Mechanische camera's

M8 part 1: first encounter september 2006
Leica M8: a small step for mankind, but a big step for Leica
Introduction


On the surface the M8 offers the design elements and the operative characteristics of the current M-line. A 10Mp CCD-chip inside the body captures and transforms the images projected by the lens onto the sensor area in image files that can be manipulated by a suite of software programs. The development team has been quite successful in the transmutation of the film-loading M7 into the CCD-based M8.
Staying as close as possible to the M-concept sets limits and brings opportunities.
You cannot pack a host of features into a compact body and it would violate the underlying design principles that have been the guideline for every M-model since 1954. The Leica M3 was designed as a camera with full manual control of the main photographic parameters (focus of the lens and exposure setting by selecting a combination of shutter speed and aperture). These controls had to be laid-out as intuitively as possible. The photographic style that has become possible with the M-camera is rooted in the seamless integration of the camera with the mind-hand-eye coordination necessary for creative and spontaneous imagery. The Leica designers assumed with reason that photography is primarily a mental process with the camera an unobtrusive part of that process.
The Leica camera has always been designed as a precision tool. This designation is usually interpreted as referring to the technical and manufacturing aspects of the camera. But is also applicable to the photographic process, where precision implies a sharply focussed, well-exposed and well-framed photograph.
The technical development of the camera took a different direction when the industry introduced more and more electronic components into the camera that automated the basic photographic controls.
From M3 to M7 Leica has carefully balanced on that dividing line between manual operation and automatic operation. Sometimes one would assume that Leica stayed too close to the manual directive. The R8 is a clear example of a camera that was designed form scratch to incorporate a motor-driven transport, but here Leica hesitated to cross the line.
The Konica Hexar RF implemented a vertically running shutter with integrated motor drive in the body shape of the M-camera and showed that a sprinkle of automation was feasible and attractive for M-photographers.
The ultimate level of automation has been reached with the shift to digital recording of the image, where the image file can be manipulated on the level of individual pixels
Much has been written about the reasons for Leica to keep away from these developments.
The Epson exercise to incorporate a CCD sensor into a Cosina Bessa has not been very convincing and may be living proof for the decision by Leica to invest in a longer development cycle.
The main problem for the Leica designers to incorporate the sensor into the M body has its roots in the past. M lenses have always been designed with the twin goal of compactness and superior optical capabilities. The M camera owes its fine haptical qualities to the slim body contours. The depth of the body has been kept quite narrow. These two aspects, the narrow body with the famous 27.8mm flange-to-film gate distance and the compact size of the lenses forced the designers to locate the exit pupil as close to the film surface as possible. What was a brilliant move to optimize performance when capturing on a non-opaque film layer, became a liability when designing for a solid-state opaque sensor surface.
The solution took a long period and involved the use of dedicated shapes for the micro-lenslets over the individual pixels, the absence of the low-pass filter and the coating of the surfaces of the filters before the sensor surface. Once these problems were solved, the camera could be developed.
The M8 body
The body of the M8 is a Solms design, made in Portugal. It follows as closely as possible the classical M shape and I have to say that on a first glance the resemblance is striking. The build quality is outstandingly good and the body feels solid as a rock. A very small gripe would be the easy feel of the shutter speed dial when selecting shutter speeds. The camera now has the R8/9 shutter assembly with speeds till 1/8000, a first for a rangefinder camera. The sound of the shutter is extremely subdued; it produces less noise than the M7 shutter: the noise when the heavy curtains must be stopped by the braking device is clearly absent. On the other hand I have to say that the sound of the motor-driven shutter cocking is less pleasing, but that is a matter of adjustment.
The high top speed is fine, but more important is the fact that we now have really accurate speeds around 1/500 and 1/2000.
The dimensions are not identical to the M7 shape. See the table below

Dimensions in mm	M8	M7
Width of top cover	37	34
Width of top cover with eye-piece extension	40	36
Width of bottom cover	32.5	32.5
Overall length	139	139
Height	80	80
Width of body with display/speed dial and flange	42.5	37

The main difference is the extrusion of the bayonet flange that at 2mm is more than the M7 where the flange hardly extends from the body. This extension is required to keep the distance from lens flange to sensor surface the same as in all M bodies. The sensor is located slightly in front of the film gate and that explains the difference. The small increase in thickness gives the body a more boxlike impression that is somewhat less elegant than the classical M shape.
At last we find in the bottom cover a centrally located tripod attachment. On the other hand we may deplore the disappearance of the X-contact for external flash connection.
Features
The M8 has borrowed heavily from the M7: it has the same level of automation: AE with aperture priority and manual speed dial. The shutter is the same as we find in the R8/9. The speeds are 8 sec to 1/8000 in manual with half steps and in AE the range is from 32 sec to 1/8000 in quite fine steps. Flash synch is now 1/250 sec. You will miss an exposure override in automatic mode, unless you set in one of the menu options, which is quite cumbersome. When operating in manual mode, you can compensate by the red triangles in the finder as is the norm for every M since 1984.
The battery is lithium ion, and can fire about 300 shots (depending of course on the level of studying the pictures in the display). As a comparison the Canon 5D can hold 600 to 800 pictures. The CCD is a currency guzzler!


The monitor has a diameter of 2.5 inch (the industry standard) with 230000 pixels. The display is quite clear and shows fine detail, but is hardly visible in sunlit conditions (all displays are lousy in this respect). You can enlarge the picture to 100% (1:1 view) to study sharpness details.
The camera has two positions S and C for Single and Continuous Mode. In C mode the camera can take ten shots (RAW) in a row (about 2 per second) and then has to wait for transferring the buffer content. This takes about a minute and is quite slow. Canon 5D figures (17 RAW in 5 seconds, buffer empty in 20 seconds) are better, but hardly an industry standard
ISO speeds are a bit queer form ISO 160 to 2500 in full steps. Leica gives the basic sensor sensitivity as ISO160, a move that can be applauded. Why Leica has not opted for the more traditional 200, 400 etc range is a bit of a mystery.
The menu items are very simple and you can set sensor sensitivity, exposure correction, white balance, picture size and picture resolution. In fact there is no need for more options: the load of options found in other cameras do not improve the quality of the photograph.
The shutter release
The fully mechanical release for the horizontally running cloth curtain shutter has a very precise pressure point. It is part of the essence of the M camera. 
You can fire the shutter at exactly the moment you want and you can depress the shutter release with hair-trigger accuracy. The M8 has an electro magnetic-shutter release that is less responsive. After pressing the release, there is a slight moment of hesitation and indecision before the shutter actually is released. This might be a minor issue, but the lack of this precision feel gives the camera its own profile, which may or may not be important. When we add the absence of the transport/shutter wind lever, we must conclude that the camera has its own personality.
The finder
The finder has three sets of frame-lines: 24/35, 28/90 and 50/75 and a magnification of 0.68, slightly less than the classical 0.72, but required to add the 24 frame line. The 24/90 frame-lines correspond to angles of view for the 32 and 120mm lenses in 35mm photography. This reduction was required to accommodate the frame lines of the wide angle lenses, that were a bit cramped at the edges of the finder in the classical M finder. The accuracy of the rangefinding is not impaired by this reduction. The 135mm lens can be used with confidence as far as RF accuracy is concerned.
But there is another caveat: the 1.33 reduction in viewing angle implies an additional 33% enlargement factor and that implies a 33% enlargement of the diameter of the Circle of Confusion. Any rangefinder inaccuracy will be translated in a 33% additional inaccuracy. The reduction in viewing angle to the equivalent of an 180mm would imply an extremely small frame that makes it impossible to frame and compose the scene.
The 1.33 factor for the reduction in angle of view (also referred to as crop factor) is of course reflected in the frame-lines being used. When putting the 24mm on the body, it is a bit of a surprise to see a 30mm frame being selected. Normally one would not bother, but when you are used to the M7 body, and select the 24mm lens you know that the finder cannot show the full view of the lens. On the M8 the finder gives frame-lines that you would expect to cover the 35mm lens.
The frame-lines themselves are a bit less accurate than what you can expect form the M6/7/P etc finders. You get more on the sensor than is framed in the finder. At distances less than 1.5 meters the finder is quite honest. There is also some flare in the rangefinder patch when strong light sources are present in the scene.
Sensor issues
Sensor is CCD with 18x27mm area and 3916 x 2634 pixels in DNG format. Pixel size is 6.8 micron. In addition you have 4 different JPEG sizes and you can take pictures in RAW + JPEG sizes together. In my view all these JPEG options can be thrown away: the computer software can handle this more efficiently. The choice for DNG looks sensible, but there is no guarantee that this DNG format will survive and in fact many programs cannot read this program. Leica supplies a new version of Capture One LE (3.7.5) that is not yet available on internet. At least not at this moment of writing.
The sensor is made by Kodak and has as one of its ancestors the Olympus E1 sensor architecture. The software is written by a German company (Jenoptik), based on specifications by Leica.
Conclusion
The M8 fits in seamlessly in the evolution of the M camera. It is however not an M7 with a chip implanted. It is a camera with its own personality that leans much more than any previous M model in the direction of automated functionality. The M8 enables the continuation of the classical M-style of photography in the digital domain. The main argument for investing in the M8 (buying an M8 would be a too simple action), is the continued use of the Leica M lenses and anyone who owns a suite of Leica lenses would be happy to put them on an M8 and enjoy the fine optical characteristics.



