Fuji X-T2 announced

[Anthony]

Photo Ninja, which I use, does not read Fuji lens correction parameters.

I use PT Lens for lens correction.

PT Lens has determined that the Fuji 14mm does not require any correction.  http://epaperpress.com/ptlens/profileInt.html#FujinonLenses  So it seems likely that this lens at least does not have lens correction parameters in the raw file.

[bjornthun]

That Fuji 14/2.8 must be a very fine lens. Very few lenses get the note "no correction needed". The Olympus 75/1.8 and the Sony Zeiss 135/1.8 are in the same distinguished club.

Most double Gauss normal lenses will have at most 1-2% distortion, many less than that. Double Gauss lenses, or Planar in Zeiss speak for DSLRs have a front group that is more negative than the rear group in order to increase the back focal distance to allow for the mirror box.

[bjornthun]

Photozone use jpeg files from D3x with a flat profile to get vignetting data for Nikon mount lenses. They go on to say that Canon files show 40% more vignetting due to a more agressive profile in their Canon test camera. One must wonder... 

[Les Olson]

Let's compare these figures to two Nikon mount 20/21mm lenses. The Nikon AF Nikkor 20/2.8 has a vignetting of 2.26 EV at f/2.8 and the Zeiss ZF.2 21/2.8 has a vignetting of 1.99 EV at f/2.8 which is less than 0.5 EV difference from the Fuji 14/2.8 at f/2.8. It is well known that all lenses will vignette less as you stop them down, so comparing a lens that starts at f/1.8 to one that starts at f/2.8 may be a skewed comparison.

Let's consider distortion, and compare the Zeiss Loxia 21/2.8 for mirrorless Sony and the Zeiss Milvus/ZF.2/ZE 21/2.8 for DSLRs. Both are retrofocus Distagons.

Zeiss Loxia 21/2.8: http://www.zeiss.de/content/dam/Photography/new/pdf/en/downloadcenter/datasheets_loxia/loxia_2821.pdf

Zeiss Milvus/ZF.2/ZE 21/2.8: http://www.zeiss.de/content/dam/Photography/new/pdf/en/downloadcenter/datasheets_milvus/milvus2821.pdf

From the graphs for distortion, we see both lenses have very similar, about 2% barrel distortion without any digital correction applied.

The graph for vignetting indicates approx. -2 EV for the Milvus/ZE/ZF.2 (DSLR) lens. The graph for the Loxia (mirrorless) indicates less, but I think that one is digitally corrected.

Both the mirrorless and the DSLR Zeiss 21/2.8 show very similar quality MTF graphs, though the mirrorless lens has a more gentle curve drop off toward the corners, where the DSLR lens drops more abruptly near the edge.

The takeaway for Zeiss is that distortion is comparable for the DSLR and mirrorless lenses.For the Fuji 14/2.8 and corresponding full frame DSLR Nikon 20/2.8 and DSLR Zeiss 21/2.8 the vignetting is actually comparable rangin from 1.99 EV to 2.4 EV. My guess for the Zeiss Loxia 21/2.8 is an uncorrected vignetting between 2-2.5 EV.

In summary it appears that little, if anything is given up with mirrorless wide angles. In the case of Zeiss, the mirrorless Loxia 21/2.8 weighs 398 grams, about half of it's DSLR sibling, so some weight is lost.

It is worth noting for other people that Zeiss' data sheets give data for radial distortion, while Imatest and other test systems use TV distortion, which is a related (and generally better) measure: Zeiss' 2% is not the same as the 2% you see in lens reviews. 

I didn't say that high levels of distortion were a consequence of the mirrorless design: peripheral light fall-off is a consequence of the mirrorless design and allowing barrel distortion mitigates peripheral light fall-off, and it appears that mirrorless manufacturers have taken advantage of that in some lenses.  So the fact that some mirrorless lenses do not have more distortion than their SLR equivalents is irrelevant. 

Zeiss lenses consistently (notoriously, one might say) have a lot of light fall-off, so the fact that a mirrorless Zeiss and an SLR Zeiss both have a lot of light fall-off has more to do with Zeiss than with the mirrorless design.  And, as you point out, we don't actually know what the uncorrected light fall-off is for the Loxia because they don't tell us.  Why do you suppose that is?  The Nikon 20/1.8 has 1.2 stops of light fall-off at f/2.8 - and it weighs 355g.  It is not stopping down that clears light fall-off, it is the relative aperture: the fact that the Nikon goes to f/1.8 is irrelevant. 

[pluton]

According to Photozone the Fuji 14/2.8 has 0.4% measured distortion and the 23/1.4 less than 0.3%, in both cases the same in RAW and JPEG.  Those are incredibly (literally) low values.  There has to be a suspicion that they are correcting the distortion in RAW. 

Les, Your suspicion is justified based on all known principles.  I questioned the results I was seeing also.  However, I can report that with both my 14mm/2.8 and 23mm/1.4 Fuji lenses, I have mounted them to the camera(Fuji XE1) with plastic tape inserted between the lens and camera electronic contacts, so the camera had no way of 'knowing' what lens was attached. In the Fuji menu system it's called "Shoot Without Lens". It's the same setting as when one shoots with a 'dumb' 3rd party adapter that has no communication with the body.
Results:  Same apparent lack of distortion.
Results from the Fuji 14mm are reminiscent of...although not the same as... the clean, distortion-free images as one got with the a Leica M camera and the 21mm/3.4 Super-Angulon. 

[Bjørn Rørslett]

Even with a "dumb" lens the camera might do some trickery on its own ... For example, a Nikon has the same ability to reduce CA for in-camera generated jpgs irrespective of the lens being known to it or not.

