Acceptable Level of Focus Shift on new 35 Lux FLE

[pgk]

Advertisement (gone after registration)

To insert a disruption - if film accommodates minor focus shift, why cannot digital modulate to achieve the same? Is the firmware  not smart enough? Is it time to consider engineering focus to the sensor in a more intelligent way?

 

If you integrate lens design and software correction you don't need to because you can probably build a flatter field lens and correct for other faults in software. But we are discussing M cameras here and as has been stated numerous times pure optical excellence, as demanded by Leica rangefinder cameras, has its disadvantages (as in extreme high cost and even then some residual problems). I think we have to live with it because the idea of Leica financing a specialist focusing sensor might be interesting but probably defies their economics.

[mp58]

I have had my 35 summilux asph non-FLE (bought new) since 1999. It is one of my favorite lenses, and i have never experienced focus shifting. After I dropped a camera that had this lens on it (saved by the Overgaard hood which took the brunt of the fall), the front element came loose and I sent the lens to DAG and had him work his magic. I asked him to optimize the lens for shooting at 1.4. Love this lens.

[henning]

Aberrations happen. Focus shift is mostly a function of under- and sometimes over-corrected spherical aberration. It is also usually dependent on the zone of focus, i.e., close, mid distance or far. Then, particularly with ever more complex lenses, the imaging field isn't flat, but wavers about, so that some things are in front and some things behind the plane of focus. Optical designers try their best, but designing a high speed lens for distances from near to infinity and for all colours of light while making the lens reasonably small and relatively affordable limits their choices at this time. Compromises are made by all companies. BTW, focus shift is almost all a designed parameter, and sample variation only plays a minor role. Film, with its depth and varying position in the 'film plane' was much more forgiving. Digital imaging does occur almost completely on one plane, in the same position for each shot so it's a lot less forgiving. Designing more 'give' into the digital sensor would just make higher resolution sensors pointless.

 

I've had my 35/1.4 ASPH since just after it was on the market, and I'm happy with it. No, it's not perfect and has some focus shift, but the additional dram of 'perfection' offered by the FLE, which is also not perfect, holds no interest for me. I still shoot with my 50/1 and 75/1.4, both of which have significant focus shift, but I'm used to them and can accommodate their eccentricities as well. To me, learning your lenses is more important than that they are 'perfect'.

 

Also, focus shift is a problem not only for rangefinder users, but definitely also for SLR and DSLR users, as in those cases focussing is done at maximum aperture, and when stopped down, problematic lenses will cause problems. At least rangefinder lenses can be adjusted for specific apertures, i.e., for an f/1.4 lens it could be adjusted for f/2.8 if that is what the lenses is used at much more than f/1.4. Yes, f/1.4 will then be out of focus unless compensated for, but f/2.8 will be right on. I like to have my lenses optimized for maximum aperture, as that is the most critical for me. For the 75/1.4, for example, I know that at f/2.8 at a distance of about 3m I have to move the camera back a certain amount after focussing to actually obtain best focus.