The design of the M8 is focussed on this enabling aspect: the use of the M lenses. Given the technological infrastructure inside the camera, we have to admit that the superior optical imagery of the Leica lenses cannot (yet) be exploited in all dimensions. The overall performance of the M8 (optics, sensor technique and post processing) will be most certainly a match for the best players in the market (the advanced and professional DSLR-models), but it will be difficult to surpass them.
But the Leica package is much smaller than the bulky DSLR's and the ease of use is very convincing. Of course a comparison with the new smaller DSLRs would shift the balance a bit in the size issue, but the Leica is the more compact camera.
The M8 then is a camera model for existing M users. Given the features that are quite important in today's market environment (battery power, speed of use, an overload of selectable options), the M8 has little to offer. Professional photographers (with a handful of exceptions) have long ago abandoned the M-system and it is unlikely that they will return to this system. The enthusiast photographer (as it is known now) wants the long list of features now commonly provided by the major manufacturers and here the M8 has its weaknesses. The short list of features is on the other hand precisely what the M photographer wants, and it is this watershed line where Leica now seems to feel a bit insecure. It is clear from the design of the M8 that Leica is firmly committed to continuity and small incremental changes. Where the rest of the photographic world is preparing for the deep changes that follow in the wake of the technological possibilities, Leica has an eye on values from the past.
There is nothing wrong with this approach, but Leica now knows very well that a broader product line with more modern products is required to survive. The Leica M in whatever livery is a niche product.
Fifty years ago, Leitz had only one product for the photographic community, the M3 and M2 models, and they stayed very long in this monoculture, partly by conviction, partly by economic necessity.
The monoculture of the M has been abandoned and replaced by a full product line.
One would even say that the range is too wide: there is hardly any mention nowadays of the film-based line of products, but we still have the M7 and MP, the R9 and the C/CM-range. With the new D-Lux3, the Digilux3, the V-Lux1 (all Panasonic derivatives) Leica offers a range of cameras with 10 MP sensor sizes, excepting the Digilux with 7.5 Mp. The internal competition for the M is of course the Digilux3 4/3 system. The Digilux is a descendant of the Olympus E1 lineage and that is most interesting as the E-1 was designed to become the Leica camera for the digital arena.
A fuller discussion of these issues will be presented in the next part. 



The Leica M8, the first professional digital rangefinder in Leica speak has been introduced in september 2006 at the Photokina 2006, exactly 81,5 years after the original Leica at The Leipzig Fair in spring 1925 and 52 years after the introduction of the M3 at Photokina 1954. There seems to be one new Leica system for every generation of photographers.
The rumor circuit gave the M8 extremely high marks and this was the camera that would learn the manufacturers of digital cameras a lesson. The M8 camera has been showered with praise and also with critique. 
The true verdict may lay between these two extremes: the camera is not the world beater some had anticipated it to be and it not so bad as some severe critics may assume. 
It is a camera that needs a clean look and appraisal that cuts through the dense fog of faithful admiration and the deception that follows when expectations are set unreasonably high. 
This series of articles looks at the various aspects of the M8 and tries to evaluate the camera on its measurable qualities and some subjective reflections. 
This series has been written over several months of working with the camera. Some changes in view and opinion will be noticed in the course of the reviews. 


M8 part 2: image quality and performance
introduction
In my previous report I noted that the Leica M8 feels and operates like a true M-camera, but with enough unique points to support the claim that the camera has its own personality, that fits into the current notions of the digital workflow. Leica has always stated that their main expertise is focused on the realization of high quality imagery with filmbased and sensorbased photographic instruments. In this part I will present an in-depth analysis of the optical properties of the sensor of the M8, the limits of the testing procedures and the often neglected issue of the many factors that influence the feasible results when taking photographs in real-life situations. It is assumed in all reports that the maximum attainable quality is also the quality that can be consistently expected in practice. My testing in very controlled lab-situations indicate that this is a myth. The bandwidth of the quality of the practical results is much wider that one assumes and even the bandwidth of test results is also much wider that the tester will acknowledge.

This second part of my M8 report took a longer time to realize than expected because of a number of themes that I encountered during testing.
Leica has upped the ante with claims that their solution is best when you wish to exploit the inherent optical performance of the Leica lenses.
Therefore a more stringent method of testing was adopted to detect small, but meaningful distinctions. The ubiquitous approach of visual comparison on screen does not suffice, nor can we rely on most types of testcharts. I have reported on these issues in previous posts. One of the (neglected) problems with on-screen visual inspection is the influence of the quality of the screen. This parameter is not often included in the quality equation.

What is the influence of the low-pass filter?
Basically Leica has selected a very thin glass plate before the sensor surface with the argument that a thick glass plate will degrade the optical performance. This is undoubtedly true. A glass plate of 3mm in front of the sensor areas does seriously degrade the quality of the image. We also know that beyond the resolution limits defined by the Nyquist frequency, the fine structures are not recorded with the true (actual) frequencies, but with some false frequency. This phenomenon has been identified as aliasing and moire-effects, actually two different phenomena. The low-pass filter (LPF) has been introduced to cut off the fine frequencies at the limit of the Nyquist frequency in order to reduce the recording of details beyond this limit. There are many variants of the LPF and without knowing more features, one cannot state the actual influence. The basic working of the LPF can be seen below (a matlab simulation). The fine detail is indeed eliminated, but also the contrast has been reduced severely.


The resulting MTF would be quite low and this cannot be accepted. We may assume then that the major companies like Canon, Nikon and Sony, to name a few, will employ all kinds of mathematics to create their dedicated LPF. The question is then not: is there an LPF or not, but what kind of LPF has been employed. In any case the reduction in contrast and the loss of fine detail needs to be counteracted and here the major difference between Leica and the companies that use LPF is the post-processing software. In this area the Japanese companies have a major lead and in fact it is amazing what the software can accomplish. But there are limits as usual. These are not often discussed.

Above the Nyquist frequency (the safe zone so to speak), the contrast may be quite high, but the contrast is dependent on the phase of the periodic structure that is being detected by the sensor-raster. A high contrast can be found with a 'wrong' phase and will introduce artifacts in the image. The world is not so clean as we want it to be.
The Leica M solution is less dependent on the post-processing software and will allow the optical quality of the lenses to be recorded as honest as possible.

In the end both solutions may deliver the same result at the limits of resolution, but we cannot emphasize enough that these limits are not always relevant, nor defined with accurate measurements.
The test
I used for the test a Siemens star pattern with very fine structures. This pattern has a good correspondence with actual imagery and when the test pattern is well recorded, so we can confidently predict that practical results will be quite good as well.
The chart has been photographed at a distance of about two meters. At this distance the magnification of the chart will be such that all periodic structures will be recorded and more pictures are made at this distance than at infinity, where most tests are conducted. But the performance of the lens is different at infinity and at medium distance ranges. I used a variety of lenses form Leica and Zeiss. (a separate report will discuss this). For this test report I will concentrate on the new 1.4/50 asph (SLA) and the new Summicron 75 (ASC).

As comparison system I selected the Canon 5D. This has a sensor with the classical 24x36m format and delivers excellent, if not superior imagery. The lens was the macro-2.5/50mm. Form previous MTF measurements I know that this lens at medium apertures is quite good and should be a better match for the Summilux than the Canon version of the 1.4/50mm I am not interested here in comparing lenses, but comparing sensor quality. We would expect that the Leica M8 must be at least the equal of this system.

The following illustrations are enlarged sections of the sensor area. Leica recommends the use of Phase One Capture One LE software that is in the box. I compared the results of the CO LE with Photoshop CS2 9 and Camera Raw. The differences in processing are very small, so I opted for Adobe. Camera Raw settings are: sharpness: 25, colour noise reduction: 0. All adjustments are on Auto. The picture represents the center part of a much larger image file; in fact the selection is 1/40 of the total area.

Both pictures are raw versions with minimal adjustments. On screen you can detect colour artifacts in the fine details of the spokes of the star as imaged by Leica.

The Canon image too has this type of artifacts, but slightly less visible. It is now easy to jump to conclusions, but one must take into account some additional facts. This star-test is quite heavy on the image detector and the fineness of the detail is such that one will not encounter these fine details in practical work quite often. Secondly there is a difference between screen representation and print representation. I made prints of the whole test chart image with the Epson R2400, controlled by the Photoshop print options. On the A3 prints, the differences are almost negligible! We should then add the printer software as an additional element in the digital image equation. Thirdly and most significantly we should realize that at these high frequencies the results are not reliable in a consistent way and an element of chance is introduced. There is an almost universal tendency to interpret figures as absolutes and attach great value to small differences. Some magazines and also some websites present resolution figures in absolute terms like this: sensor/lens combo A has a resolution of 1836 lines per image height and combo B ' only' 1470 l/ih. It is then assumed that 1800 is better than 1500 and that these values are attainable in practical work. Both assumptions are wrong. Without the additional parameter of contrast these values are hardly insightful.

