Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
FE 16-35, focus shift and firmware
#1
Been playing with the 16-35 and corresponding a little with Lloyd Chambers about it.

 

He has detected interesting forward focus shift with it using firmware 1.01 on the a7r

He says (preliminarily) it seems to go away using the firmware 1.1 (which came out just before the lens)

 

DXO are showing the periphery getting slightly worse at f5.6 from f4, which on a full frame lens suggest focus shift.  I wonder if they used firmware 1.01?

 

I'm on 1.1, and haven't found any yet (but have not exhaustively looked)

 

 

Hypothesis: either there is a bug in 1.01, or more plausibly, there is native focus shift in the lens but it's corrected for in the 1.1 firmware.

 

If so that's cunning: if you can make a lens cheaper or sharper by allowing a little focus shift, and then compensate for that by moving the focus  (assuming not too much curvature of field) in firmware, that's the kind of lossless goodness I'm all in favour of.

 

Anyhow, if Klaus tests this one it'd be good to know which FW he uses!

#2
Focus shift has nothing to do with the camera firmware. It is a characteristic of the optical design.

 

That being said, the focus shift characteristic could be stored into the lens' firmware. The camera could then compensate the shift during AF based on this data. However, the correct approach would be to AF with the lens stopped down which is, of course, possible on mirrorless cameras (except in very low light). 

 

Thus firmware v1.1 may have such a compensation algorithm in place (Pentax does it as well I think).

 

However, this wouldn't help in manual focus mode.

 

As of today I can't promise a test in the immediate future.
#3
Of course.

 

That's exactly what I was thinking: that the native optical focus shift might be being compensated for in firmware.

 

And it *could* work in manual focus mode in a fly-by-wire camera with a lot of af sensors: you focus manually, and it compensates as you take the picture if the lens is stopped down.

 

Of course to do it accurately without the knowing roughly where you are focussing  would be hard 

#4
DXO has problematic or nonsensical results in other lens "tests" too, so I would not put too much weight in their graphs and findings, regarding this lens.

 

Besides that, I am pretty sure a focus shift makes the center be less sharp.

#5
Lloyd chambers has tested it and finds that there is indeed peripheral focus shift -- i.e. more edge shift than centre shift (another way of thinking of that is that there is field curvature which gets worse as you stop down)  And sadly the initial thought that it is compensated in FW seems not to be true.

 

Not that uncommon in good UW zooms - the Nikon 14-24 has it just as bad according to Lloyd.

 

Other than that, though, he finds the lens to be exceptional - as good or better as any uw zoom, and very flat field wide open.

 

BTW his tests seem in conformity with DXOs   

#6
Looks like, on closer inspection, there's not FW related difference in focus shift.

But there is quite a bit of focus shift, especially at close focus, which I think explains the DXO data showing it getting worse as you stop down.

 

That's a bugger. But other than that it's very nice (except at 35mm). Sharp across the frame from wide open, contrasty, and with surprisingly smooth bokeh in situations where you can see any (i.e. at close focus wide open)

the long end is a bit uncontrasty, and needs stopping down to get sharp. But it's a terrific 16-18-21-24-28 f4...

#7
Makes me wonder why Sony isn't doing stopped-down AF metering as long as there's sufficient light.

  


Forum Jump:


Users browsing this thread:
1 Guest(s)