Leica M240 review

[ErikKaffehr]

Hi,

You happen to deny image processing science. Once you pass about 10% MTF at Nyquist aliasing will occur. This is well known. It is actually a possibility that incredible feather separation you see in 20x30" prints is fake detail. If you post a raw image or an actual pixel crop it would be possible to see if you or   I am right.

Here are some sure signs of aliasing effects:

• Strains of hair shifting RGB colours
• Strains of hair that is not continuous
• Jagged thin lines

Here is an article about it. Schneider has a better one but I have not found it.
http://www.dvxuser.com/articles/article.php/20


I enclose a test shot by Imaging Resource from a Leica M9-P. It is converted by ACR. I show it in monochrome to avoid distracting color artifacts.
The image is here: http://216.18.212.226/PRODS/M9/FULLRES/M9hRES5216F.DNG

What this image shows is fake detail left of the red line. Note that it is not easy to count the lines, the shade of the lines differ.

The other crop also shows some fake detail.

If you want to demonstrate these effects on you DMR you could just take a sharp picture at f/8 or larger aperture of Norman Koren's test target: http://www.normankoren.com/Tutorials/Lenstarg_50_5906p_15g_25is.png

Photograph it at 2.5 m with a 50mm lens.

You will see an area where the lines turn into either a grey mass, or an irregular pattern, right of that area you will se an emerging line pattern that is an alias of higher frequency detail.

http://www.normankoren.com/Tutorials/MTF5.html

Making the pixels larger will increase aliasing. Stopping down to say f/11 or f/16 will reduce aliasing, as diffraction will act as a nice kind of OLP filter.

Note: the views below are at 200% (twice actual pixels).

I also added a screen dump comparing Leica M9 with Nikon D800 (not the E-version). The third image shows identical fields in two test images from Imaging resource. In this case the sensor of the Leica doe not properly resolve the line pattern and produces low frequency aliasing artefacts the Nikon D800 resolves the line pattern cleanly and produces no artefacts. The reason the Nikon does not show artefacts is because the sensor resolves higher.

I guess that the M9 has a similar pixel size to the DMR and I presume that the Leica M lenses are of similar quality, so I would presume that the DMR would perform pretty like the M9. The M9 has microlenses, at that may reduce aliasing to some extent. I don't know about microlenses on the DMR.

Best regards
Erik

Talk is cheap, Eric, especially when you've never shot with a DMR.  Show me a single file with this fake detail you insist exists.  That's hogwash and you write about this as if you have the definitive answer.  Let's see it!  

I wouldn't be able to make 20x30" prints of Hummingbirds showing incredible feather separation if there was fake detail.  This is yet another load of bollocks by someone who has never shot with the DMR.

Lawrence

[ErikKaffehr]

Hi,

The document from Schneider Optics here: http://www.schneideroptics.com/pdfs/whitepapers/optics_for_digital_photography.pdf describes aliasing pretty well. The relevant information is on pages 8-9.

Best regards
Erik

[wildstork]

"Another observation about the DMR is that it is a camera that obviously produces a lot of fake detail. You cannot put a very good lens on a camera with large non OLP filtered pixels without producing a lot of artefacts. It seems that photographers object to colour moiré but largely ignore or even enjoy other aliasing artefacts. Aliasing will produce fake detail, that is a fact."

"I guess that the M9 has a similar pixel size to the DMR and I presume that the Leica M lenses are of similar quality, so I would presume that the DMR would perform pretty like the M9. The M9 has microlenses, at that may reduce aliasing to some extent. I don't know about microlenses on the DMR."

Your first comment was with respect to the DMR, Eric.  You then go on to post a sample of a test chart taken with an M9P.  Different sensor, different microlenses, different software... yet this doesn't matter to you as you use it as evidence to support your claim. 

You then go on to state that you presume that "the M9 would perform pretty like the M9." 

You're very presumptuous.  Making assumptions like this with no evidence is of no use to anyone on any forum since you can't substantiate your claims with evidence from a different sensor, software etc.  Yet still you assume.

I've never known anyone who bought a DMR to shoot test charts.  Most of us use the camera so shoot everything but test charts and I've never seen the effects you attribute to the DMR.  In your infinite wisdom, you've proven all of the imaging engineers at Kodak, Imacon, Leica, Sigma, Sinar, Phase, and everyone else who uses CCD and CMOS chips devoid of OLP filters to be wrong, with your statement "You cannot put a very good lens on a camera with large non OLP filtered pixels without producing a lot of artefacts."  So not only are these companies selling faulty sensors on their cameras but they're peddling us mediocre lenses as excellent lenses given your assertion above. 