My measured results for the M8/75 combo are 1300 lp/ih (2600 l/ih) at 10% contrast. At 50% contrast the resolution is about 900 lp/ih or 50 lp/mm! The Leica claim then is substantiated? Not exactly: depending on the direction of the spokes of the Siemens star the MTF varies between 50% and 30%. Here you see clearly the danger of presenting figures without context. Based on the same measurements we can present the MTF as best value (50) or worst value (30) or average value (40). The mere figure in itself does not convey relevant information. There is also the issue of averaging the results over the image area: it makes a difference whether you present values from the center of the image or averaged over the whole sensor area. In the first case you test the maximum performance of the sensor and in the second case you present numbers that represent the average quality of the lens behavior over the image field. I have analyzed countless measurements with the M8 system and several lenses at all apertures. My view at this moment can be summarized as below.

The theoretical Nyquist limit for the M8 sensor is 73.5 lp/mm. My best results (see above) are close to that value for the Leica M8 system at aperture 4 (beyond this value the diffraction errors become visible) : about 1300 lp/ih or about 72 lp/mm. More is not attainable with this sensor. But you can infer from the pictures above, that at these extreme resolutions all kinds of artifacts are introduced. Experience tells us that good definition should be found at values that are 0.7 or 0.5 of the theoretical Nyquist value: in this case around 40 to 50 lp/mm.

When you study a large amount of MTF graphs that represent the M8 level of performance (and the same applies to the Canon set of graphs!) it is safe to say that up to 800 lp/ih the results are representative of practical performance and above 900 lp/ih we enter a zone where large margins of error must be assumed and the results around and above 1000 lp/ih are at best interpreted as tendencies.

Results compared with Canon 5D
The Canon picture presents equally good imagery, but see the comments above). At least in my lab conditions the results are comparable. But one needs additional information to see the results in perspective. I will not present the graphs as I have noticed that people will interpret the values at a level of precision that is not realistic. The Canon system has an inherent 30% advantage in imagery because of the 1.33 angle reduction for the Leica M8. In normal parlance: when we want to print an image at the same size (A3 as example) the smaller sensor needs a higher level of definition to present the same quality.

The measured result for the Canon system at aperture 5.6 to 8 is about 1200 lp/ih at contrast values from 10% to 30%. That is 50 lp/mm at quite low contrast values. But when we correct this resolution with the 1.3 factor, we can see that Leica needs 65 lp/mm to bring the same detail on the print with the same printsize. At these values the Leica has a contrast of 10%. But more meaningful is this comparison: At 1000 lp/ih Canon brings a contrast of 80% into the scene and Leica has 60% (75mm ) and 50% (50mm). All values at optimum apertures. The Canon behavior is quite intriguing. All Leica graphs follow the familiar pattern of a best value at lower frequencies and then a gentle roll off to the zero position. Canon has a different behavior: Here the values stay quite high from lower frequencies to mid frequencies (about 1000 lp/ih) where the contrast gets a software induced boost. A value of 80% contrast at frequencies around 30 to 50 lp/mm cannot be generated by the lens alone. Here the post processing must help. The upshot of the Canon approach is a high contrast image over a wide range of frequencies that is quite pleasing for the eye and also brings that digital look/fingerprint in the picture.

I am not taking sides which approach is best: the Leica pictures are more in line what you expect from silver-based negatives.
I do not have the intention to pursue the comparison Canon-Leica ad nauseam. Suffice it to say that Leica with its smaller sensor and thin glass plate presents results that up to A3 can compare quite well with the Canon images and do present a different look. If you are a number-crunching person, you might say that the Canon system has the edge even with less superior lenses.
All tests and evaluations were also done with the Summilux-M 1.4/50 ASPH. The results are not that different form the 2/75 that a whole paragraph should be devoted to this topic.
Variations in image quality when sensor speed is changed and JPEG sharpness is changed.
The same series of tests done with Raw pictures was also conducted with the various JPEG compressions and the various ISO values.

For the comparison of JPEG sharpness results I used the Summilux at f/5.6. Again there is a bandwith of values in the high-low range. At the meaningful 1000 lp/ih we see these values: Off: contrast = 5 to 10%; low: contrast = 20 to 30%; medium: contrast = 20 to 40%; Med high: contrast = 25 to 45%; high: contrast = 40 to 60%;
The difference between the Raw image and the JPEG image at medium sharpness level is not that great. In many cases one can shoot pictures confidently with JPEG without having to fear that definition will suffer. Other aspects of course underlie the usual JPEG constraints.


ISO values were compared using the 2/75. Again at 1000 lp/ih the values are ISO 160: 20 to 40%; ISO 640: 5 to 30%; ISO 1250: 5 to 20%; ISO 2500: 3 to 10%. The results indicate that (noise issues apart) ISO 1250 is the maximum speed at which you will see good contrast values to get good definition of fine detail, but is safer to set the maximum at ISO 640. I leave it to the user of the M8 to find out the best combination of ISO speed and JPEG compression to do the job as required.

The consistency of results in practical work.
Under the lab conditions it is quite easy to find the best focus of the camera/lens combo, you would expect. But it took some effort to focus accurately to be able to capture the finest possible details. I also noted that some movement of the camera (by changing the distance unintentionally or by taking handheld shots as a comparison with the tripod-based pictures) could degrade the image to some degree. Most users of the Leica M will take pictures in handheld conditions and we all know from practice that it is quite easy to move the camera when taking portraits or photographing persons in action. The lab-related pictures may show the maximum possible quality, but it would be unrealistic to assume that these results can be reproduced in normal shooting. I also became aware of the fact that when making a series of pictures where you try to find the best focus per every individual picture, you may find that some pictures are way off the mark. That is normal and again substantiates the claim that testing must be done in controlled situations to avoid all kinds of error introduced by accuracy and consistency issues.

Below you find a picture of the Siemens star with the lens wide open and deliberately defocused, but not by very much. Within the central part of the star image you see a blurred area where some patterns can be recognized. This is a phase jump of the periodic pattern and is an indication that the focus is off. The drop of contrast is most visible. The resulting picture looks similar to what you get when using a LPF before the sensor! I have presented this picture to show that there are major differences in image quality, simply by a defocus blur. On the other hand we may be happy that the mechanical precision of the Leica is so accurate that the results can be very good. But we should realize that image quality may be much lower than we anticipate: in may cases we have only our own level of competence to blame!


The picture displayed here is a really critical object: it is flat and has extremely fine lines. In real life we will take pictures of three dimensional objects with depth and then the defocus will be less visible: the sharpness plane will be at some location on the object and there is visible sharpness, when not at the intended spot.

Here again we see that trying to get consistent and repeatable results is less easy than is often assumed. Results from tests may vary according to the care and expertise of the tester.

Conclusion
The results of this test indicate that the Leica M8 can deliver results that are excellent and will not unfavorably compare to other top quality systems. Especially when one takes into account the smaller sensor area.
The Leica M digital solution does exploit the image quality of Leica M lenses to the limit of the 50/50 rule (50 lp/mm at 50% contrast). For best range finder accuracy, the use of the 1.25 magnifier is strongly advised. I have also indicated that there is a wide margin between the best possible (maximum) results and practical results. I did note that the mirror movement of the Canon could destroy the delicate fine lines of the target, even when the camera was on tripod.
One of the issues I wish to ask attention for is the focus topic. With digital imagery, the cost of focus bracketing is zero and it will definitely help to get better imagery in many situations. When using the 2/75 with focus bracketing I noted that the steepness of the focusing curve in the lens mount implies a definite on-off position for sharpness. That is fine for quick and accurate focusing, but the steepness makes it difficult to accomplish fine tuning in the focus bracketing mode.
For best definition we should employ Raw images and stay within the ISO range from 160 to 640.

Preview
In the next part I will compare the M8 images with the M7 images made with the same lenses. Can film go further than the 50/50 rule states?
In addition I will compare pictures of 3-D objects to see whether the conclusions from this lab report need fine tuning for real life situations.
I will also look at the dynamic range, the noise and the color aspects of the M8 and discuss the IR-issue, that has popped up in the internet.


The Leica M8, the first professional digital rangefinder in Leica speak has been introduced in september 2006 at the Photokina 2006, exactly 81,5 years after the original Leica at The Leipzig Fair in spring 1925 and 52 years after the introduction of the M3 at Photokina 1954. There seems to be one new Leica system for every generation of photographers.
The rumor circuit gave the M8 extremely high marks and this was the camera that would learn the manufacturers of digital cameras a lesson. The M8 camera has been showered with praise and also with critique. 
The true verdict may lay between these two extremes: the camera is not the world beater some had anticipated it to be and it not so bad as some severe critics may assume. 
It is a camera that needs a clean look and appraisal that cuts through the dense fog of faithful admiration and the deception that follows when expectations are set unreasonably high. 
This series of articles looks at the various aspects of the M8 and tries to evaluate the camera on its measurable qualities and some subjective reflections. 
This series has been written over several months of working with the camera. Some changes in view and opinion will be noticed in the course of the reviews. 


