I'm no defender of Canon L zooms, but those off-axis results look abysmal to the extent that the testing methodology needs clarification, if not a complete overhaul.
However...
A useful rating of lens sharpness might factor the ease by which that sharpness is achieved. braindeadmac alludes to this factor in his follow-up postings. By using a certain manual-focus technique of a less-than-excruciating nature, or simply relying on AF, the practical level of sharpness is then measured, and the best lens is declared. I consider this a valid (and useful) test, provided the body and lens are calibrated to factory standard.
For critical focus, we ignore the pain and tedium of the procedure, and go for the gold: every pixel etched to the optical limits of the glass - no excuses! The only way to achieve this result is bracket focussing, a large number of shots with manual-focus increments, followed by host-side pixel-peeping. Due to mechanical slop, remember to move the focus ring in one direction only:
rico wrote:
I'm no defender of Canon L zooms, but those off-axis results look abysmal to the extent that the testing methodology needs clarification, if not a complete overhaul.
However...
A useful rating of lens sharpness might factor the ease by which that sharpness is achieved. braindeadmac alludes to this factor in his follow-up postings. By using a certain manual-focus technique of a less-than-excruciating nature, or simply relying on AF, the practical level of sharpness is then measured, and the best lens is declared. I consider this a valid (and useful) test, provided the body and lens are calibrated to factory standard.
That's nicely articulated, rico. I'm the first to admit that the lens testing methodology I used is hardly one for the ages. But it seems pretty relevant to me that I was able to EASILY achieve critical focus with the Leica lens, but found it difficult with the Canon L zooms--particularly with manual focus.
braindeadmac wrote:
That's nicely articulated, rico. I'm the first to admit that the lens testing methodology I used is hardly one for the ages. But it seems pretty relevant to me that I was able to EASILY achieve critical focus with the Leica lens, but found it difficult with the Canon L zooms--particularly with manual focus.
It's rather unfortunate, but I don't think that the L zooms had manual focus in mind when they designed the mechanism. Some are better than others (the 100-400 is not very good - too sticky) but the AF reigns king over MF, and it seems to be added almost as a second thought. The L primes that I have are much smoother (the 300/4 seems to be rather easy to manually focus).
Of course, on the 1.6 cameras, the viewfinder presents an obstacle to manual focus as well.
Looks like camera shake to me on those Canon lens samples - not suprising a Leica prime beats Canon Zooms but I've never seen anything that bad from my 16-35.
It would not be surprising that a prime is sharper than a zoom however
Focus needs to be apples to apples
Most lens have field curvature. Some concave some convex. A zoom lens can vary in curvature by distance. A concave is better close. A convex better far. So you should choose your lens based on your predominant distance. Landscape convex or flat. Close on concave.
My point is the leica might be set up for closer in and the zoom might be at a worse point (requiring focus on middle off centre) and may just never catch up but be better at landscape (far).
Adequate testing requires identify preferred subject and varying focus spot from centre to edge and distance. Or just understand you are testing a narrow range.
In addition some lens are sharper and some have higher contrast. Some might be better at both varying by distance.
Anyone buying a zoom needs to understand its field curvature.
Old thread but interesting to read again. I'm looking now for a 28mm or 35mm Leica R to use with my 5DSr. I have the longer focal lengths covered with several other Leica R lenses.
Also considering adding a Leica SL to my kit as my old eyes are having trouble now seeing well enough in the viewfinder to manually focus accurately and quickly.
Liveview does work great when shooting slow and on a tripod though.