It's truly unfortunate that all of the medium format sensor makers and all of the camera makers that use CCD and CMOS sensors devoid of OLP filters didn't consult with you before dumping untold resources into image making devices that create so much fake detail.  I suggest that you present your credentials to them as it would save them a lot of money spent on technology that produces "a lot of fake detail." 

You've taken sophistry to a new level.  This is one of the unfortunate aspects of the internet, as some will read your post and make assumptions, just as you have, repeating your nonsense on other forums until it becomes commonplace.  And all the while those of us shooting with Leica DMR's, M8's, M9's, M240's, S2's Sigma Merrill's and all medium format digital backs will be producing beautiful images with our cameras using sensors that produce so much "fake detail."

You're digging a wonderful hole for yourself in full view of the entire internet world to see.  And no doubt, I've provided you with a new shovel to continue your task.  Have at it.

Viva los detalles falsos!!!
Lawrence 
 

[BJL]

Newer cameras haven't approached the DMR's colors so there's no need to 'upgrade' every few years.

Since output colors depend greatly on the details of conversion from raw sensor data to the final RGB file (JPEG pr whatever), how on earth do you make this judgement of unrivalled DMR color superiority? Comparing default raw conversions or out-of-the-camera JPEG's at default settings? Comparing the colors that you get with a work-flow you have developed based on considerable experience with that camera to what you can get when trying another camera that you are less familiar with?

The latter is what I will call "familiarity bias", which leads some Mac OS and Windows users to confidently declare the OS they know best to be easier to work with in some absolute sense, when all they really mean is "for me". it is also, I suspect, what leads some long-time rangefinder users like Mark D., Ph. D. to declare that rangefinder focusing is more accurate than focusing with magnified live view.

[wildstork]

I won't presume to speak for Doug, BJL, but in answer to your question I'll mention that I corresponded at length with any number of photographers who moved from the DMR to the Sony A900.  The DMR was a hybrid stop-gap effort by Leica to fill a void in their digital SLR line while they worked on an all-digital R.  Unfortunately the entire R line was shelved in favor of the S2.  The A900, technologically speaking, was light years ahead of the DMR in terms of features.  But I've yet to hear from a single A900 user who moved over from a DMR who didn't say the DMR produced better color. 

I bought a Sony A77 because I need the cropped sensor focal length magnifier for bird photography as well as other subjects I shoot.  The DMR was off to Somls for as long as 5 months and I needed a substitute.  I was told by some that the A77 was the next best thing to the DMR though it was a technological tour de force by comparison.  While the bells and whistles were a most welcome improvement... the color, in short, sucked compared to that of the DMR.  And if I have to spend a significant amount of time to tweak the color to approach that of the DMR. only to suffer the significant amount of color noise this Sony high pixel density sensor produces... I'm not interested.

I used to work with a 4x5 camera as well as medium format cameras.  I ran my own film on a Jobo processor and did high contrast masking for Ciba prints that were exhibited at Photography West in Carmel in the late 80's and early 90's.  It took a lot of time to get from film exposure to a final print.  Today, even with the kludge that is the DMR, I can get from A to Z in a fraction of the time and with better results.  I'll never go back to photochemical printing and it hurts to lose a venue like Photography West simply because I refuse to dump toxic chemicals into the toilet.  P West won't even look at digital output. 

What matters to some of us who used to shoot and print our film is the color and sharpness Leica optics provide.  These qualities may not be important for those who post jpgs on the internet, or print to a maximum size of 11x14", but it matters when your market is for extreme enlargements.  The sharpness difference Mark Dubovoy points out in his review is of the utmost importance when it comes to extreme enlargements, for just as you can't make a silk purse from a sow's ear... you can't make a sharp print (speaking of extreme sizes here) from a soft file. 

Sharpness matters... and anyone who comments: "My main problem with Mark's review is something that he doesn't even discuss -- that's the implication that the qualities he discusses are relevant to the practice of photography, which they aren't" is is simply pigeonholing every photographer into their "famous" photographer great image but out-of-focus category.  Statements like this show a complete disregard for the specific needs of different artists.