M8 part 2: image quality and performance
introduction
In my previous report I noted that the Leica M8 feels and operates like a true M-camera, but with enough unique points to support the claim that the camera has its own personality, that fits into the current notions of the digital workflow. Leica has always stated that their main expertise is focused on the realization of high quality imagery with filmbased and sensorbased photographic instruments. In this part I will present an in-depth analysis of the optical properties of the sensor of the M8, the limits of the testing procedures and the often neglected issue of the many factors that influence the feasible results when taking photographs in real-life situations. It is assumed in all reports that the maximum attainable quality is also the quality that can be consistently expected in practice. My testing in very controlled lab-situations indicate that this is a myth. The bandwidth of the quality of the practical results is much wider that one assumes and even the bandwidth of test results is also much wider that the tester will acknowledge.

This second part of my M8 report took a longer time to realize than expected because of a number of themes that I encountered during testing.
Leica has upped the ante with claims that their solution is best when you wish to exploit the inherent optical performance of the Leica lenses.
Therefore a more stringent method of testing was adopted to detect small, but meaningful distinctions. The ubiquitous approach of visual comparison on screen does not suffice, nor can we rely on most types of testcharts. I have reported on these issues in previous posts. One of the (neglected) problems with on-screen visual inspection is the influence of the quality of the screen. This parameter is not often included in the quality equation.

What is the influence of the low-pass filter?
Basically Leica has selected a very thin glass plate before the sensor surface with the argument that a thick glass plate will degrade the optical performance. This is undoubtedly true. A glass plate of 3mm in front of the sensor areas does seriously degrade the quality of the image. We also know that beyond the resolution limits defined by the Nyquist frequency, the fine structures are not recorded with the true (actual) frequencies, but with some false frequency. This phenomenon has been identified as aliasing and moire-effects, actually two different phenomena. The low-pass filter (LPF) has been introduced to cut off the fine frequencies at the limit of the Nyquist frequency in order to reduce the recording of details beyond this limit. There are many variants of the LPF and without knowing more features, one cannot state the actual influence. The basic working of the LPF can be seen below (a matlab simulation). The fine detail is indeed eliminated, but also the contrast has been reduced severely.


The resulting MTF would be quite low and this cannot be accepted. We may assume then that the major companies like Canon, Nikon and Sony, to name a few, will employ all kinds of mathematics to create their dedicated LPF. The question is then not: is there an LPF or not, but what kind of LPF has been employed. In any case the reduction in contrast and the loss of fine detail needs to be counteracted and here the major difference between Leica and the companies that use LPF is the post-processing software. In this area the Japanese companies have a major lead and in fact it is amazing what the software can accomplish. But there are limits as usual. These are not often discussed.

Above the Nyquist frequency (the safe zone so to speak), the contrast may be quite high, but the contrast is dependent on the phase of the periodic structure that is being detected by the sensor-raster. A high contrast can be found with a 'wrong' phase and will introduce artifacts in the image. The world is not so clean as we want it to be.
The Leica M solution is less dependent on the post-processing software and will allow the optical quality of the lenses to be recorded as honest as possible.

In the end both solutions may deliver the same result at the limits of resolution, but we cannot emphasize enough that these limits are not always relevant, nor defined with accurate measurements.
The test
I used for the test a Siemens star pattern with very fine structures. This pattern has a good correspondence with actual imagery and when the test pattern is well recorded, so we can confidently predict that practical results will be quite good as well.
The chart has been photographed at a distance of about two meters. At this distance the magnification of the chart will be such that all periodic structures will be recorded and more pictures are made at this distance than at infinity, where most tests are conducted. But the performance of the lens is different at infinity and at medium distance ranges. I used a variety of lenses form Leica and Zeiss. (a separate report will discuss this). For this test report I will concentrate on the new 1.4/50 asph (SLA) and the new Summicron 75 (ASC).

As comparison system I selected the Canon 5D. This has a sensor with the classical 24x36m format and delivers excellent, if not superior imagery. The lens was the macro-2.5/50mm. Form previous MTF measurements I know that this lens at medium apertures is quite good and should be a better match for the Summilux than the Canon version of the 1.4/50mm I am not interested here in comparing lenses, but comparing sensor quality. We would expect that the Leica M8 must be at least the equal of this system.

The following illustrations are enlarged sections of the sensor area. Leica recommends the use of Phase One Capture One LE software that is in the box. I compared the results of the CO LE with Photoshop CS2 9 and Camera Raw. The differences in processing are very small, so I opted for Adobe. Camera Raw settings are: sharpness: 25, colour noise reduction: 0. All adjustments are on Auto. The picture represents the center part of a much larger image file; in fact the selection is 1/40 of the total area.

Both pictures are raw versions with minimal adjustments. On screen you can detect colour artifacts in the fine details of the spokes of the star as imaged by Leica.

The Canon image too has this type of artifacts, but slightly less visible. It is now easy to jump to conclusions, but one must take into account some additional facts. This star-test is quite heavy on the image detector and the fineness of the detail is such that one will not encounter these fine details in practical work quite often. Secondly there is a difference between screen representation and print representation. I made prints of the whole test chart image with the Epson R2400, controlled by the Photoshop print options. On the A3 prints, the differences are almost negligible! We should then add the printer software as an additional element in the digital image equation. Thirdly and most significantly we should realize that at these high frequencies the results are not reliable in a consistent way and an element of chance is introduced. There is an almost universal tendency to interpret figures as absolutes and attach great value to small differences. Some magazines and also some websites present resolution figures in absolute terms like this: sensor/lens combo A has a resolution of 1836 lines per image height and combo B ' only' 1470 l/ih. It is then assumed that 1800 is better than 1500 and that these values are attainable in practical work. Both assumptions are wrong. Without the additional parameter of contrast these values are hardly insightful.

My measured results for the M8/75 combo are 1300 lp/ih (2600 l/ih) at 10% contrast. At 50% contrast the resolution is about 900 lp/ih or 50 lp/mm! The Leica claim then is substantiated? Not exactly: depending on the direction of the spokes of the Siemens star the MTF varies between 50% and 30%. Here you see clearly the danger of presenting figures without context. Based on the same measurements we can present the MTF as best value (50) or worst value (30) or average value (40). The mere figure in itself does not convey relevant information. There is also the issue of averaging the results over the image area: it makes a difference whether you present values from the center of the image or averaged over the whole sensor area. In the first case you test the maximum performance of the sensor and in the second case you present numbers that represent the average quality of the lens behavior over the image field. I have analyzed countless measurements with the M8 system and several lenses at all apertures. My view at this moment can be summarized as below.

The theoretical Nyquist limit for the M8 sensor is 73.5 lp/mm. My best results (see above) are close to that value for the Leica M8 system at aperture 4 (beyond this value the diffraction errors become visible) : about 1300 lp/ih or about 72 lp/mm. More is not attainable with this sensor. But you can infer from the pictures above, that at these extreme resolutions all kinds of artifacts are introduced. Experience tells us that good definition should be found at values that are 0.7 or 0.5 of the theoretical Nyquist value: in this case around 40 to 50 lp/mm.

When you study a large amount of MTF graphs that represent the M8 level of performance (and the same applies to the Canon set of graphs!) it is safe to say that up to 800 lp/ih the results are representative of practical performance and above 900 lp/ih we enter a zone where large margins of error must be assumed and the results around and above 1000 lp/ih are at best interpreted as tendencies.

Results compared with Canon 5D
The Canon picture presents equally good imagery, but see the comments above). At least in my lab conditions the results are comparable. But one needs additional information to see the results in perspective. I will not present the graphs as I have noticed that people will interpret the values at a level of precision that is not realistic. The Canon system has an inherent 30% advantage in imagery because of the 1.33 angle reduction for the Leica M8. In normal parlance: when we want to print an image at the same size (A3 as example) the smaller sensor needs a higher level of definition to present the same quality.

The measured result for the Canon system at aperture 5.6 to 8 is about 1200 lp/ih at contrast values from 10% to 30%. That is 50 lp/mm at quite low contrast values. But when we correct this resolution with the 1.3 factor, we can see that Leica needs 65 lp/mm to bring the same detail on the print with the same printsize. At these values the Leica has a contrast of 10%. But more meaningful is this comparison: At 1000 lp/ih Canon brings a contrast of 80% into the scene and Leica has 60% (75mm ) and 50% (50mm). All values at optimum apertures. The Canon behavior is quite intriguing. All Leica graphs follow the familiar pattern of a best value at lower frequencies and then a gentle roll off to the zero position. Canon has a different behavior: Here the values stay quite high from lower frequencies to mid frequencies (about 1000 lp/ih) where the contrast gets a software induced boost. A value of 80% contrast at frequencies around 30 to 50 lp/mm cannot be generated by the lens alone. Here the post processing must help. The upshot of the Canon approach is a high contrast image over a wide range of frequencies that is quite pleasing for the eye and also brings that digital look/fingerprint in the picture.

