35/1.4 shootout take II

[MCTuomey]

thank you jon, enlightening and fun, appreciate you taking all this time and sharing results.

 

[bennyng]

Focus was not adjusted between shots. Only the aperture was adjusted. The camera was on a tripod and the shutter was fired using a cable release. I think the main problem is that the Nokton is a bit soft wide open which makes the focus point not as clear as it otherwise could be.

Perhaps these might help with the rest understanding the test parameters.

First, the focus alignment target was levelled to the ground.

Next, we level the camera horizontally. Then we align the focus pane (target) to the sensor pane.

With everything locked down, we attached the cable release and fired away, only changing lenses in between.

We could have arranged for the test to be done with the Sony A7s too. But timing was a bit tight. What a shame... However, I doubt the results will be any different from what we see here, except for the ISO performance.

Jon, perhaps we can repeat the test when I get the Zeiss 35mm f/1.4 ZM.

Cheers,

 

[ian_watts]

Comparisons of image center

@ F1.4


@ F2


@ F2.8


@ F4


@ F5.6

These are interesting and corroborate my own experience of the FLE that it starts to get less sharp at distance in the centre between F4 and F8. It isn't classic focus shift as such but a kind of ball of unsharpness that develops in the middle as the very curved plane of focus becomes more pronounced in the middling apertures (and focussed at distance). The FLE is a phenomenal lens in the close to mid distance range (and at any distance when opened up between F1.4 and F2.8) but can be exasperating when I've used it for landscape work. I've owned more than one copy of this lens and talked to Leica about it quite often and it appears to be a characteristic of the design which has prioritised wide-open, close-ish performance above overall consistency. It's a truly great lens (if a little large for my taste) but you have to know it's foibles to get the best from it.

 

[thompsonks]

Thank you, Jon. Love that pre-asph 'glow.'

Kirk

 

[mrmeadows]

That's a lot of tedious work that the OP has done for us in making these comparison images. Thanks for all that work.

I believe, however, that the Singapore comparison shots show a subject which actually only compares a fraction of the possible differences among these lenses. I know a bit of optics, but I'm no lens designer, so please correct my analysis if I'm wrong.

First, the "glow" of the Summilux pre-Asph is probably due at least as much to less effective older coatings as it is to lens design.

Second and more importantly, the lower right corner images are overwhelmed by astigmatism, so that all other aberrations are almost completely masked and cannot be reliably assessed. Thus these particular test images show primarily what to expect with these lenses when vastly overexposed light sources at the corner are so bright that the meridional rays forming the tangential focal line supply so much overexposure that little else can be seen. The sagittal focal line is much shorter and therefore is nearer the sensor plane. This would argue for under-corrected astigmatism in all four lenses, although the correction seems somewhat better for the two Asph lenses than for the other two.

My main point is the following: For more general photographs without overexposed light sources and for photos made at distances less than infinity, aberrations other than astigmatism are likely to actually be as important or more important in practice. Consequently, I would be very hesitant to assign too much weight to these particular tests.

My experience with a Summilux pre-Asph is that the corner aberrations are dominated by coma, not astigmatism. I can detect this aberration for the images of the Nokton and pre-Asph in the interior lights the top center images. Note that the size of the "glow" is much larger than the size of the coma which argues for coating deficiencies as its cause. No significant coma is present for the newer Leica lenses. I cannot be certain of coma in the lower right corner images, because the fluorescent lights in the building above the corner are so dim and small. I find it quite curious that the tangential lines of the Nokton are shaped more like gull wings than lines, and those of the Asph like an upside-down umbrella. Anybody know how that could be? The FLE-Asph

As the apertures are closed down, all of the lenses seem to show prominent diffraction spikes radiating from the light sources. The spikes are present in all the images, but are mostly masked by the astigmatism in the wider aperture shots. All of the sources seem to actually be multiple, since a count of the diffraction spikes is greater than the number of diaphragm blades.

--- Mike

 

[jonmanjiro]

Perhaps these might help with the rest understanding the test parameters.

Jon, perhaps we can repeat the test when I get the Zeiss 35mm f/1.4 ZM.

Cheers,

Thanks for posting those pics, Benny. When you get the Zeiss ZM 35/1.4, we're definitely gonna have to put it through its paces!

I believe, however, that the Singapore comparison shots show a subject which actually only compares a fraction of the possible differences among these lenses.

Mike, I agree completely.

First, the "glow" of the Summilux pre-Asph is probably due at least as much to less effective older coatings as it is to lens design.

I personally don't think the lens coatings are a much of a factor in this particular comparison. The Summilux Pre-ASPH I used is a very late Made in Germany lens (s/n is 350xxxx) manufactured around 1990, so it has the later improved coating compared to earlier Pre-ASPH Summiluxes. Incremental improvements to coatings were no doubt made since the early 1990s, but I don't think any technological leaps were made until recently when we started seeing Nano coatings from Nikon, and similar technology from other manufacturers. I don't know if Leica has adopted these latest coating technologies though.

My main point is the following: For more general photographs without overexposed light sources and for photos made at distances less than infinity, aberrations other than astigmatism are likely to actually be as important or more important in practice. Consequently, I would be very hesitant to assign too much weight to these particular tests.

Absolutely. The test shows just two scenarios among an infinite number of possible scenarios. Just a little more data in the public domain with which to compare these lenses.

 

[HenningW]

I don't have the FLE, nor have used it, but I have had the others for various lengths of time and these examples fully corroborate my experiences. My late pre-ASPH's behaved exactly like this one (my early one was softer and had these aberrations to a noticeably greater degree). My 35/1.4 CV was also like this one; in the style of the pre-ASPH, with lesser aberrations except pronounced focus shift which made the lens intolerable for me. I tried another copy and found the same focus shift. The 40/1.4 is a lot better lens in most respects, especially focus shift. It has some, but is manageable. The coma and astigmatism of the 40 are also a lot less.

BTW, the 35/1.2 vI is generally a very well behaved lens, with very good performance wide open, extremely little focus shift, fairly flat field and decent bokeh. The only thing is that the performance doesn't improve all that much on stopping down, so while it is almost competitive with the ASPH at f/1.4, it's not at f/5.6. It really is a specialist lens.