Strange Field Curvature

[Dan Finlayson]

Yesterday I needed to evaluate the basic condition of a lens quick and dirty - no fancy equipment at my disposal.  I decided to shoot some stills through it at a grassy field to get an idea of what field curvature the lens has and if there's any weird tilt.  I was expecting a gentle curvature with the plane of focus curving toward camera at the edges.  What I got was pretty surprising:

https://imgur.com/a/GZDs7Wy

In this shot the lens is set to focus just a few feet in front of me.  The plane of focus sweeps away from me dramatically toward the edges.

This is a Zeiss standard speed.  I've shot on standards probably more than any other lens and I've NEVER noticed one do this.  At the same time, I've never really gone looking for the focus plane like this before.  Is this anywhere close to normal?  If not, what would cause it?

[Simon Wyss]

Zeiss standard speed is not very precise. Could you clarify a little?

[Dom Jaeger]

What I think you’re seeing isn’t necessarily field curvature but increased depth of field at the edges.

The plane of focus is set quite near, with only the very bottom of frame in sharp focus. With many older lens types, the way the focus falls off is much steeper in the middle of the frame, with the edges defocussing much less, which is what’s happening here.

If you were to take the lens to a rental house with a lens projector and project it slightly defocussed, you would see that the test pattern is more defocussed in the centre than the edges.

Have a look at the lens projections in this blog post I made about different lens characteristics (note that these are quarter field photos, so the top right is centre and bottom left is the corner):

http://cinetinker.blogspot.com/2013/10/comparing-lenses-blur-characteristics.html

Notice how a more modern lens like an Ultra Prime projects a more evenly defocussed  image compared to the older lenses like Speed Panchros or Schneider’s where the corner is far less defocussed than the centre.

[Dom Jaeger]

4 hours ago, Simon Wyss said:

Zeiss standard speed is not very precise. Could you clarify a little?

Zeiss Standard Speeds are very well known T2.1 cinematography lenses, they have been used on movies for decades. They were an earlier cine lens design compared to the T1.4 Zeiss Super Speeds. There are several iterations, the first date back to the 60s, but the optical formula remained the same. Google them.

[Simon Wyss]

I wasn’t aware of the focal length in question. Seems to be 28 mm, so the Distagon design.

Googled and found https://www.opticallimits.com/nikon_ff/586-zeisszf2820ff?start=1.

All retrofocal wide-angle lenses have some residual field curvature.

[Dan Finlayson]

3 hours ago, Dom Jaeger said:

Have a look at the lens projections in this blog post I made about different lens characteristics (note that these are quarter field photos, so the top right is centre and bottom left is the corner):

http://cinetinker.blogspot.com/2013/10/comparing-lenses-blur-characteristics.html

Thank you so much Dom!  I'm so glad I posted here - your lens test is an amazing resource!

I've definitely seen this sort of phenomenon as you describe and demonstrate on your website.  New lenses like the Leica cine lenses have so little of this that I don't enjoy their images.  What I'm seeing on your test charts lines up with my anecdotal experiences.  Something about this wide shot I took (shot open at T2.1) just feels more pronounced than anything I've seen before.

I know this is hard to evaluate from my shots but does this strike you as normal performance out of a 28mm Standard Speed?

Also - based on my very limited understanding of the history of these lenses I'm under the impression the 28mm was one of the later designs they manufactured.  Is this true?  Did they make multiple iterations of this lens?

 

[Dom Jaeger]

4 hours ago, Dan Finlayson said:

I know this is hard to evaluate from my shots but does this strike you as normal performance out of a 28mm Standard Speed?

Also - based on my very limited understanding of the history of these lenses I'm under the impression the 28mm was one of the later designs they manufactured.  Is this true?  Did they make multiple iterations of this lens?

 

What size sensor did you use to capture the still image? The 28mm Zeiss Standard Speed has about a 36mm diameter image circle - depending on how much of that circle you use the edges will have deeper depth of field. I can't say if it's normal or not just from one mainly out of focus image of a field, but I took some pics of a test projection of one I have here to show the defocus characteristics.

Here's the lens focussed a few feet in front of the screen:

It's a bit low res due to the low file size allowed, but you can clearly see the difference between the defocussed centre and the corners. Outside the S35 frame the Siemens stars are even more in focus. 

In case anyone thinks this is caused by field curvature, there is actually little field curvature to this lens, as seen when the lens is focussed on the screen:

The field is quite flat all the way to the edges. The top image is purely due to the way the centre defocusses more dramatically than the corners.

The 28mm was not in the original Mk1 line-up of Standard Speeds, but added during the Mk2 period in the (I believe) 80s. It still behaves like the earlier Distagon designs though, which makes sense as they were made to be a matched set.

 

 

 

Edited February 25, 2021 by Dom Jaeger

[Dan Finlayson]

3 hours ago, Dom Jaeger said:

I took some pics of a test projection of one I have here to show the defocus characteristics.

Ahh thank you!  That dramatic defocus in the center is definitely reading here.

I captured my test images with a Sony A7s.  I then cropped the image to 2814px across which, according to my math, should closely match the horizontal FOV of an Alexa Classic.  Here's one more image, this time focused at about infinity:

https://imgur.com/a/BPu8Xuu

3 hours ago, Dom Jaeger said:

In case anyone thinks this is caused by field curvature, there is actually little field curvature to this lens, as seen when the lens is focussed on the screen:

I'm really glad you included this in focus image as well because it raises some questions about the particular lens I shot through.  In the infinity focus image there's quite a bit of softening at the edges.  Is this a characteristic of infinity focus on these old Distagon lenses or should I be seeing flatter focus across the image circle than this like on your lens projector shot?  I shot this image wide open.

[Dom Jaeger]

The example I have seems quite sharp to the corners, but I can't project it past 20 ft.

That infinity image looks softer than it should to me, perhaps have it checked by a tech.