I am not taking sides which approach is best: the Leica pictures are more in line what you expect from silver-based negatives.
I do not have the intention to pursue the comparison Canon-Leica ad nauseam. Suffice it to say that Leica with its smaller sensor and thin glass plate presents results that up to A3 can compare quite well with the Canon images and do present a different look. If you are a number-crunching person, you might say that the Canon system has the edge even with less superior lenses.
All tests and evaluations were also done with the Summilux-M 1.4/50 ASPH. The results are not that different form the 2/75 that a whole paragraph should be devoted to this topic.
Variations in image quality when sensor speed is changed and JPEG sharpness is changed.
The same series of tests done with Raw pictures was also conducted with the various JPEG compressions and the various ISO values.

For the comparison of JPEG sharpness results I used the Summilux at f/5.6. Again there is a bandwith of values in the high-low range. At the meaningful 1000 lp/ih we see these values: Off: contrast = 5 to 10%; low: contrast = 20 to 30%; medium: contrast = 20 to 40%; Med high: contrast = 25 to 45%; high: contrast = 40 to 60%;
The difference between the Raw image and the JPEG image at medium sharpness level is not that great. In many cases one can shoot pictures confidently with JPEG without having to fear that definition will suffer. Other aspects of course underlie the usual JPEG constraints.


ISO values were compared using the 2/75. Again at 1000 lp/ih the values are ISO 160: 20 to 40%; ISO 640: 5 to 30%; ISO 1250: 5 to 20%; ISO 2500: 3 to 10%. The results indicate that (noise issues apart) ISO 1250 is the maximum speed at which you will see good contrast values to get good definition of fine detail, but is safer to set the maximum at ISO 640. I leave it to the user of the M8 to find out the best combination of ISO speed and JPEG compression to do the job as required.

The consistency of results in practical work.
Under the lab conditions it is quite easy to find the best focus of the camera/lens combo, you would expect. But it took some effort to focus accurately to be able to capture the finest possible details. I also noted that some movement of the camera (by changing the distance unintentionally or by taking handheld shots as a comparison with the tripod-based pictures) could degrade the image to some degree. Most users of the Leica M will take pictures in handheld conditions and we all know from practice that it is quite easy to move the camera when taking portraits or photographing persons in action. The lab-related pictures may show the maximum possible quality, but it would be unrealistic to assume that these results can be reproduced in normal shooting. I also became aware of the fact that when making a series of pictures where you try to find the best focus per every individual picture, you may find that some pictures are way off the mark. That is normal and again substantiates the claim that testing must be done in controlled situations to avoid all kinds of error introduced by accuracy and consistency issues.

Below you find a picture of the Siemens star with the lens wide open and deliberately defocused, but not by very much. Within the central part of the star image you see a blurred area where some patterns can be recognized. This is a phase jump of the periodic pattern and is an indication that the focus is off. The drop of contrast is most visible. The resulting picture looks similar to what you get when using a LPF before the sensor! I have presented this picture to show that there are major differences in image quality, simply by a defocus blur. On the other hand we may be happy that the mechanical precision of the Leica is so accurate that the results can be very good. But we should realize that image quality may be much lower than we anticipate: in may cases we have only our own level of competence to blame!


The picture displayed here is a really critical object: it is flat and has extremely fine lines. In real life we will take pictures of three dimensional objects with depth and then the defocus will be less visible: the sharpness plane will be at some location on the object and there is visible sharpness, when not at the intended spot.

Here again we see that trying to get consistent and repeatable results is less easy than is often assumed. Results from tests may vary according to the care and expertise of the tester.

Conclusion
The results of this test indicate that the Leica M8 can deliver results that are excellent and will not unfavorably compare to other top quality systems. Especially when one takes into account the smaller sensor area.
The Leica M digital solution does exploit the image quality of Leica M lenses to the limit of the 50/50 rule (50 lp/mm at 50% contrast). For best range finder accuracy, the use of the 1.25 magnifier is strongly advised. I have also indicated that there is a wide margin between the best possible (maximum) results and practical results. I did note that the mirror movement of the Canon could destroy the delicate fine lines of the target, even when the camera was on tripod.
One of the issues I wish to ask attention for is the focus topic. With digital imagery, the cost of focus bracketing is zero and it will definitely help to get better imagery in many situations. When using the 2/75 with focus bracketing I noted that the steepness of the focusing curve in the lens mount implies a definite on-off position for sharpness. That is fine for quick and accurate focusing, but the steepness makes it difficult to accomplish fine tuning in the focus bracketing mode.
For best definition we should employ Raw images and stay within the ISO range from 160 to 640.

Preview
In the next part I will compare the M8 images with the M7 images made with the same lenses. Can film go further than the 50/50 rule states?
In addition I will compare pictures of 3-D objects to see whether the conclusions from this lab report need fine tuning for real life situations.
I will also look at the dynamic range, the noise and the color aspects of the M8 and discuss the IR-issue, that has popped up in the internet.

M8 part 4: silver (M7) versus silicon (M8)
Introduction
The Leica brand name has a strong iconic value: Leica was the first practical 35mm camera, and was quickly developed into the best precision miniature rangefinder camera of the photographic world. Leica has always been dedicated to the rangefinder concept and the typical style of photography that is possible with this instrument. The Leica rangefinder camera has its band of loyal followers who strongly advocate its virtues and possibilities. Any high profile product will also be criticized and there are at least as many Leica bashers as there are Leica devotees. The discussion between these groups sometimes looks like the culture wars we are experiencing in this first decade of the 21 century.
The current digital transformation of the photographic image and the method of recording pictures has profound implications for the future of photography. Photography has strong instrumental roots and for many camera users this aspect is the main focus of interest. This was the case in 1950 when the 35mm camera entered its first golden age. It was the heyday of magazines like Modern Photography, that explained the technicalities of photography to an eager public. Users of rollfilm cameras were not taken seriously and every self-respecting photo amateur wanted a 35mm reflex camera. This was the period of the great debates about the superiority of German lenses and engineering and of SLR versus RF and of Japan versus Germany as the best camera manufacturers. We have a rehearsal of that period now with the digitalization of photography, boosted by the ubiquitous and cheap use of the computer.
In the debate about finding the best Raw converter, you can see the shadows of the classical debate about the best film developer. In the war of the words about all technicalities of the digital way of image recording and manipulation there seems to be one victim: the simple joy of using a fine instrument to get the images you want. Use the M8 and experience the ease of handling and the advantages of the immediate feedback you get from digital capture. With film you do not have that large array of software to manipulate your images. The method of recording an image in layers of silver halide grain is pure and honest and lets you focus on the act of capturing the image.
The main advantage of the M8 is its integration in the world of M-photography and its easy migration path from film to digital while keeping the basic spirit of the Leica M camera line.
Then it is natural to inquire about the co-existence of both recording methods within the M world. And it is also of some interest to analyze the performance of current film technology versus the capabilities of the M8. All negatives have been set to TIFF in Photoshop with smart sharpening at 25% and radius 2.1.
Kodak Portra 160.
For the first test I used the newest version of Kodak's Portra film in NC and VC versions. The speed of ISO 160 happens to be the same as what you find in the M8 as basic sensitivity. The negatives were scanned with the Nikon Coolscan V ED at 4000 ppi. The files (in TIFF format) have a size of 70 Mb, compared with the 30 Mb we get from the M8 files also in TIFF. The area of the M8 sensor is 0.56 the size of the negative area of the M7. The same lens (2/75) has been used and the same model too.
Below: Kodak Portra: Bootom M8 at ISO 160


The Coolscan cannot record every detail that the emulsion is capable of capturing, but that is not that important as I intend to show the main differences. In addition the sizes of the pictures differ by the 0.56 factor, but I did compensate for this difference in scale, but not fully.
The Kodak film has excellent color reproduction and a very large total scale. The grain is amazingly fine for a speed that once was considered high. Comparison with the digital image show clearly two main differences: the grain pattern which gives a more natural look to the picture and the lower level of detail definition. The M8 image is sharper, shows more detail in the eye, but some would say that the skin is too plastic, because of the smoothness.
Spur Orthopan UR
This is a modern BW film, made in a country close to the Netherlands. It has a monodisperse emulsion with outstandingly fine grain and high contrast. The Spur developer adds some acutance as can be seen easily form the example. This film cannot hide its heritage as a microfilm, but the tonality is very good, if not as smooth as the Kodak.



Definition of detail is excellent and captures more detail that does the M8. The additional detail in the iris of the eye. 
Texture details
A second example (a very small section of the full image showing some cloth texture) indicates that a suitable film can capture more detail with the same lens that is possible with the sensor of the M8. The film has a limited appeal (ISO20 and relatively steep contrast), but it is extremely easy to work with. A full development cycle (pour developer in tank to final rinse with water) is about 10 minutes. The digital image heas been exposed to extreme posytprocessing and then the difference with the picture captured on film is less visible. But remember: here I have extracted every detail form the file: the overall picture would not be acceptable.
Some reflections
The high performance of the M8 is amazing, if one considers the relatively small sensor area of 18 x 27mm compared to the classical 35mm format of 24 x 36mm. It is questionable if one needs more image quality when the goal is an eye catching A3 picture on the wall.
If have reported very often that the quest for the ultimate image quality is like searching for that pot of gold at the end of the rainbow: it is tantalizingly close, but it will escape you forever. Most Leica M users operate their camera handheld: that is the best recipe for image degradation. Even at shutter speeds above 1/1000, that the M8 allows, there is the real problem of exact focusing.
Best imagery you can get when you photograph on tripod, with flash and a stationary subject in order to focus accurately and make a focus range. In all other situations, the basic M8 image quality will be higher than what is possible based on the imperfections of the human operator.
Let us have no false illusions. The M camera has always been an instrument that operates perfectly within quite narrow borders. The popularity of the (D)SLR is based on its broad use in all situations. The M8 follows the M-tradition of being perfect with a small footprint. With the M8 you can exploit the characteristics of the Leica lenses without the usual compromises. The sensor has a resolution limit and then the M7/MP take over and give you when loaded with suitable film the best imagery that is possible with 35mm photography.
My view is this: the image quality that the M8 delivers and the sheer joy of using the M camera, is the best argument to select the M8 as the main camera. (I sold my 5D). And the M7/MP stay in my bag loaded with Spur Orthopan and slide film (Velvia 50) when I need or want the ultimate performance of the Leica lenses.In the next installment I will analyze the performance of the M8 with a range of Leica and Zeiss lenses. The results will be surprising!

