Archived posting to the Leica Users Group, 2008/11/02
[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]If an analysis and explanation from one who is neither an engineer nor a Leica insider is worth reading, here goes. The reason has to do with the flange (or lens)-to-sensor distance. This distance is short in RF cameras, resulting in a relatively highly off-axis incidence for the light rays going from the exit pupil at the rear element of the lens to the edges of the sensor, especially for retrofocal or other WA lenses. SLRs, with the necessary space for the mirror box, have a much longer lens-to-sensor distance and correspondingly shallower angles for edge light rays. As far as the sensor itself goes, this is important because the photosites have difficulty with light coming in highly off-axis. A smaller sensor suffers less because its extremities are not at such a high angle. As far as the IR filter goes, it is important because the filter acts at least in part by means of optical interference, which depends sensitively on the distance (fractions of a wavelength) that the light travels between the front surface of the interference coating and the surface of the glass substrate. The specific thickness of the coating optimizes its effect over a small range of wavelengths, e.g IR, centered on an exact wavelength at perpendicular incidence. For the wavelength of interest, the thickness could be chosen to be optimal for perpendicular rays, in order to optimize the effect at the center of the image, where the center of visual interest usually lies. It could also be chosen to be optimal for slightly angled rays, in order to produce a zone of acceptable effect that extends from the center as far toward the periphery as possible. A significant deviation from this ideal angle increases the wavelength of maximal effect. (Which is why the color of reflections from the front surface of a lens--and the IR filter!--varies with the angle of viewing.) It's a complex engineering decision that depends on esthetic and technical considerations--how far off-axis the effect should extend, how suboptimal the on-axis effect can be allowed to be, how the IR sensitivity of the sensor varies with wavelength, and so on. The angles are less extreme in front of the lens, which is why a filter placed there is feasible. (I would guess that even then, the effect at the extremes of an image produced with a WA lens would be noticeable. I'll run a test with the 21 and the 24 Elmarits as soon as I can find a piece of black polyester cloth to put in front of my M8....) In the case of the M8, these two factors underlie the sub-35mm sensor size and the absence of an IR filter at the sensor. (For SLRs, the angles are sufficiently less extreme that both a full-size sensor and a filter directly in front of the sensor are workable. This also applies to MF reflex cameras, of course. Leica presumably made what it deemed the best of a situation that necessitated compromise. We live with it, or not, according to our own priorities. --howard On Nov 2, 2008, at 9:41 AM, Frank Filippone wrote: > Can you point to a technical discussion of why this is so? I have > only heard the Leica point of view. > > It seems to me that if there were a good technical reason., others > would fool the path... Hasselblad, and third party esoteric camera > backs for MF and larger studio camera included.... > > But Leica, in the M8 only, is the only believer....... > > > Frank Filippone > red735i@earthlink.net > > > There is a perfectly sound technical reason for it to be better for > the IR filter to be on "the other side" of the lens with rangefinder > lenses. > > > > > > _______________________________________________ > Leica Users Group. > See http://leica-users.org/mailman/listinfo/lug for more information