To begin an argument with the aforementioned quote and end with "Who cares if you can see the pupil of an eye, if you can feel the birdness of the bird?" only illuminates the reader as to the lack of understanding the poster has about wildlife photography as as well printing images large.  The eye is critical in wildlife photography.  It is the eye that connects us with the subject...  and failure to properly focus upon the eye results in an image that's far less compelling.

Lawrence 

[Telecaster]

...is pretty simple. Some people are incapable of accepting (or unwilling to accept) the notion that points-of-view other than their own are valid. Thus all the proselytizing and emotionalism.

-Dave-

[John Camp]

"But saying that the M240 is sharper than the D800e -- a controversial statement, to say the least -- is like saying my tires are very slightly prettier than yours."  

John,  That has to be some of the faultiest logic I've encountered on this forum. Seriously!  Sharpness can be qualitatively measured and agreed upon by all who have reasonably good vision and view two similar images.  Prettiness has nothing to do with it and shouldn't even enter into this argument as prettiness is a subjective value that cannot be quantified.  The only controversy here is your statement... with all due respect.  A very weak argument...

Your argument is now changing a qualitative review into an emotional statement about which is prettier or which image is worthy of being considered "famous".  Mark's review had absolutely nothing to do with what you're proposing here and I don't see how anyone can agree with your comments.  Obviously some do.  Mark spoke of sharpness and of one camera providing sharper results (and better color) than another.  He neither implied nor suggested that the M240 could produce images that could be judged more famous so why even bring this into the discussion?

And with all due respect, I've never had a post I've made here misinterpreted as thoroughly as you've done. I wasn't talking about prettiness, except to suggest the slight differences in prettiness would be irrelevant to tires...If you're buying tires, you really want something that's going to carry your car, and most of us don't worry much about small differences in attractiveness. This is called an analogy. It's an analogy to Mark's emphasis on sharpness, in which one camera is judged to be better than another if you can take out a microscope and detect a difference in sharpness. Well, no. By the time all the other aspects are taken into account, that kind of difference in sharpness is meaningless.

Also, I was using "famous" photos as examples from which we can all make comparisons and conduct a reasonable discussion, because we all know them. I asked if anyone could think of a famous photo in which sharpness was critical, because I couldn't. (Although I did think of one -- exactly one -- last night, although it's a news/feature photo, not an art photo, and that's the shot of the Afghan woman by the National Geographic photographer. In that shot, as much sharpness as he had was critical, but he didn't need any more.)

As for your focus on a bird's eye, I would suggest that's your craft. Not an art form, but a very specific kind of craft that's not much relevant to anything but your craft. And you're welcome to the craft; I have no problem with that at all. I would suggest, however, that you've probably never taken a photo as good as

http://www.soulcatcherstudio.com/exhibitions/trees/caponigro.html

Would an eye make the photo more compelling?

 

[ErikKaffehr]

Hi,

I have not invented aliasing, Nyquist limit or signal processing.

It is quite basic, if you put a high resolution lens in front of a low resolution sensor the sampling will introduce low frequency aliases. Cameras without OLP filtering produce aliasing, unless resolution is reduced by some means. Here is a good example of aliasing artifacts for Pentax 645D:



The example is for Pentax. I am pretty sure that similar issues would be seen on the DRM but raw files from DRM are not that abundant on the net. It would also be problematic to discuss about a feathers of a bird, you can show the cameras interpretation of the bird but it says little if you don't have a sample of the same bird.

There are two ways to reduce aliasing, one is to reduce MTF using an optical low pass filter or stopping down to small apertures, the other one is to increase resolution.

Regarding aliasing, the combination of a good lens and sensor with large pixels and no OLP is a bad combination. Microlenses can help, as they increase the fill factor. A system with higher fill factor will produce less aliases.

On everyday subjects it would be hard to tell fake and real details apart, on test charts it is more obvious, because you can count the line pairs, if you see the same number of line pairs in the image as in the subject the detail is probably real, but if you fewer line pairs than in the subject the details are obviously fake.

For real world subjects it may matter little, but once you are aware of the matter it will be much more obvious.

Sorry for my words, I am an engineer by profession. If I see something I can't understand I try to collect information, design tests and try to find out.

Just to mention, my normal print size is A2. I essentially print A2 if I print, that is 15.5"x23.4".  I would like to print larger but it is not practical. A2 prints make a nice picture when framed at 50x70cm with a passepartout. If I print larger I would send the file for printing, don't do that so often. Also, I always shoot raw.