 M8 part 5: Performance of Leica and Zeiss ZM lenses
Introduction
In previous parts I have compared the M8 performance with a high-end DSLR and with film. The results lead to the summary conclusion, that the M8 image quality is superior to film, with the exception of microfilm technology and on a par with the high-end segment of the competition. It was also noted, and that is for me at least an important fact that the quality of definition is very close to the silver-halide based capture medium. Tonal scale, sharpness, definition and fine texture reproduction are reminiscent of the film based era. In this respect too, the Leica M8 is a heritage camera, transmuting the Barnack/Berek approach into the 21-century with digital means.
In all comparisons I used top-quality modern Leica M lenses (SL-A 1.4/50 and ASC 2/75). The nature of the digital capture is such that the recording of the finest structural details is limited by 0.7x the Nyquist frequency and the low frequency components of the image are artificially enhanced by the post recording algorithms. One might expect then a levelling off of absolute performance differences when comparing lenses, used on the M8. Using test charts is the best method for comparing lenses to study relative and absolute quality differences. The chart used is the extra fine Siemens star chart.
For easy comparison I have segmented the charts in four parts, together producing the complete chart. The presentation is always the same. The top row has the full aperture and the bottom row has the stopped down apertures, from 5.6 to 11, depending on the widest opening of the lens. The left column shows the centre portion of the image and the right column the edge position. See scheme below.
full aperture center and corner
stopped down center and  corner
The lenses I selected are the Summicron-M 2/35 ASPH, compared with the Biogon 2/35, the Summilux 1.4/35 compared with the Summilux-M 1.4/35 ASPH, the Elmarit-M 2.8/21 ASPH compared with the Biogon 2/35 and the C-Biogon 4.5/21, the Planar and identical the Summicron 2/35 with the Summilux-M 1.4/50 ASPH, and without comparison the Ultra-Wide Heliar 5.6/12, the Distagon 2.8/15 and the Apo-Summicron 2/75 ASPH.
The methodology for this comparison was: pictures made in DNG mode. Imported with Camera Raw with all settings on zero and transformed in Photoshop to JPEG, again with all settings as efault: no addition sharpening or contrast boost has been done. The goal was to see the pure or raw performance of the lens as recorded by the M8 sensor.
Apo-Summicron-M 2/75 ASPH
The SCA 2/75 performs as planned. Extremely fine structures are recorded with crisp micro contrast. Hardly any difference between stopped down (f/8) and wide open performance. A slight drop in quality in the edge position wide open. Note the dust spot at the edge of the image. 
UltraWide Heliar 5.6/12mm
The focal length of 12mm is the shortest available for the M camera. The performance is quite respectable, especially at the smaller apertures (f/11). Note the strange artifacts in the centre of the image at the small aperture. Vignetting is quite low and this shows the effectiveness of the special construction of the microlenses of the M8.

Distagon 2.8/15mm
This lens has more lens elements and is much heavier than the 12mm. The performance at 2.8 however is already better than what you get from the UWH. Quality wise the ZM Distagon is a very fine lens indeed. Stopped down is f/11. A pity it is so heavy. The level of vignetting at the corners (not shown: I used the edge position) is high, at least three stops.


Elmarit-M 2.8/21 ASPH and Biogon 2.8/21mm and C-Biogon 4.5/21mm
Can we see differences between the lens design with and without aspherical surfaces? If there is a difference, it can be found in the curvature of field, which is better in the Zeiss version. And the contrast wide open which is better in the Leica version. The digital images show the same type of quality difference that has been reported in the film based images. The Biogon 4.5/21 should be pronounced the winner in this contest. This lens exhibits excellent behavior at all apertures and field positions. It has an 'unfair' advantage of 1.5 stops of course, but if you can live with this, the lens itself can hardly be faulted. Stopped down is f/11 for all three lenses.

Summicron-M 2/35 ASPH and Biogon 2/35mm
I noticed a change of character in the most important focal length of the M camera: the 35mm. Now in the M8 it will be the replacement for the 50mm lens in the film-based environment. The Summicron 35mm ASPH has been the standard bearer of lens performance for a long time. But the Zeiss friendly revenge has created a challenge: the Biogon 35mm is in most aspects the better performer. In the centre and wide open the Summicron is still unequalled as contrast goes, but measured on most other criteria the Biogon is the current winner. Note the higher level of flare for the Summicron in the corners.

Summilux 1.4/35 and Summilux 1.4/35 ASPH
This is no contest. The older version of the 1.4/35 wide open has low contrast, low resolution and a lot of flare. Stopped down the performance is much better. Now you know why you need the ASPH version.

Planar 2/50 and Summicron 2/50 and Summilux-M 1.4/50 ASPH
Wide open at 1.4 and at the edge the new Summilux design shows some weakness in the contrast and resolution area, compared to the standard f/2 designs, but do not forget that there is a full stop difference! The new 1.4/50 Summilux deserves the conclusion as the best high speed standard lens in the world. The Planar and Summicron are identical in performance!

Vignetting
A full study of the vignetting behaviour needs another article. For now I can conclude that the film based versions have an advantage here, specifically if we take into consideration that the M8 has a reduced angle of view and should have a natural advantage. My measurements indicate that a 1.4/35 mm ASPH lens has more vignetting on the M8 than the Summilux 1.4/50mm ASPH has on an M7 loaded with film. Both lenses have roughly the same angle of view on the M8 and M7.
This result is typical for all the other lenses tested above. In all cases the film based behaviour was better than the sensor based behaviour. And the M8 has a restricted angle of view. The new microlens design is really needed, but even this construct cannot cope fully with the light fall off at the extreme corners.
Conclusion
The smaller angle of view of the M8 implies that a lens, designed for the full 35mm format will not be asked to bring into effect the extreme corners and edges. So most lenses will operate with a safety margin and the results above do testify this effect. In a sense the M8 does reduce the inherent differences between lenses, as the effects of the problematic edges and corners can be neglected.. Differences do exist as can be inferred from the examples shown above.
My earlier conclusion that in this digital age the optical designs should be concentrated on vignetting, flatness of field and reduced astigmatism and flare wide open and even performance over the image field are substantiated. The ZM lenses from Zeiss have done this already and the comparison between the ZM 35 and the Summicron 35 is an indication of this changing battleground. 
 
