Puts news

[Herb Sennet]

Advertisement (gone after registration)

M9 final verdict

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[Anguish]

Well stated:

 

"you need to adapt to the characteristics of the camera to get the most out of it."

 

Agree

 

"The great value of the M9 is the concept of compactness, simplicity, direct vision, lack of lowpass filter and the rangefinder/manual focus mechanism, coupled with excellent engineering and accurate assembly."

 

Agree

[scott kirkpatrick]

There's a lot to agree with in his article, and I have always enjoyed his approach, but what does he mean by

 

"The shape of these micro-lenses is not exactly identical, element for element. In the case of the Leica M camera, the shape of the micro-lenses is such that some color fringing is exaggerated and even caused by that shape. "

 

Since the microlenses are inded " bubbles" laid down uniformly over the expanse of a wafer which is probably 8-12 inches in diameter, out of which some number of good image chips are then cut out, I can't see how the shape of the bubbles would cause the asymmetry of the Italian flag pattern, although the microlenses may be part of the process. But Mr Puts talks to people with knowledge of Leica's engineering, so I wonder what he has heard???

 

scott

[ho_co]

Excellent assessment as usual, IMHO.

 

Every time I read one of Erwin's articles, I feel like going out and buying another Leica lens or body.

 

 

Scott, I'm interested as well by his mention of "color artifacts at the Nyquist limit." Since his terminology differs from what we tend to use on the forum, I've no idea which 'defect' that refers to. (I mean, we pixel-peepers must have seen it, right? )

 

My guess in regard to the 'bubbles' would be that there is some misalignment problem, that each microlens doesn't line up with the Bayer matrix exactly, and indeed can't line up exactly for the ray patterns of all lenses, since they differ so much in exit pupil location.

 

It would be nice to see an example of the "color artifacts at the Nyquist limit," and a description of the involvement of the microlenses in the red edges problem. He may be waiting for Leica to publish its official explanation of the latter.

[scott kirkpatrick]

The Nyquist limit refers to the spatial frequency at which the pattern you are trying to resolve involves lines closer together than the pixel spacing, so that the gaps between a series of lines fall sometimes on a pixel, sometimes between. In a Bayer color sensor, this leads to Moire and also to color substitutions in very finely detailed parts of an image, such as distant tree branches.

 

The microlenses are displaced inwards towards the center of the image, so that they will capture incident rays that arrive at angles away from the vertical and reflect or refract them down into the pixel's collecting surface. I read somewhere that the displacement is proportional to tangent (theta), where theta is the deviation from the vertical -- zero at the center and as much as 45 degrees at the edges of the chip. I suspect that is an approximate result, valid for small angles, and some more complete simulation is used to optimize their layout near the chip edges. And of course, theta at each radius on the chip depends on the position of the particular lens's exit pupil, so the array of "bubbles" has to be designed for some typical set of lenses, and the rest will require some firmware corrections.

 

scott

[Guest stnami]

C/N philosophy????.......... D/L better watch out so should G/O/D.... a calling from the L ranch...... Nyquist is the man!!!!!

[LondonM]

Advertisement (gone after registration)

Nice review. And very well said. People who complain about the M9 remind me of people who complained about the Omega Seamaster when the new generation of watches with the coaxial caliber came out. Ohhh, they moaned, my $10 digital Timex keeps better time. My $10 digital Timex is lighter. etc. etc. etc.

 

C'mon folks. These are expensive, hand made mechanical objects whose sole purpose is to keep you as much in contact with the picture taking or time keeping process as is possible in this modern world of throwaway stuff. If you don't "feel" that connection, you bought the wrong piece of equipment.

[smb]

I contacted Mr. Puts ref. red edge. His reply was that up to 21mm all lenses possess this problem "in several shades of intensity." Personally, I like Mr. Puts evaluations. You need to learn to use the camera to its maximum potential which he states is quite high.

[Jerry_R]

his mention of "color artifacts at the Nyquist limit." Since his terminology differs from what we tend to use on the forum, I've no idea which 'defect' that refers to. (I mean, we pixel-peepers must have seen it, right? )

Indeed, pixel peepers can only see that: http://www.l-camera-forum.com/leica-forum/leica-m9-forum/117511-m9_moire.html

[diogenis]

There's a lot to agree with in his article, and I have always enjoyed his approach, but what does he mean by

 

"The shape of these micro-lenses is not exactly identical, element for element. In the case of the Leica M camera, the shape of the micro-lenses is such that some color fringing is exaggerated and even caused by that shape. "

 

Since the microlenses are inded " bubbles" laid down uniformly over the expanse of a wafer which is probably 8-12 inches in diameter, out of which some number of good image chips are then cut out, I can't see how the shape of the bubbles would cause the asymmetry of the Italian flag pattern, although the microlenses may be part of the process. But Mr Puts talks to people with knowledge of Leica's engineering, so I wonder what he has heard???

 

scott

 

My guess is that he means, that it's not an elegant solution, because it also introduces some problems. It's something has to be there out of necessity. Maybe with some manufacturing improvement in the future these can be resolved and M9 show even beter final results

[diogenis]

His final thoughts as presented in the last paragraph generally reflects what these cameras are. However, I remember that he did not like M8 even though one can say that apart from the obvious bigger sensor both cameras are identical. So, the bigger sensor makes that much of a different camera or did he exaggerated a bit when he referred to the M8?

[chris_tribble]

So, the bigger sensor makes that much of a different camera or did he exaggerated a bit when he referred to the M8?

 

I genuinely think that the bigger sensor makes the difference. I can shoot my 5D2 and M9s side by side for the same project and there's no compromise. I now have some of the best tools in the world for shooting in the 28-50 range and the 50-300 range (the longest prime I own). I also have a highly portable, compact reportage system which lets me work from 18 to 135.

 

Full frame is the thing that really made the difference. That's why I took the financial pain of transferring from M8 to M9.

[rosuna]

Puts doesn't talk about the reliability problems of the M9 (or the M8). He has connections with Leica engineers, isn't? Those problems has to be present in any valuation of the camera.

[jaapv]

I have seen very few M9 reliability problems in this forum. There seems to have been a small series where there has been a problem with the IR filter cracking, There have been some users with SD card problems, usually solved by changing brand, and there has been Mitch. But he must be the most unlucky Leica buyer in the universe.Nothing points to a structural problem with reliability, these things appear to be incidents..