Best regards
Erik

"Another observation about the DMR is that it is a camera that obviously produces a lot of fake detail. You cannot put a very good lens on a camera with large non OLP filtered pixels without producing a lot of artefacts. It seems that photographers object to colour moiré but largely ignore or even enjoy other aliasing artefacts. Aliasing will produce fake detail, that is a fact."

"I guess that the M9 has a similar pixel size to the DMR and I presume that the Leica M lenses are of similar quality, so I would presume that the DMR would perform pretty like the M9. The M9 has microlenses, at that may reduce aliasing to some extent. I don't know about microlenses on the DMR."

Your first comment was with respect to the DMR, Eric.  You then go on to post a sample of a test chart taken with an M9P.  Different sensor, different microlenses, different software... yet this doesn't matter to you as you use it as evidence to support your claim.  

You then go on to state that you presume that "the M9 would perform pretty like the M9."  

You're very presumptuous.  Making assumptions like this with no evidence is of no use to anyone on any forum since you can't substantiate your claims with evidence from a different sensor, software etc.  Yet still you assume.

I've never known anyone who bought a DMR to shoot test charts.  Most of us use the camera so shoot everything but test charts and I've never seen the effects you attribute to the DMR.  In your infinite wisdom, you've proven all of the imaging engineers at Kodak, Imacon, Leica, Sigma, Sinar, Phase, and everyone else who uses CCD and CMOS chips devoid of OLP filters to be wrong, with your statement "You cannot put a very good lens on a camera with large non OLP filtered pixels without producing a lot of artefacts."  So not only are these companies selling faulty sensors on their cameras but they're peddling us mediocre lenses as excellent lenses given your assertion above.  

It's truly unfortunate that all of the medium format sensor makers and all of the camera makers that use CCD and CMOS sensors devoid of OLP filters didn't consult with you before dumping untold resources into image making devices that create so much fake detail.  I suggest that you present your credentials to them as it would save them a lot of money spent on technology that produces "a lot of fake detail."  

You've taken sophistry to a new level.  This is one of the unfortunate aspects of the internet, as some will read your post and make assumptions, just as you have, repeating your nonsense on other forums until it becomes commonplace.  And all the while those of us shooting with Leica DMR's, M8's, M9's, M240's, S2's Sigma Merrill's and all medium format digital backs will be producing beautiful images with our cameras using sensors that produce so much "fake detail."

You're digging a wonderful hole for yourself in full view of the entire internet world to see.  And no doubt, I've provided you with a new shovel to continue your task.  Have at it.

Viva los detalles falsos!!!
Lawrence  
 

[hjulenissen]

Is it inconcievable that some cameras, while appealing to some photographers, and producing appealing images in the right hands, have technical limitations that other cameras may not have (to the same degree)?

If you have a successful career or a rewarding hobby as a photographer not caring about measurements and test-charts, why should you care if someone comments on the measurements and test-chart abilities of your chosen tools?

-h

[Rob C]

Is it inconcievable that some cameras, while appealing to some photographers, and producing appealing images in the right hands, have technical limitations that other cameras may not have (to the same degree)?

If you have a successful career or a rewarding hobby as a photographer not caring about measurements and test-charts, why should you care if someone comments on the measurements and test-chart abilities of your chosen tools?-h

Take great care: sounds as if you are turning into me!

;-)

Rob C

[BJL]

... I've yet to hear from a single A900 user who moved over from a DMR who didn't say the DMR produced better color. ... even with the kludge that is the DMR, I can get from A to Z in a fraction of the time and with better results. ... What matters to some of us who used to shoot and print our film is the color and sharpness Leica optics provide.  These qualities may not be important for those who post jpgs on the internet, or print to a maximum size of 11x14", but it matters when your market is for extreme enlargements.

You did not directly give any answers to any of my questions about what raw conversions are used in these color and sharpness comparisons, but you silence on that point and a few of your comments quoted above make it sounds like you are indeed comparing default (in-camera?) JPEG conversions, for both color and sharpness.  If so, then it puzzles me that with all your emphasis on sharpness and color accuracy and "extreme enlargements", you ignore even the basic post-processing that most digital photographers would apply in preparing such prints --- especially since this effort is so much less effort than was required with making large, fine prints from film. For example, it seems that many DSLR's deliberately do not apply "print-ready" levels of sharpening to their default output, and err if anything on the side of low contrast, and so produce default output that is under-sharpened and slightly flat if viewed large: this is deliberate, on the expectation that sharpening and such will be part of the preparation for large prints, and such sharpening is better applied later in the process, with the intended print size taken into account. And if you really want out-of-camera files that are ready to print large, then the cameras settings should be adjusted before you judge a camera by its default output.