M8 part 6: Comparison Hasselblad H3D 39 Meg
Introduction
Many years ago, when silver halide capture was the only method of fixing a shadow (as the photographic act has been described) 35mm format was the rule and medium format the niche. Prominent standard-bearers of both systems were the Leica (35mm) and Hasselblad (120 format). Common sense and tests indicated that the larger capture area of medium format compensated for the inherent quality advantage of 35mm optical designs.
A number of tests were done (among others by Geoffrey Crawley who compared a top grade 35mm SLR with a low grade Chinese medium format twin lens reflex and myself who compared Hasselblad/Zeiss with Leica M/Leica optics) and there was acceptance on three facts: the smaller format required higher quality optics to squeeze the same amount of image information on a smaller area, the best 35mm optics captured the same level of image detail and the main advantage of medium format was the lack of grain at higher enlargements (A3 level), where the prominence of grain in the 35mm picture broke up image detail and smoothness of luminance gradation. So the area advantage of 2x of the medium format size can be largely offset by the better optical quality of the small format lenses, given equal emulsion characteristics.
Given the sizes of the negatives and the resulting format diagonal, the inherent advantage of medium format was a factor of about 1.6 to 1.9 at larger print sizes and zero at smaller print sizes, when using the same ISO speed film. Or the same quality at the same final enlargement size was possible with ISO 400 at medium size and ISO100 at 35mm size. Comparing the 2:3 size of the 35mm format with the square format of 120 roll film cannot be done at full size. So the roll film format has to be reduced to the 2:3 format of 35mm. The basic 58 x 58mm has to be reduced to 38.7 x 58mm with a diagonal of 69.7mm. Then the relationship is 1.7. It so happens that the diagonal of the M8 sensor and the H3D sensor are related 1:1.9 or identical as the original miniature and medium format sizes (82 to 43.3). The M8 sensor size has a relation of 2:3 (1.5) and the H3D has a relation of 3:4 (1.33).
24 x 36mm has a diagonal of 43.3mm
45 x 55 mm has a diagonal of 71.1mm
45 x 58mm has a diagonal of 73.4mm
56 x 56mm has a diagonal of 79.2mm
58 x 58mm has a diagonal of 82.0mm
36.7 x 49mm (H3D) has a diagonal of 61.2mm
18 x 27mm (M8) has a diagonal of 32.5mm
One should not analyse these figures to death: photographic processes do not operate with mathematical precision. Now in the digital domain, we do not have the simple basic position that the emulsion is equal. We have different sensors and different algorithms. With the M8 and the H3D we have roughly the same relationship as with silver halide emulsions.
The M8 does not possess a full 35mm area sensor, but has in effect a half-frame sensor, with a 1.3 reduction compared to the Barnack negative size. I could test recently a new HasselbladH3D with 39 Million sensor, that has a 1.1 reduction compared to the original 6 x 4.5 cm format. The Leica sensor has 10 Million pixels and the Hasselblad has 40 Million pixels. How do they compare and what differences do we see?
I do not intend to review the H3d (the review will be published elsewhere). Here I will focus on the differences in image quality. But a small digression may be of interest. The main and dominant shape nowadays in cameras is the Canon pioneered EOS shape from about 1985. Most companies have followed that lead and by now any DSLR that wants acceptance in the market has the same layout and ergonomics. The basic shape has been designed by Barnack in 1913 and the Leica M8 has all the main characteristics of that day, including the aerial view through a separate viewfinder.
The SLR Shape is of course dominated by the pentaprism that allows direct viewing through the taking lens. The other shape was pioneered by Victor Hasselblad in 1948 and is quite visible in the new H3D contours. The handling of the H3D is superior to everything I have seen and the viewfinder is simply stunningly good. When the shape of Barnacks design dominated the 20th century, the Hasselblad shape might define the 21st century. It is a beautiful camera to use in a studio, and even handheld. The main problem with a large sensor (producing 200Mb TIFF files) is the heat. Operating the camera, you hear a cooling fan that is almost constantly running, draining the battery power quite heavily. With current technology, we cannot expect a large sensor in a small body without cooling fans and this is maybe one of the reasons with small format DSLRs are limited to 16 Mb pixels.
Sensor sizes
The Hasselblad H3D 39 Meg sensor has dimensions of 5412 x 7212 pixels and the Leica M8 10 Meg sensor has dimensions of 2626 x 3936 pixels. The 4 times difference in size translates nicely into file sizes (30 Meg for M8 and 120 Meg for H3D).
Pixel size is with 6.8 micron the same for both. This value implies that the theoretical Nyquist frequency of maximum resolution should be the same too. It is, but factual resolution is different. The Siemens star pattern shows the real definition of the systems. M8 was fitted with new Summilux 1.4/50 asph and aperture was set to 5.6. H3D was fitted with the standard 80mm (Fuji) lens and aperture was set to 8. The M8 has somewhat less maximum resolution than H3D, but the quality difference is quite small, given the factor 2 advantage.
Leica left: Hasseblad right


As I remarked in my M8 report, it does not make sense to use the Nyquist limit as a practical reference. In practical use the factual resolution limit is 20% to 30% lower than the Nyquist limit. And contrast is not taken into account, when one just calculates maximum Nyquist frequencies. The same fallacies that plagued the silver halide debate, are now being used in the digital domain. The small advantage of the H3D can be attributed the larger sensor size. In fact we should consider that a certain image detail has to be recorded on the Leica on let us say 4 pixels, whereas the H3D can use 16 pixels. With this larger amount of capture area the interpolation is a much easier and more secure process.
Quality comparison
The flower picture used for comparison purposes is quite interesting. It clearly shows the extended depth of field of the Leica system. overall image (Hasselblad and Leica)

Quite often this is an an advantage (you do not have to focus carefully and it was the main incentive around 1930 to buy the camera ), it sometimes becomes a bit difficult to work with. A wider aperture will do the trick of course, but then the depth of field for depicting the main subject with sufficient depth, becomes quite narrow. The H3D picture shows more detail in the ribbon and especially in the reflections of the metal and glass. When we blow up the Leica picture to 200% (to get the same picture size), we note the lack of detail that cannot be reproduced by the Leica system, but is easily captured by the Hasselblad system.

Left: Leica 100%; right: Hasselblad 100% (small section!! to show the size difference). Leica image at 200%



Color reproduction
Color reproduction (by the way) is quite good when one compares both systems and is also an indication that post processing is a powerful tool for correcting whatever defects can be found in the Raw negative.
Conclusion
The main differences between the Leica picture and the Hasselblad picture are related to definition of fine detail and reproduction of textural structures. One can also note in the luminance gradation in the white background that the Hasselblad picture has more potential for modulation of gradation nuances. This is of course the same conclusion as we got from the silver halide comparison. In the domain of definition of detail, the Leica M8 has to retreat a bit when comparing the H3D imagery.
It is evident then that the H3D (or for that matter any large sensor system) is required stuff when you need to capture every detail and gradation nuance of the subject and want to blow up the print to A2 format without interpolation losses. In this respect, the H3D performs flawlessly. The cost of the system (buying price, capture costs etc) is beyond normal means in most cases.
The M8 cannot match the H3D quality in absolute terms, but offers surprisingly good quality when one compares the sizes of the sensor. It is difficult to quantify the quality differences. The sensor sizes differ by a factor of two, but I would suggest that the quality difference in imagery is a factor of 1.2. Quite relevant for the person who wants the ultimate in performance.
The Leica engineers can be satisfied with this result. The Leica image is not as good as the Hasselblad one, but the difference is quite small and especially in the prime area of the Leica (dynamic hand held shooting) probably not of supreme importance.
The advantage of the 39 Meg pixel area compared with the 10 Meg area is less great that one could expect: here we note that pixel count is not of paramount importance , just as resolution was in the silver halide domain. 
 M8 part 7: results with the IR blocking filter
Background
Over a long period of time, Leica could be considered as a company that was engineering driven: high quality products, made with the obsession of engineers who wanted to realize the best possible equipment. Customer wishes were not that relevant. Engineers are no marketers, but they possess one enviable characteristic: they are honest about the products they design and realize. Information from the Leica company could be relied upon and that was also the time that being proud about your results was not mixed up with over-hyping your product. This latter attitude might be necessary in today's overcrowded camera scene to be heard and seen, but it is a far cry from the previous single minded approach.
Nowadays Leica has shifted in the direction of a market driven company: that is good for the customer who has an eager ear, belonging to market people who need customer exposure in an world where product cycles are short and the proliferation of products is beyond imagination.
The M8 is a camera definitely designed from an engineering view: the engineers will tell you that they designed a product with a certain goal (maximum image quality at the beginning of the image manipulation cycle), that this goal could only be accomplished within certain fixed parameters (the Kodak sensor characteristics, the Leica lens characteristics, production techniques and tolerance levels) and that therefore some characteristics of the product can be optimized, while others had to be compromised: the choice where to optimize and where to compromise was made in the context of the stated goal.
Marketing people do not want to hear that: they want to tell the buyers that the product is superior to anything else on the market and is a no-compromise product that will suit any buyer.
And in the case of the M8 we may also point to the fact that long before the product was constructed the Leica internet community, in the best tradition of swarm intelligence, assumed that any new Leica product designed for the digital workflow, being a direct offspring from the redoubtable M series of cameras would be as superior in the sensor-based world as it is in the film-based world. In the film based world Leica could rely on the emulsion makers to provide the best possible films and they could also assume that the emulsion makers would cater for the customer wishes providing a broad range of films from Kodachrome to Velvia and from Techpan to Delta 3200. But in the sensor world Leica has to rely on their own expertise to develop software that will interpret the raw sensor data into a colour space that will accurately reflect the real world colours. The Leica M8 colour reproduction is not truly accurate as several analyses with colour charts will show. Again this reproduction behaviour is part of the equation as outlined above. (Colour film by the way has never been accurate: consumers do not buy these films as manufacturers who tried to market such a film have discovered)
The IR blocking filter and ist use
High sensitivity to Infrared radiation and a higher sensitivity to the red part of the spectrum are characteristics of the M8 that influence the colour reproduction.
The natural inclination of marketing people is to play down the issue (it occurs incidentally, only in extreme situations) but then the engineering spirit of the company took over and the IR blocking filter is the proposed answer to the IR problem. The filters are now being distributed and some publications claim that this is an excellent solution and the recommendation is to use the filter as a standard on every lens fitted to the M8.
My testing leads to a different conclusion. To start with, we should be aware that any filter will have adverse effects on the image and its quality. The two main points that are inherent in every filter are reduction of contrast of the image, sometimes severe, and additional reflections from specular highlights and strong light sources.
The test results below show that the filter does create quite visible reflections: the filter does not only cause this but also by the high reflection values of the sensor surface. Right: with filter; Left:  without filter. 
The MTF graph does show that the filter does not reduce the final resolution, nor are the main low frequencies influenced, but the critical medium and high frequencies suffer a loss of contrast. One may say that the post processing software can counter this and for some part this is indeed the case, but an artificial contrast boost has its own problems. Blue line is without filter, red line is with filter; resolution rage is from 100 lp/image height to 100 lp/image height. 


