Upload & Sell: Off
| p.2 #13 · First Shots with the Otus - Bokeh Samples |
Could you talk a bit about these numbers, and how they will compare in large prints and such? I am not really good at interpreting numbers like this.
How do results compare visually? I mean, if you have shots of anything other than test charts, how would you describe the boke from these two lenses? I am considering making one of these lenses a replacement for a bunch of lenses I don't use much, but I am not sure which. Normally I would just go for the Zeiss, but this Nikkor has me intrigued. I would be especially interested in shots...Show more →
That's REALLY hard to tell... If you're comparing lenses that are somewhat similar in general rendering and "other" qualities, the numbers are accurate down to a certain degree. And certainly after the lens(es) have been stopped down by a full step.
At least with large-aperture lenses like the F1.4's, Imatest did at least USE to show very inflated numbers for lenses with low global contrast, wide open. So for lenses like the Canon/Nikon/Zeiss 50/1.4's, the F1.4 MTF50 values were much higher than what they "should" have been. That took me a while to catch on to, and to get some kind of analytical explanation for.
In Imatest master, I think you can now insert the "veiling flare" MTF baseline into the SFR measurement. This sets the "starting point" for the MTF - low frequencies like 2-5 lp/mm - lower. Then the lens will also reach the set MTF70/50/30 cutoff points earlier.
In some older lenses I found a 15% increase in Imatest indicated MTF50 compared to a real traversing slit measurement, wide open. That was lowered to a few percent (or nothing) after one or two stops down.
This will also upset quite a lot of people with lenses like the new Nikon 58. The MTF numbers extracted by software suits like Imatest or the simpler DxO packages just will not show the "correct" difference towards the more normal 50mm lenses. Weighing systems like the DxO "perceptual MP" - which I don't like for other reasons - or the SQF series in imatest - that include the baseline contrast compensation - will show other differences.
You really have to have both together (or alternatively, a real bench measured MTF!) to get the "whole picture".