I know others have reviewed it, but we got our first copies in today. I just finished testing them and the 25 f/2.0 does something I never thought I'd ever see: it had higher resolution than the fabled Zeiss 21mm f/2.8. (They were actually dead even when the 25mm was shot at 2.0, the 25 was easily better at f/2.8).
I'm asking Denoir if he'd do one of his fancy MTF overlays. I've only compared the MTFs at f4 for the 25/2 and 5.6 on the 21.
Can't find others.
The 25/2 @2 is remarably close to the 21/2.8 @2.8 - wide open they are amazing.
It seems to be a hair worse than the 21 when it's apeture is at 4 and the 21 is at 5.6 - that's to be expected. Look at the contrast
I'd still like to see more shots from this lens at distance at mid apertures. I know the f2.0 probably does some magical things close-in and wide open but for 'normal' landscape/cityscape shots at f5.6 focussed to infinity this lens doesn't seem to be worth the premium over the 25/2.8 which excels at infinity.
RCicala wrote:
I know others have reviewed it, but we got our first copies in today. I just finished testing them and the 25 f/2.0 does something I never thought I'd ever see: it had higher resolution than the fabled Zeiss 21mm f/2.8. (They were actually dead even when the 25mm was shot at 2.0, the 25 was easily better at f/2.8).
You would expect the 25/2 to be better up to 2.8 and then at f4 I would expect the 21 to pull ahead from then on.
And I assume you are talking about center resolution only, right? As we know from MTF's and user reports/pictures that the 25/2 is weak in the corners till f8 or f11.
wayne seltzer wrote:
As we know from MTF's and user reports/pictures that the 25/2 is weak in the corners till f8 or f11.
At longer distances- close to infinity focus - I believe it remains very weak in the corners even closed down...but I would love to see some examples which contradict this.
And is this weakness in the corners due only to field curvature and thus the plane of focus just shifts forward when you get near the corners or is it just not sharp period and field curvature is not evident?
wayne seltzer wrote:
I assume you are talking about center resolution only, right? As we know from MTF's and user reports/pictures that the 25/2 is weak in the corners till f8 or f11.
Center and overall resolution (weighted average). The absolute corners are a bit soft, but the edges not bad at all. Imatest MTF50 results (LP/IH, unsharpened raw images) for 9 copies tested today, along with 9 copies of the 21mm f/2.8 we tested as a control (since we know what values we expect with it). Left column is peak spot MTF50, right column weighted average (1.0 center 1/4; 0.75 middle 1/4; 0.5 outer 1/3). The bottom row, after the gap, is average of the column.
Absolute corner values ran about 400-450 for the 25mm (about the same corner numbers we see with the 21 and 35 f2), but that is at 15 feet, not infinity. I can't speak to infinity at all. But this is the first lens shorter than 85mm that has ever resolved as well as the 21mm in our tests. The 50mm f/2 is close. None of the 35mm are.
Wayne, we don't go that far (we're testing for decentering and tilt, which is most apparent wide open) but I'll put them back on the machine tomorrow and get some f4 and f5.6 numbers.
Wayne, we don't go that far (we're testing for decentering and tilt, which is most apparent wide open) but I'll put them back on the machine tomorrow and get some f4 and f5.6 numbers.
RCicala wrote:
Wayne, we don't go that far (we're testing for decentering and tilt, which is most apparent wide open) but I'll put them back on the machine tomorrow and get some f4 and f5.6 numbers.
Roger,
Oh ok, then only do f4,5.6 testing if you want to and if you have some spare time as I know you probably have better things to do.
Would be interesting what you measured for 24Lmk2 and especially my beloved 24G/1.4 which I first rented from you before buying it later. Maybe later sometime I will rent this 25/2 from you and test it out.
wayne seltzer wrote:
Would be interesting what you measured for 24Lmk2 and especially my beloved 24G/1.4 which I first rented from you before buying it later. Maybe later sometime I will rent this 25/2 from you and test it out.
Wayne, I've only got f/1.4 numbers handy on the two 24mms. The Canon Mk II measured the same way as above had average values of 620 peak (range 640-575) and 540 weighted average. The Nikon was 650 (range 690-605) and 560, again at f/1.4. I'll see if I can find the f/2.8 numbers for those.
My Zeiss dealer mentioned that all of the new Zeiss are aimed at high resolution performance in anticipation of future high resolution sensors. I don't recall the exact numbers he claimed, but 100MP sticks in my mind.