[bjornthun]

It is worth noting for other people that Zeiss' data sheets give data for radial distortion, while Imatest and other test systems use TV distortion, which is a related (and generally better) measure: Zeiss' 2% is not the same as the 2% you see in lens reviews. 

I didn't say that high levels of distortion were a consequence of the mirrorless design: peripheral light fall-off is a consequence of the mirrorless design and allowing barrel distortion mitigates peripheral light fall-off, and it appears that mirrorless manufacturers have taken advantage of that in some lenses.  So the fact that some mirrorless lenses do not have more distortion than their SLR equivalents is irrelevant. 

Zeiss lenses consistently (notoriously, one might say) have a lot of light fall-off, so the fact that a mirrorless Zeiss and an SLR Zeiss both have a lot of light fall-off has more to do with Zeiss than with the mirrorless design.  And, as you point out, we don't actually know what the uncorrected light fall-off is for the Loxia because they don't tell us.  Why do you suppose that is?  The Nikon 20/1.8 has 1.2 stops of light fall-off at f/2.8 - and it weighs 355g.  It is not stopping down that clears light fall-off, it is the relative aperture: the fact that the Nikon goes to f/1.8 is irrelevant.

I know perfectly well that there are different meassures of distorion, which is why I stuck to Zeiss's data sheets for comparing Zeiss lenses.

Btw. relying on photozone.de, it possible to obtain two values for vignetting for the same lens at the same aperture by testing it on different cameras. Check out Zeiss 21/2.8 on Canon and Nikon respectively. Which do you believe?

This test of the Zeiss Loxia 21/2.8: http://phillipreeve.net/blog/rolling-review-zeiss-loxia-distagon-2-821mm-t/ , shows real vignetting at f/2.8 to be at 2.5 EV, and I guessed at an interval of 2-2.5 EV. Zeiss knows perfectly well that data for uncorrected vignetting can be extracted by just taping over the electrical contacts. Thus speculation about why they didn't publish it are futile. The aperture of Loxia lenses is mechanical, like Leica M lenses.

The Loxia 21/2.8 has a 52mm filter thread compared to 77mm for the Nikon 20/1.8. The Nikon lens is a modern AF design, and the Zeiss lenses are both old school MF lenses, which is why I compared the Zeiss lenses at the outset. Fewer variabels.

[bjornthun]

Even with a "dumb" lens the camera might do some trickery on its own ... For example, a Nikon has the same ability to reduce CA for in-camera generated jpgs irrespective of the lens being known to it or not.

Correction of CA is my happiest memory from Nikon Capture (NX?). It was the first software I had, that could do it.

I think in camera distortion correction is impossible without knowing the lens, but I wonder more about vignetting.

[Eddie Draaisma]

The barrel distortion of the XF 35/2 WR can easily be observed in PhotoNinja, which as already mentioned, does not autocorrect distortion. I would never call such an amount of distortion insignificant. Both the 23/1.4 and the 14/2.8 do not show any significant distortion in PhotoNinja at all.

In other words, the low distortion figures for both the 14 and the 23 are most probably the true optical distortion figures.

[Akira]

I figured that the distortion data published on the popular online reviews are generally based on the test charts hanging just a few meters from the camera/lens combo.  So, the lenses tend to show more distortion than focused at distances.  The condition is more disadvantageous for faster lenses which tend to show the strong barrel distortion when focused closely.

As for the retro, design, I don't necessarily dislike the dials.  But I don't like the hardware design in which the lesser used dial keep taking up the space (like the shutter speed dial in A or P modes).  Also, I don't like the fact that the increment step cannot be changed.  Who would still need to set the ISO "manually" by 1/3 steps?  (I want Nikon to provide a whole step option in the menu on all models.)  Also, I always feel the turning of the dials while pushing the lock button awkward.  The command dials do require an extra hand or finger to push the desired function button, but its handling is much nicer than the dials of retro-dsign.

Also, I don't like the film camera design with the extra space on the left-hand side of the body.  That space was only necessary to accommodate the film patrone and makes no sense on a digital camera.  That extra space makes the holding of the camera feel awkward when switched the orientation of the camera between landscape and portrait.  I remember that even Bjørn, the Df advocate, complained about it.

[Bjørn Rørslett]

Nah. You got the wrong end of the stick, Akira. I positively prefer the "extra" space on the left, as it makes hand holding so much easier. The occasional trend of launching cameras chopped-off to the left (with the rationale as you explain), makes me cringe as handling becomes awkward and ignores the fact that cameras tend to be used by humans with two hands.

[Akira]

Nah. You got the wrong end of the stick, Akira. I positively prefer the "extra" space on the left, as it makes hand holding so much easier. The occasional trend of launching cameras chopped-off to the left (with the rationale as you explain), makes me cringe as handling becomes awkward and ignores the fact that cameras tend to be used by humans with two hands.

Bjørn, I remember you claimed the added left-hand side by attaching RRS L-bracket in your "Living and Working with Df" thread.  That wouldn't happen on a cameras of the "correct" DSLR design.

[Bjørn Rørslett]

No, that was *not* my point. You misread me and cannot derive the conclusion you put forth. The RRS factory L-bracket was too thick to allow my fingers access to the aperture ring whilst holding the camera correctly. This is simply an interface adjustment not a change of principle. Slimming the grip by 3-4 mm put the situation right.  For other bodies, such as the D600, this problem does not prevail although its RRS bracket is at least as thick, if not thicker, than the stock RRS for the Df. Each camera has to be evaluated on its own for handling ease. It is a 3-D shape.

[Akira]

Okay, I stand corrected in terms of the interpretation of your problem with Df and RRS bracket.

But my complaint about the extra space stays true (at least) to me.

[Bjørn Rørslett]

We agree to disagree. I want both hands to participate actively in the operation of the camera, the lens,  and their controls. I want to see easily how the camera is set up for a given purpose.