufcturers have off-axix microlenses on mirrorless cameras, and possibly even in the DSLRs (though there the benefits would be significantly less).
Wrong. All manufacturers have this capability. Off-axix microlenses are standard technology. And to be even more precice, Kodak and Dalsa are far from being anywhere near the leaders of the pack regarding microlens technology.
Also, they are part of the "toppings", not integral to the sensor itself. Leica could make them themselves (though they don't necessarily have the tools inhouse and the expertise in this special field is certainly lesser than of Sony of other major players in image sensor bussiness).