[telyt]

You did not directly give any answers to any of my questions about what raw conversions are used in these color and sharpness comparisons, but you silence on that point and a few of your comments quoted above make it sounds like you are indeed comparing default (in-camera?) JPEG conversions, for both color and sharpness.  If so, then it puzzles me that with all your emphasis on sharpness and color accuracy and "extreme enlargements", you ignore even the basic post-processing that most digital photographers would apply in preparing such prints --- especially since this effort is so much less effort than was required with making large, fine prints from film. For example, it seems that many DSLR's deliberately do not apply "print-ready" levels of sharpening to their default output, and err if anything on the side of low contrast, and so produce default output that is under-sharpened and slightly flat if viewed large: this is deliberate, on the expectation that sharpening and such will be part of the preparation for large prints, and such sharpening is better applied later in the process, with the intended print size taken into account. And if you really want out-of-camera files that are ready to print large, then the cameras settings should be adjusted before you judge a camera by its default output.

You are making numerous unfounded assumptions.  Lawrence is not comparing in-camera jpg files.

[Ray]

You are making numerous unfounded assumptions.  Lawrence is not comparing in-camera jpg files.

They don't seem like assumptions to me but more like questions. For those of us who are used to adjusting to taste, the color, white balance, contrast, sharpness and vibrancy etc. of each image in Lightroom or ACR as we process it, any claims that one particular model of camera produces more pleasing color and contrast than another should immediately raise the suspicion that the images being compared are either jpegs straight out of the cameras, or 'default' RAW conversions without further adjustments having been made.

Whenever I compare the images of identical scenes from different cameras and/or different lenses, I always try to get the adjustable broad variables the same, such as color and contrast, so that I can more clearly see the advantages that one camera or lens may have over the other, such as lower noise with equal sharpening, or higher resolution.

[ErikKaffehr]

Hi,

This is one of the questions I have. How much different is color rendition between different sensors, and how much of that is coming from the sensor, the color conversion matrix in the camera and from the color conversion pipeline in the raw processor.

Regarding the sensor, what matters is the CGA (Color Grid Array) which are known to be different between cameras and sensors.

Some examples and thoughts are listed here: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/71-mf-digital-myths-or-facts?start=9

What is correct color is pretty much up to taste. How green is the grass? Some liked Velvia in film times and some liked Ektachrome, those films were very different.

To an extent, the accuracy of color reproduction can be measured. DxO presents something called Sensitivity Metamerism Index, which is based on reproduction of color checker values. We see millions of colors and the Color Checker has just eighteen colors, so SMI obviously doesn't say it all, but if a sensor cannot reproduce the 18 CC colors correctly it is probable that it cannot reproduce the other millions of colors either.

For me, there are far more questions about color reproduction than there are answers.

Best regards
Erik

They don't seem like assumptions to me but more like questions. For those of us who are used to adjusting to taste, the color, white balance, contrast, sharpness and vibrancy etc. of each image in Lightroom or ACR as we process it, any claims that one particular model of camera produces more pleasing color and contrast than another should immediately raise the suspicion that the images being compared are either jpegs straight out of the cameras, or 'default' RAW conversions without further adjustments having been made.

Whenever I compare the images of identical scenes from different cameras and/or different lenses, I always try to get the adjustable broad variables the same, such as color and contrast, so that I can more clearly see the advantages that one camera or lens may have over the other, such as lower noise with equal sharpening, or higher resolution.

[OldRoy]

Whilst I tend to fall asleep when reading some of the - undoubtedly highly informed - discussion in this thread, which is in any case of no more than casual interest to me, I keep coming back to the observation made in Erwin Puts' piece. In it he compares Mark Dubovoy's assessment of the M8 series with the current camera.

"The big question: the Leica M8 gets the identical comments (stunning, superb) that the Leica M is receiving, but the M8 is now “a bust, it was not a good camera”, but then it offered “exceptional shooting experience and extraordinary image quality”.
What will Mark Dubovoy say about the Leica M when a new Leica M (2014) will be announced. "

Call me mean-spirited, but such glaring contradiction completely undermines any value in the review. Did I miss something whilst asleep or has anyone found a way to explain away this 180 degree swerve?
Roy

[Rob C]

Whilst I tend to fall asleep when reading some of the - undoubtedly highly informed - discussion in this thread, which is in any case of no more than casual interest to me, I keep coming back to the observation made in Erwin Puts' piece. In it he compares Mark Dubovoy's assessment of the M8 series with the current camera.