I am still working (presumably very old fashioned) from the film based approach where the best quality of the image at the start of the process must be preserved at all cost. That is one of the main arguments for using a Leica anyway: the best optical quality is delivered in front of the camera. In fact I would prefer to go directly from the Raw image to the printing software (Q-image) to emulate the classical darkroom process as much as possible. As less manipulation as possible is the best method to get celan and honest pictures in let us say the tradition of a master photographer like Andreas Feininger.
The IR sensitivity for IR-radiating fibres and synthetics is reduced to a very low level, as the shot of the model indicates. It is not fully gone however and the red sensitivity is also not reduced. The skin tones are still on the verge of reddish. The working of the IR blocking filter has been restricted to the IR range and there is a very sharp cut off in the transmission characteristic: only the IR is blocked and the rest of the spectrum is not affected. Below: with filter; bottom: without filter.

The Macbeth analysis shows no significant difference between pictures with and without filter. Note the high noise in the red layer. Above: with filter; below:  without filter.

Conclusion
The IR blocking filter then is a good solution for pictures where the fibres and materials do emit IR radiation. For all other pictures one should not add the filter to the lens: it does not improve the colour rendition and it only adds the classical filter effects of contrast reduction and additional reflections. Using the filter habitually on the lens is not the best option: one has to very careful to look at potential strong light sources. If you have time, you should take a picture without filter: if this is OK, proceed, if not add the filter. When you print in black and white (and the classical oriented Leica M photographer should at least do this frequently) the IR problem is a non-issue. And the filters are in this case a slight help in some situations. 
 M8 part 8: Rangefinder accuracy
Background
Rangefinder accuracy in the sensor based camera
Rangefinder accuracy is the result of several factors: the mechanical accuracy of the camera/lens coupling in combination with the rangefinder unit alignment on the camera, the focus shift of the lens in question when stopping down, the accuracy of the location of the focal plane (film gate or sensor surface) related to the bayonet location and not to forget the quality of the eye of the photographer.
The mechanical accuracy of the lens/camera coupling is the result of the match of the shape of the roller and the shape/adjustment of the roller arm (prism arm) and the shape/steepness of the lens cam. Location of film gate and bayonet and the alignment of the rangefinder unit on the body are also accurately fixed. The accuracy is needed over a large range of distance settings from 0.7 meters to infinity and some machined corrections are necessary to guarantee the accuracy over the focus range, the close range being especially critical. Leica therefore adjusts the rangefinder settings at 1meter and 10 meters.
With film based cameras the adjustments can be done within an accuracy of several hundreds of a millimetre (0.02 mm is cited as film gate tolerance). This is very accurate given the fact that the film itself has a certain depth of emulsion (normally around 20 micron) and has always a certain curvature itself. The camera and lens designers assume a certain location were the optical focal plane of best performance will be located. That is the position where the focus plane will be located when the rangefinder spot is accurately aligned. Some latitude is acceptable here because of the emulsion thickness and the film curvature. Both phenomena will ensure that the focus plane will be located inside the emulsion layer and will generate a sharp image.
Film based cameras
Assuming that the mechanical parts are accurate, we can be sure that the plane of sharp focus is there where the designers want it to be. But we know that the any lens has some degree of spherical aberration (SA). This implies that we do not have a fixed point of sharp focus, but we have some latitude in locating the image plane. SA will force rays form the outer parts of the lens to focus at a different location than the central rays do and we will have the situation where rays from a point will at first converge to a small circle and then diverge again. This small circle is of course related to the circle of least confusion. We know that the smallest circle will consist of a small patch of concentrated light with a larger halo of diminishing intensity around it. Here we find the best resolution. But just before and after this location we find a circle of light that is a bit larger but has less halo around it. Here we find a pont with maximum contrast and somewhat lower resolution. The designer of the lens has to decide where to locate the focal plane and thus the image plane of best performance. But wherever that location is, it is designed for the wide open performance of the lens. When stopping down the rays at the edge of the lens are cut off and the focal plane starts to shift: this is the phenomenon of focus shift that I discussed for the first time when assessing the Noctilux performance. It has been the first time that attention was drawn to the phenomenon of focus shift. In practice that implies that you focus for wide open performance but get the location of the sharpness plane for a smaller aperture. For critical work that can be unpleasant.
These lens characteristics cannot be changed. Zeiss has claimed that their lenses are designed to exhibit less focus shift than is usual, but part of the claim is based on a different location of the focal plane: a kind of average position so to speak. Not as good for best wide open performance, but better for stopped down performance. Leica has adopted the opposite approach: optimize for wide open performance and let the focus shift be compensated by depth of field.
Sensor based cameras
The main characteristic of the sensor surface is its absolutely flat surface: there is no depth like we have in film emulsions. Here we cannot count on emulsion thickness to compensate for mechanical errors in accuracy or focus shift.
Leica does know this of course. And the tolerance level of the whole rangefinder adjustment chain has been narrowed accordingly. The tolerance level in film based cameras is a few hundreds of a millimetre, in the M8 that level has been reduced to a few thousands of a millimetre. That does not imply that the factual tolerance level is a factor ten narrower! In reality the factor is about three to four times. This makes the M8 the most accurately machined and assembled M camera in history. This is done to compensate for the lack of image capture thickness.
When focussed accurately the focus plane is spot on on the sensor surface.
The other side of the medal is the fact that focus shift might be more noticeable when stopping down: another argument not to stop down too much when you need critical sharpness at the focus point.
We should also realize that any mechanical part and any manual adjustment has a small level of tolerance that may cancel out (good) or add up (bad). So some misalignment of the rangefinder mechanism should be kept in mind. You cannot work at zero tolerance. Here we run against the limits of mechanical precision equipment.
The test
To find out how narrow the tolerances are, I conducted the following test. I took a picture of the Siemens star target at 2.5 meters distance and focused with the magnifier on the vertical black bar of the chart. With this bar we can optimize the accuracy of the vernier alignment. The actual thickness of the bar is one millimeter.
Then I did a series of pictures of pictures wth focus bracketing to ensure that this is indeed the plane of best sharpness. Then I deliberately shifted the focus by the width of one millimetre (or thickness of the bar) and took again a series of pictures with focus bracketing. The result you see below. The misaligned picture has reduced contrast and reduced resolution. All pictures wide open to evade focus shift problems. Left: in focus; right out of focus by one millimeter.
The small shift in focus inaccuracy is really small: at this distance of 2.5 meter the shift by one millimetre amounts to a focus displacement of 1/43 of a degree or roughly one second of arc. The whole movement of the lens covers three degrees and this small misalignment will happen almost always: you are in a hurry; you cannot focus accurately because of handholding movement or because of eye fatigue etc.
When you encounter unsharpness where you expect a sharp image plane you may have to think about operator problems first (eye fatigue, hand movement), then focus shift (when stopping down) and then about possible mechanical tolerance issues. A good test is to do focus bracketing (cheap because digital files do not cost a cent) and to repeat the test several times over a week to see whether there are operator induced errors.

 M8 part 9: UV/IR filter effect of 6-bit coding
Introduction
There is still a certain amount of misinformation around the use and effects of the new IR filter. Specifically the topic of image quality related to the use of the IR filter in combination with the new software update 1.102 and lenses with the 6-bit coding is creating some confusion.
My first tests were done with the older firmware versions and with lenses without 6-bit coding and it is natural that some questions were raised about the validity of this test. My tests were conducted with the 50mm and 75mm lenses and with theses focal lengths some of the specific problems of extreme wide angle lenses do not occur. Basically I was most interested in the claim that the overall quality of the images would be improved.
The test showed that the UV blocking filter did what the name suggested: it blocked the UV radiation, but did not have any effect on the rest of the spectral colours.
IR Filter revisited
I have now a lens with the 6-bit coding and the cuurent firmware update. So I did a new test, now comparing the three options: lens detection off, lens detection on and lens detection with UV IR filter on. In every case I took pictures with and without the filter of a Macbeth chart.
The WB was automatic as was the exposure. The image files were Raw converted with Lightroom and here all parameters were set to zero. Color temperature was 5900 and Tint was +9.
Below you see the results as reported by Imatest. Left column is without filter, right column with filter. First row is detection off, second row is detection on and third row is detection on and  IR filter.

The values as presented by Imatest in the charts differ by a few percentage points, but are within statistical error range and certainly within the latitude of he human visual system. Note also that differences in chroma are presumed to be linear, which is not the case. Compare below the two extreme positions: detection off and no filter, with detection on and IR filter. The latter has more noise and less contrast as can be expected.

Upshot
The conclusion as reported in the previous test does not have to be altered. At least for all lenses above 24mm (and possibly including that focal length too) the firmware update and the IR filter in combination with the 6-bit coding do not improve on the general image quality that can be delivered by the basic combination of the lens without coding and the attachment of the IR filter. The firmware update may be required stuff for extreme wide angles but not for the more moderate angles of view, at least in relation to the color reproduction of the camera sensor. The IR filter has its obvious merits but introduces some flare, additional noise and lower contrast.