Vignetting can vary with focus

[rfdesigner]

Hi

Thought I'd post this after a conversation with TDP.

I noticed they had the 50STM down as having vignetting at 1.25EV, whilst DxO had measured 2.5~3EV, and I had measured mine as being 2.0EV.

We exchanged files and agreed TDPs measurement was indeed 1.25EV, whilst mine measurement yielded 2.0EV

I was trying to think why this could be when the penny dropped:

The 50STM is a classic 50mm design in that the entire optic moves in and out to focus.

What I realised was that if you're focussed at infinity the optics are very close to the sensor so getting light into the corners is rather difficult, where as at minimum focus distance the optics are well away from the sensor and so the ability to light the corners is much improved.

I therefor re-measured what I got with my lens at both extremes.

I found around 1.5EV vignetting at minimum focus distance and 2.6EV at infinity.

I therefore believe that vignetting numbers need to be given for minimum and maximum focus distance.. many lenses ought to have little difference, but that should not be assumed, it might also explain some of the differences between different testers.

I am yet to check my other lenses.

 

[LordofTackle]

Interesting finding, thanks for posting

Maybe then there is still hope for the 16-35 III.
Do you know whether Brian measures all his lenses at minimum focus distance? or Maximum? or random?

 

[rfdesigner]

Interesting finding, thanks for posting

Maybe then there is still hope for the 16-35 III.
Do you know whether Brian measures all his lenses at minimum focus distance? or Maximum? or random?

He said maximum, he shared a copy of his vignetting file, and that claims focus is at infinity, which ought to be high vignetting case, I'm wondering if it's possible for the camera to get confused (it's a 1dsMkIII).

My old 30D never recorded focus distance in the file, I don't know about the 1DsIII.

And yes we have the 16-35f2.8III to blame for this finding, it made me start chaseing down vignetting figures and I happened to choose the 50STM as a starting point as I had one.

 

[Alex_M]

Is that right? Focus at infinity should produce the lesser amount of vignetting than at minimum focusing distance. What I am missing?
Actually, can the lens owners confirm that the rear element is at the most distant position from the sensor plane when focused at infunity? All that needs to be done is: rotate the focusing ring while looking at rear element and taking note of it's position relative to the lens mount.

He said maximum, he shared a copy of his vignetting file, and that claims focus is at infinity, which ought to be high vignetting case, I'm wondering if it's possible for the camera to get confused (it's a 1dsMkIII).

 

[rfdesigner]

Is that right? Focus at infinity should produce the lesser amount of vignetting than at minimum focusing distance. What I am missing?
Actually, can the lens owners confirm that the rear element is at the most distant position from the sensor plane when focused at infunity? All that needs to be done is: rotate the focusing ring while looking at rear element and taking note of it's position relative to the lens mount.

He said maximum, he shared a copy of his vignetting file, and that claims focus is at infinity, which ought to be high vignetting case, I'm wondering if it's possible for the camera to get confused (it's a 1dsMkIII).

Here's the evidence, for both the 50STM and my 28f1.8, they have very similar levels of vignetting and both see about a stop of difference in vignetting between infinity and minimum focus, focus can be verified in DPP.

https://dl.dropboxusercontent.com/u/44641599/IMG_2817_50_at_min.CR2

https://dl.dropboxusercontent.com/u/44641599/IMG_2816_50_at_infinity.CR2

https://dl.dropboxusercontent.com/u/44641599/IMG_2815_28_at_infinity.CR2

https://dl.dropboxusercontent.com/u/44641599/IMG_2814_28_at_min.CR2

 

[Alex_M]

rfdesigner

All correct, Sir!