Voigtlander Nokton 25 mm f/0.95
4. Image resolution
At the very beginning it’s worth reminding here that in the case of tests based on the E-PL1 sensor - densely packed with pixels and devoid of any AA filter - the decency level is positioned near 44-45 lpmm and the best “primes” can exceed the value of 80 lpmm. If you take it into account you can easily notice that the image the Nokton gives us by f/0.95 is not fully useful. We shouldn’t deceive ourselves and expect it to be so, though. Lenses with such an extreme aperture value are rarely useful when wide open – the maximum aperture is so huge that the full usefulness can be obtained only after a slight stopping down. If we recall the results of most system 1.4/50 or 1.8/50 class lenses it turns out they weren’t useful at the maximum relative aperture either.
On stopping down the image quality improves very quickly so by f/1.4 we can enjoy a fully sharp image reaching the level of almost 53 lpmm. The rest of performance can be called outstanding without any doubts. It is the effect of a good optical construction and record-breaking fastness. Almost all fast lenses reach the peak of their possibilities on stopping down to about 3 EV. Such stopping down already marginalizes optical aberrations but at the same time we are still in a safe distance from significant limitations imposed by diffraction. As a result, fast f/1.4 fast lenses achieve maximum MTF values by the aperture near f/4.0. The Nokton, being over 1 EV faster, can reach maximum MTFs not by f/4.0 but near f/2.8 so in a place where diffraction disturbs us even less. Small wonder it is the perfect recipe for resolution records – and we can certainly observe them here. By f/2.8 we get the result of over 82 lpmm – the level unattainable for slower constructions. It’s enough to mention here than the Panasonic 1.7/20, so praised by us, its maximum result of 75 lpmm reached by f/4.0. Its noticeably worse performance stems from two factors: a slower aperture and a specific, full for compromises, “pancake” construction.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
When we consult the results on the frame edge they don’t seem sensational at first glance. The image is useful neither by f/0.95 nor by f/1.4 and it becomes so only near f/2.0. Such a way of assessing the performance is false, though. In this case the maximum achievements for the frame edge have been the best of all Micro 4/3 lenses, tested by us so far. What’s more, if we compare the results on the edge of full frame, reached by 50 mm system lenses, we’ll see that most of them had to be stopped down to almost f/4.0 to give fully useful images. Compared to that the results of the Voigtlander present themselves quite differently – there is really nothing to complain about.
The crops below traditionally show us our test chart’s centre. They were taken from JPEG files saved along RAW files – the basis for the analysis posted above. As always we used the JPEG files taken with the lowest sharpening level available. You must remember, though, that in the case of the E-PL1 that level is still pretty high.