[malcolmdavies]

I made the following general observation when testing my M 50 1.4 on a full frame K1:

At all apertures that when focused at far distance in the centre of the frame, the focal plane ('field') comes forwards significantly towards the borders.

I would expect the focal plane to be equidistant from the camera across the frame and therefore slightly curved, but the focus in the borders is far forward of an equidistant curve. I researched this further and the phenomenon is called Petzval Curvature, where the focal plane ('field curvature') is paraboloidal and worse at open apertures. I understand that is it possible to mitigate this in lens design and I haven't noticed it particularly on my other lenses (although I have only recently upgraded to full frame) but it is clearly significant in the M 50 1.4 lens on full frame a K1.

It is not something I have ever seen measured in a lens review test but I now think should be; what appears to be edge softness on a flat test chart is in fact more likely to be this effect. It is a real issue.

I am surprised that this doesn't come up in Lens Sharpness discussions on Pentax Forums, including Ken Rockwell's Lens Sharpness missive.

---

Last edited by malcolmdavies; 04-29-2021 at 04:20 AM.

[D1N0]

The issue goes away as you stop down. It as a characteristic of a fast 50 especially when they are very compact. DXO mark tests for it Pentax smc FA 50mm F1.4 - DxOMark (FA version of the 50/1.4) look for sharpness and the field map. Imaging resource does to (or used to anyway) Pentax 43mm f/1.9 Limited SMC P-FA Review Also this was all magazines used to to when testing lenses. Resolution and contrast center and corners and rating lenses accordingly. Still dpreview e.a frown upon cat eye bokeh balls (a sure sign of field curvature). But in the vintage world these flaws are something good as they add character and uniqueness to a lens. Only CA is hated by everyone

[Serkevan]

Flat field lenses are generally larger and more complex, so there's no free lunch.

As long as the curvature is symmetric and you know how your lens behaves, it's a photographic tool like any other. For example, if the plane of focus is closer to you on the edges, it will probably give a centered portrait creamier bokeh on the sides as the background is more separated.

Of course, for astro or other aberration-sensitive types of photography, field curvature is really annoying. Horses for courses

[malcolmdavies]

The issue goes away as you stop down. It as a characteristic of a fast 50 especially when they are very compact. DXO mark tests for it Pentax smc FA 50mm F1.4 - DxOMark (FA version of the 50/1.4) look for sharpness and the field map. Imaging resource does to (or used to anyway) Pentax 43mm f/1.9 Limited SMC P-FA Review Also this was all magazines used to to when testing lenses. Resolution and contrast center and corners and rating lenses accordingly. Still dpreview e.a frown upon cat eye bokeh balls (a sure sign of field curvature). But in the vintage world these flaws are something good as they add character and uniqueness to a lens. Only CA is hated by everyone

Thank you this. I didn't know that the results of a field map were attributable to this phenomenon; I had thought the loss of acuteness was just 'softness'..

[ChristianRock]

Back in the day, before websites tested lens test charts in a basement, lens designers designed the lenses to achieve certain looks and rendering. For example, Hirakawa Jun, the former Pentax designer that was responsible for the FA Limiteds, published a paper where the described the intended look they wanted to achieve with their lenses - how bokeh transition and bokeh quality was a chief concern, and how skin and hair softness was to be achieved through the lens design, while still maintaining sharpness in the image. It specifically said that the lenses were not designed for absolute sharpness across the field and that some field curvature was to be left built into the design as it was seen to improve how subjects were portrayed. Pentax has even said in the past that they have left aberrations uncorrected to a certain extent, because correcting them further would be detrimental to the out of focus rendering.

The fact is that field curvature is desirable for some applications because of the effect it has on the image. Designers did not use to design lenses for test charts - and I would argue that this is the reason why a lot of us love how old lenses lender and think that a lot of newer lenses look "flat" and the images just lack that artistic quality to them.

[stevebrot]

Field curvature?

The first question to ask oneself might be, "why would one focus to the center if the subject is in the corner?"

Flatness of field is highly desirable for flat subject copy and for certain macro subjects (stamps?), but irrelevant for most field work where the topography generally contains sufficient relief to make field flatness the least of one's concerns. (Lower corners are generally much closer with upper being further away.)


Steve

[swanlefitte]

Lensrentals shows field curvature
Each lens review has it and many meta studies such as this one.
Lens Rentals | Blog

Edit: In one of the blogs a takeaway is that by slightly front focusing on many lenses in the center you loose a touch of center sharpness but gain alot across the frame because of the curve and fall off rate from the curve.

---

Last edited by swanlefitte; 04-29-2021 at 10:00 AM.

[TwoUptons]

There's a two-part thing Roger Cicala (of Lensrentals) did at the other site...
I found it very informative and entertaining...

Roger Cicala: Understanding field curvature for fun and profit: Digital Photography Review

-Eric

[swanlefitte]

There's a two-part thing Roger Cicala (of Lensrentals) did at the other site...
I found it very informative and entertaining...

Roger Cicala: Understanding field curvature for fun and profit: Digital Photography Review

-Eric

Thanks for the heads up. I lost touch with Rodger after he left lens rental.

[DWS1]

It's part of understanding the len's character and learning how to utilize it to an advantage.

[Lowell Goudge]

This is why, years ago, people used to really push for macro lenses to be used in copy applications, they were not fast, by today’s standards but generally were flat field lenses

[malcolmdavies]

This is why, years ago, people used to really push for macro lenses to be used in copy applications, they were not fast, by today’s standards but generally were flat field lenses

I note the comments on how the 'charm' of a lens that has significant field curvature may have in portraiture, but I want a 50mm lens for landscape work that does not have this 'feature' and has a flatter field.

Is there another 50mm lens that would met this criterium, for example the SMC Pentax-D FA 50mm F2.8 Macro?

[Lowell Goudge]

I note the comments on how the 'charm' of a lens that has significant field curvature may have in portraiture, but I want a 50mm lens for landscape work that does not have this 'feature' and has a flatter field.

Is there another 50mm lens that would met this criterium, for example the SMC Pentax-D FA 50mm F2.8 Macro?

I am not the best one on landscapes, I shoot mostly wildlife with 400mm plus lenses. But..........

Considering most landscapes require you to shoot at the hyper focal point of the lens for maximum depth of field, and landscapes are not normally constructed on single flat planes, I am not sure you are concentrating on the correct properties of a lens. If you said you were into architecture, I would perhaps understand your comments more, as I interpret your requirement for flat focus plane as much more important. For landscapes, I would personally think the most important criteria would be lateral CA.

[malcolmdavies]

I am not the best one on landscapes, I shoot mostly wildlife with 400mm plus lenses. But..........

Considering most landscapes require you to shoot at the hyper focal point of the lens for maximum depth of field, and landscapes are not normally constructed on single flat planes, I am not sure you are concentrating on the correct properties of a lens. If you said you were into architecture, I would perhaps understand your comments more, as I interpret your requirement for flat focus plane as much more important. For landscapes, I would personally think the most important criteria would be lateral CA.

I am an experienced and qualified photographer and I know understand what to concentrate on. I have taken landscapes with the 50 1.4 at f8 on full frame (k1) and whilst diffraction is already taking over, it so front focused at the edges as to be completely unsuitable for this task; that's a fact.
What would be helpful would be suggestions for an alternative for this task.

[TwoUptons]

I don’t know how it does for field curvature at infinity focus, but the FA 50mm f2.8 macro is indeed a very good copy lens, with a very flat field at distances out to a couple of meters. The DFA is reportedly the same in this respect.

It’s a great lens in general...

-Eric