[Klaus]

https://www.sony-asia.com/electronics/ca.../selp1635g

---

Interestingly, the new one seems to be a MAGNITUDE better than the Zeiss 16-35mm f/4 OSS

New lens:



Zeiss:

[thxbb12]

The MTF are impressive, but what's the most impressive is the size of the lens.
It's about the same size and slightly lighter than the Fuji 10-24 f4!
I hate PZ lenses to a passion though.

Do you know what is the lens' native barrel distortion?... 10%! (source: https://www.lenstip.com/154.3-article-So...rties.html)
Optical distortion was purposedly under-designed in order to take advantage of software processing and thus be as compact as possible.
With this said, what truly matters is the final result and here, it beats "fully" corrected optical lenses such as the Zeiss you mention above.
Food for thoughts...

[Klaus]

Yeah, at 10% it's a good question how much quality remains after auto-correction.

[thxbb12]

From the DPR and lenstip samples it looks very good.
Also, aren't the MTF above after correction as well?
IMO whether one corrects aberration optically or by software doesn't matter. What matters it the final output.

[Klaus]

It is highly unlikely that the Sony MTFs are reflecting the corrected results.

[thxbb12]

It is highly unlikely that the Sony MTFs are reflecting the corrected results.

Why?

[Brightcolours]

It is highly unlikely that the Sony MTFs are reflecting the corrected results.

Why?

Because correction means that the corners need to stretch out a lot, losing resolution as a result.  MTFs like these simulate what the optics do, not what is done to images captured by the lens afterwards.

[thxbb12]

It is highly unlikely that the Sony MTFs are reflecting the corrected results.

Why?

Because correction means that the corners need to stretch out a lot, losing resolution as a result.  MTFs like these simulate what the optics do, not what is done to images captured by the lens afterwards.

I know what software distortion correction does to an image thank you.
Are you sure about this statement "MTFs like these simulate what the optics do"?
I'd assume these computer generated MTFs do include the distortion correction that happens in software as one can compute the resolution decrease due to the software interpolation required for distortion correction.
That seems very silly to present MTFs charts before distortion correction as it is embedded in the lens' firmware. Hardly anyone will use this lens without barrel distortion.

Also, by looking at the lenstip samples, it does seem sharp, even in the corners.

[Brightcolours]

They look pretty sharp-ish, not thaaaat sharp in the corners and borders (not as sharp as those MTFs suggest), albeit that we can't judge there really without knowing field curvature.
https://pliki.optyczne.pl/son16-35G/son16-35fot07.JPG
https://pliki.optyczne.pl/son16-35G/son16-35fot06.JPG

Sure beats my 20-35mm, though.

[thxbb12]

Looks plenty sharp in the corners at 16mm f4: https://www.dpreview.com/sample-gallerie...6038633160