"The big question: the Leica M8 gets the identical comments (stunning, superb) that the Leica M is receiving, but the M8 is now “a bust, it was not a good camera”, but then it offered “exceptional shooting experience and extraordinary image quality”.
What will Mark Dubovoy say about the Leica M when a new Leica M (2014) will be announced. "

Call me mean-spirited, but such glaring contradiction completely undermines any value in the review. Did I miss something whilst asleep or has anyone found a way to explain away this 180 degree swerve?Roy

Stunning footwork? Pelmanism?

Rob C

[Ray]

To an extent, the accuracy of color reproduction can be measured. DxO presents something called Sensitivity Metamerism Index, which is based on reproduction of color checker values. We see millions of colors and the Color Checker has just eighteen colors, so SMI obviously doesn't say it all, but if a sensor cannot reproduce the 18 CC colors correctly it is probable that it cannot reproduce the other millions of colors either.

Good point, Erik. It's clear from the DXOMark tests that the Leica pixel in the M240 has no advantage compared with the smaller Nikon pixel in the D800E. However, at the lower base ISO of ISO 73 for the D800E, as opposed to ISO 134 for the Leica, as measured by DXO, the smaller D800E pixel seems to have equal or better performance across all parameters measured, including SNR, DR, Tonal Range and Color Sensitivity.

If we compare equal size prints, which is usually the sensible thing to do, the difference in the technical performance between these two cameras is magnified. The DR of the D800E is over one full stop better than the M240 at base ISO, and 1.46 EV better at ISO 3200.

That's a very significant difference, yet Mark Dubovoy writes in his review:

I set the two cameras at ISO 3200 to shoot the following high contrast scene: ...At ISO 3200 the Leica image is cleaner, with less noise.

Now what the heck is going on here? That's a huge discrepancy. Are you feeling quite all right, Mark? 

I can only assume that the indoor kitchen scene that Mark shot for this comparison at ISO 3200 did not have any deep shadows. It was a contrasty scene only in relation to the bright light at the window, which is blown in both shots. The SNR at 18%, for the D800E, is only marginally better than the Leica, at ISO 3200, according to DXO, and marginally better to a degree than one probably wouldn't notice.

I think what is happening here, in this article, is that Mark is rationalizing after the event, a purchasing decision he made which was not at all rational. Hence the title, "I will not buy that camera. I promise.." 

http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/844%7C0/(brand)/Leica/(appareil2)/814%7C0/(brand2)/Nikon

[hjulenissen]

Whilst I tend to fall asleep when reading some of the - undoubtedly highly informed - discussion in this thread, which is in any case of no more than casual interest to me, I keep coming back to the observation made in Erwin Puts' piece. In it he compares Mark Dubovoy's assessment of the M8 series with the current camera.

"The big question: the Leica M8 gets the identical comments (stunning, superb) that the Leica M is receiving, but the M8 is now “a bust, it was not a good camera”, but then it offered “exceptional shooting experience and extraordinary image quality”.
What will Mark Dubovoy say about the Leica M when a new Leica M (2014) will be announced. "

Call me mean-spirited, but such glaring contradiction completely undermines any value in the review. Did I miss something whilst asleep or has anyone found a way to explain away this 180 degree swerve?
Roy

Enthusiasm fuels many positive human endeavours, but tends to wear off over time.

-h

[hjulenissen]

Now what the heck is going on here? That's a huge discrepancy. Are you feeling quite all right, Mark? 

A measurement of SNR and DR does not have to correlate perfectly with perceived "SNR" and "DR". Throw in somewhat different color correction matrixes etc.

I think what is happening here, in this article, is that Mark is rationalizing after the event, a purchasing decision he made which was not at all rational. Hence the title, "I will not buy that camera. I promise.." 

Don't we all try to rationalize our (predominantly) intuitive, gut-feeling-dominated, subjective way of being?

-h

[Ray]

.Don't we all try to rationalize our (predominantly) intuitive, gut-feeling-dominated, subjective way of being?

-h

I wonder if this piece was written for April 1st but published a bit late.