[Klaus]

http://www.opticallimits.com/sony_nex/99...18105f4oss

 

Not so bad ... other than distortions.

[Guest]

Great work! Looks like the lens's popularity is justified. It's like that pincushion distortion abnormally concentrates the center resolution. So how much exactly does the resolution go down with distortion correction? Any reason to choose the 16-70 Zeiss over this other than the wider wide end and smaller stowed footprint?

[Klaus]

Well, take a balloon with some text on it and stretch it. What happens to the text ? It gets bigger. Translated to lenses this means that the pixel are "stretched". Of course, pixels are discrete so their state is stretched into neighboring pixels causing a decrease in resolution. 

If you take an image at 105mm (at 6000x4000px), the left/right center loss accumulates to about 400px that have to be interpolated/stretched.

Note: I'm not sure whether this is a 'linear' loss or whether the relative loss increases the more you move to the borders. If it is not linear the interpolation effort rises accordingly (more blur).

 

Well, the Zeiss lens is wider. Personally I'm not a lover of that 18mm setting (vs 16mm). But as a package I would prefer the Sony G lens.

[Rover]

So... really the king of pincushion distortions... and not only in the longer half of the range. This is nucking futs. How do we call this "anti-fisheye"... the "frog-eye"? Actually, no, the "frog" theme is reserved to the Laowa brand, so I'm open to suggestions.

[Brightcolours]

So... really the king of pincushion distortions... and not only in the longer half of the range. This is nucking futs. How do we call this "anti-fisheye"... the "frog-eye"? Actually, no, the "frog" theme is reserved to the Laowa brand, so I'm open to suggestions.

 

I don't know if evolution has figured out concave eyes?

My suggestion: oomph lens (vs fisheye lens). oomph being the sound one makes when being hit in the gut, which does create a concave belly for a moment.

[stoppingdown]

"At the long end the image corners are soft but then how often do you require sharp corners at this setting "

 

Well, if you do landscape, it's important.

 

I understand the reason for which this lens is popular. Let's also recall that it's cheaper than the SEL1670Z. But, in the end, I'm happy I went with the latter, even though we know it's somewhat disappointing. The test by PZ is welcome, because it's possible to compare both lenses on the same methodology. In the past I saw comparisons based on images, and I could see that basically the two lenses are overapping - the wild sample variations probably can also change things. But I didn't like the idea of "losing" hundreds of pixels for the interpolation, as Klaus described. I'm not against IQ fixes in post-processing, it's a legitimate trade-off. For this lens, they went just too far. Since Sony advertises this lens mostly for videomakers, that interpolation might be fine for them.

 

At the end of the day, since these are the only APS-C lenses in this range, I think Sigma could be able to create some interesting competition here.

[chrismiller]

Typo many rather than may:

 

Some many not like the slight lag between turning the zoom/focus ring and...

 

thanks for the review.

 

[Klaus]

"At the long end the image corners are soft but then how often do you require sharp corners at this setting "

 

Well, if you do landscape, it's important.

 

 

 

Well, we are talking about 105mm there thus "160mm".  Never say never but for landscapes that's very long. At 70mm (105mm) the corners are better.

 

However, of course, it is not a stellar lens (2 1/2 stars).

[Brightcolours]

Many do shoot tele landscape, even 200mm on APS-C...

[toni-a]

Many do shoot tele landscape, even 200mm on APS-C...

I do 😊