Pixiq

Part 1 of this article can be found here.

Camera Shake

There is no doubt that the most frequent cause of unsharp images is camera shake. None of the Leica M lenses has an image stabilization feature, and given the compact dimensions and robustness that are essential parts of the M system, it is unlikely to ever become a feature.

Although I carry the little table tripod for night photography only occasionally, in studio type situations, I do use a full-size tripod. The M is, however, primarily a handheld camera. Under normal circumstances, I think that using a tripod contradicts all the discreetness and immediacy that is the hallmark of the M style.

Many people are remarkably optimistic about their ability to hold a camera steady; even the oft-quoted rule of 1/focal length as the minimum handheld speed for sharp pictures is often too slow to ensure a consistently high number of really sharp images. My rule of thumb for minimum handheld speeds, to be absolutely certain of consistently sharp results with the M is as follows:

This is not to say that I will not use slower speeds if it is necessary. There have been many occasions when I have used speeds as slow as 1/8 or 1/15 second handheld for an unrepeatable opportunity. But I knew that I would be unlikely to get technically perfect images, and in these cases, I took every precaution I could to steady the camera. I accepted that picture content was more important than technical perfection.

An aspect related to camera shake is subject movement. It is surprising sometimes just how much unperceived motion there is in a subject. We all know perfectly well that when shooting action pictures of a racecar or aircraft we are likely to need speeds of 1/1000 or even 1/2000 second. However, even in an apparently tranquil landscape, trees, branches, and leaves are often in constant motion because of the wind, so those relatively close to the camera are easily blurred. With many subjects, the rangefinder style is to work close with a wide-angle lens, and the closer the subject, the more apparent is any movement.

There is another good reason for keeping shutter speeds high. Remember that you can focus very accurately with the M cameras and that Leica lenses give their very best performance at wider apertures, so there is no need to stop down unnecessarily. Unless you really need the maximum depth of field, 1/500 at f/5.6 will give you a better result than 1/60 at f/16! In any case, one of the beauties of the digital M is that, if necessary, you can increase the ISO speed very easily in order to handle a difficult situation.

Shoot DNG

If you are serious about your images, you will want to shoot them as RAW data. The version used by Leica with the digital M is the industry standard DNG (Digital Negative) format. Although JPEG images are a quick and easy route to prints, slide shows, or other needs, you sacrifice quality and control compared with DNG. An original DNG image retains everything that your camera captured on the sensor. It is your negative for you to process the way you want in order to maximize quality. A JPEG, on the other hand, has been subjected to significant in-camera processing and compression of the data that it contains. Also, every time you work with it and save it to a file on your computer, you lose more data and more quality.

Processing RAW images like DNG is no longer a black art. Working with Capture One, as supplied with the M8 models (or any of the other programs mentioned in Chapter 6) is very straightforward and well worth the effort.

In any case, the M8 cameras allow you to shoot DNG and JPEG together, so you can have the best of both worlds!

ISO Settings

The M8 models allow you a choice of five settings from ISO 160 to ISO 2500. As with film, the penalty for higher speed is a loss of quality. This manifests as lower detail resolution and noise in the image. Like grain in film, noise is best described as a lack of smoothness in large, evenly toned areas of the image (the sky, for instance) but is quite well masked in areas of considerable detail. With the M8 series, noise only becomes seriously apparent at ISO 1250 or higher, particularly at ISO 2500. However, even at these ISOs, detail resolution with Leica lenses is still very good. Although I always prefer to shoot at ISO 160 for maximum quality, ISO 640 is still very good and I don’t hesitate to use faster speeds, even going up to ISO 2500 if it is necessary. The quality at the highest speeds is generally better than anything available when using film of equivalent speeds. This kind of flexibility that comes along with using a quality digital camera is a real bonus!

The M8 models do have one neat trick to reduce noise in some circumstances. At shutter speeds below approximately 1/30 second, the camera takes a second black picture (with the shutter closed) and the noise in this parallel picture is subtracted from the data in the original. This takes a few seconds and for shutter speeds longer than 2 seconds, there is a warning message that it is happening. The camera should not be switched off while this is taking place.

Exposure

The dynamic range of digital sensors is more limited than that of film. Although much can be done in postprocessing to improve the tonal qualities of an image, there is no doubt that correct exposure is essential to realizing the exceptional quality of which the M8 cameras are capable.

In my experience, digital images require exposure accuracy at least equal to that needed with color transparency film. At best, the tolerance is plus or minus a half stop. It is particularly important not to overexpose and lose detail in the highlight areas. Working on an image later using the computer, it is often possible to lighten and retrieve detail from shadow areas, but if the highlights have been overexposed and all detail has been burned out, there is nothing there to retrieve.

For the great majority of situations, the metering system of the M8—whether in Auto or Manual mode—is very accurate. Nevertheless, unusual lighting or subject matter can fool any exposure meter, and it is important to recognize the situations where this can occur.

All exposure meters are calibrated on the basis that they are reading a combination of tones that equates to an average of 18% reflectance. If the average is lighter than this, the meter will indicate an exposure that will make them darker. If the average is darker, the meter will try to make them lighter. Thus, despite even the best efforts of the white balance, the meter may want to try to make pristine white snow into a grey mush. Alternatively, it will attempt to make a night scene look like day! Provided that you know just what area of the picture the meter is reading, you can compensate either by taking a reading of a more typical 18% tone and working with a manual mode setting or using the exposure lock in Auto mode.

With a digital camera, you do have an advantage over using a film camera. First, you can immediately look at the image and check whether or not it looks right and that there is reasonable detail in the shadows and/or the highlights. Secondly, you can look at the histogram. This will show whether the distribution of pixels is all bunched in either the shadows (left side) or the highlights (right side). You can also magnify selected areas of the image to check the histogram for just that area.

However, be careful not to allow the histogram to rule. Although a properly exposed image will normally have a good number of pixels showing in the middle two thirds of the graph, this will not necessarily be the case in more unusual lighting situations. In the two cases mentioned earlier, there will likely be a large proportion of pixels at the highlight end (snow) or the shadow end (night). A useful guide that shows if any highlight areas are completely burned out is that if the histogram is set to show RGB with clipping, it will show a red line at the right hand edge of the graph. The burned-out areas of the image will also be shown in red. You should be aware, however, that the in-camera histogram is a guide and no more. It is based on a JPEG preview and is not fully representative of the dynamic range of a RAW (DNG) file. At both the shadow and the highlight end, clipping will start to be indicated when there may still be recoverable detail once the DNG file is processed in the computer.

Although they are worthwhile indications, neither the image on the screen nor the tiny histogram should be relied upon totally. If time and subject matter permit, it is always worth bracketing exposures, or taking extra shots with plus and minus half- or even full-stop variations so that these can be assessed later on the larger screen of the computer. At the processing stage of a DNG image, it is often possible that some degree of recovery of detail that might be lurking in the highlights is practicable. Similarly, there may be more detail available in the shadows than first apparent. Nevertheless, as suggested above, later correction is no substitute for getting it right when the picture is taken.

There is just one option that I ought to mention, although it is rarely relevant to the M style as it applies to a static subject and requires a tripod. This is what is known as the HDR (High Dynamic Range) method for dealing with subjects with extremes of highlights and shadows. A series of images is made at different exposures in order to maximize detail individually in the shadows, mid-tones, and the highlights. These can then be merged into one single image with the full dynamic range using Photoshop or other dedicated software.