On Wed, Sep 25, 2013 at 11:03 PM, Godfrey DiGiorgi <[email protected]> wrote: > > On Sep 25, 2013, at 10:19 AM, Alexandru-Cristian Sarbu > <[email protected]> wrote: > >>> The M mount has a short registration, but the lenses designed for the M >>> mount were not originally designed for a digital sensor. That's the issue >>> there, not the short registration. The reason a short registration distance >>> is favorable to digital sensors is that you can design the lens to have >>> more elements behind the primary nodal point to help re-direct the light >>> path across the image field to intercept the sensor orthogonally. >> >> In theory, I agree - but for compatibility reasons I don't expect this >> to change. Does it mean the future lenses should be designed in the >> same manner? > > It means that, in the future, lenses should be designed with the light path > optimized for digital sensors in mind. This has no influence on their use > with film cameras, if such use is desired. If the manufacturer continues to > desire compatibility with older lenses for new bodies, it means that sensor > customization and optimization to include those older lenses must be > incorporated as well. Let me rephrase that: would a newly designed telecentric lens have issues, with offset microlenses? (it's the other way around) Of course, new sensors could make this a non-issue. > >>> This was not a consideration in designing lenses for use with film, and for >>> compactness reasons (amongst others) RF film camera lenses were designed >>> with very tight primary nodal point to imaging plane distances. SLR film >>> camera lenses had to clear the swinging mirror, which was the primary >>> reason for the deep mount registration and fostered lens designed that were >>> 'naturally' more akin to digital sensor lens designs. >> Precisely; it was a coincidence but the "old" lenses - the better ones >> - are working quite well for digital sensors, because of this. With >> the large focal range specific to a DSLR, using offset microlenses >> would not be an option. > > Offset micro-lenses is only one implementation available to optimize a sensor > for older lenses. It's not the one being used by Leica in the M Type 240, for > instance. Canon used offset micro-lenses in their first full-frame DSLR, but > it raised some interesting issues with specifically the shorter focal lengths > that they were trying to correct for. They've changed designs since then too. > >> The telecentricity Olympus loudly promoted was a natural >> characteristic of the classic SLR lenses. > > No, it isn't. SLR lenses are not necessarily telecentric designs. See > http://en.wikipedia.org/wiki/Telecentric_lens for a simple definition of a > telecentric lens design. You are, of course, correct. However, I wasn't talking about telecentric lenses in a strict sense (and AFAIK FourThirds lenses themselves are "near" telecentric designs). My point was... wait, you explained it nicely: > SLR lens designs have to clear the swinging mirror, which with short focal > lengths tends to mean using a retrofocus or inverted-telephoto design to > project the primary nodal point to the rear of the lens allowing the hard > bits to move forward out of the way of the mirror, rather than a symmetric > design which tends to have hard bits projecting deep into the camera as focal > length decreases. A side-effect of these SLR short-focal-length lens designs > is that they're more easily compatible with digital sensors because of this. > (For instance, the Pentax FA24/2 AL ... a terrific lens on film ... performed > poorly on digital sensors, showing both lateral and longitudinal chromatic > aberrations, issues with corner sharpness, etc. Other SLR lenses have done > better.) > > Many pre-digital SLR lenses do work very well indeed. Few to none work as > well as lenses which are optimized for digital sensors with respect to > measurable lens performance criteria. However, I often use older lenses > because they have pleasing aberrations, image rendering, and quirks that I > like—so what "works well" is a judgement call beyond simplistic performance > measurement. > >> I'm afraid it is; with no new FourThirds lenses since 2008, and the >> 2010's E-5 being replaced by the EM-1, as Olympus says. Maybe not >> "buried", but they're definitely phasing it out. >> Being able to adapt your lenses on another mount doesn't change this. > > I guess you want to debate the meaning of "dead and buried." Not really, I'll replace "dead and buried" with "being phased out" if that makes you more comfortable. No one likes to be told he's using "dead and buried" things... > > To me, dead and buried means the bodies that take a lens' mount are out of > production, and the lens cannot be used on any other in-production body, with > or without adapter, and provide the lens full functionality. Most dead and > buried lens mounts are also out of production, although there are some > exceptions. For example, Pentax M42 lenses are "dead and buried" as there are > no longer any bodies being manufactured that can be used with them and > provide the full features of the mount (auto-diaphragm, open aperture > metering, etc). Canon FL/FD lenses are dead and buried, as are Minolta MC/MD, > Leica R, Exacta, Konica AR, Beseler Topcon, and probably a host of others. Then, 4/3 lenses are "dead and buried" until they'll have a u4/3 body capable to focus them as fast as the E-5? :-p > > FourThirds SLR lenses, however, are a) still in production, and b) capable of > being used on other bodies with an OEM adapter for their full function. They > are fully supported by more than one manufacturer for parts and service. > People are still buying them to use on new bodies. I'd hardly call that "dead > and buried". They are in the process of being obsoleted, however, as their > replacements for the new, derivative lens mount are brought to market. But > obsolete isn't a criteria for dead and buried IMO. It doesn't really matters if they still making lenses or they're NOS, that's temporary; the mount is being phased out. > > G Alex
-- PDML Pentax-Discuss Mail List [email protected] http://pdml.net/mailman/listinfo/pdml_pdml.net to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.
