What's the right forum to ask about field curvature?

[Juan Valdenebro]

Field curvature is just one of the major aberrations, and a major part of lens design is about trading off the different aberrations against each other and against other considerations (size, price, and so on). So for example, there' s no particular type of glass that minimizes field curvature. In that Nikkor, there might have been some glass that used in one element enabled a tradeoff that reduced field curvature, but it doesn't mean that plunking that glass into some other lens or some other element would have a good effect.

Juan, in your example of a 35mm lens on 35mm film, at f/8 and focused at 2 meters, the conventional depth of field (with a circle of confusion of 0.03mm) is roughly from 1.4 to 3.2 meters. Probably not enough to bring the house into sharp focus even with a perfectly flat field lens. Try for example the DOF calculator at https://www.dofmaster.com/dofjs.html In contrast, focusing at 3 meters, the DOF is from 1.9 to 7.2 meters. While curvature of field may have affected your photographs, the other issue is that one loses DOF very rapidly as one focuses closer.

Hi,
The lens was focused at 3 meters, thanks.

 

[Juan Valdenebro]

You don’t want to see graphs yet you want to understand the aberration? I’m sorry, to truly understand what’s going on requires study of the underlying math and physics. It’s not something that can be put readily into layman’s terms to advance your understanding beyond myths, half-truths, and marketing BS. There’s just no way.

First thing to understand: Field curvature isn’t a magical cause of all a lens’ woes. Optical fabrication and assembly tolerances are significant contributors to image degradation in a modern lens, and could just as easily be the root cause to your experiences as anything. Any review or test of a lens which draws conclusions about the design based on a single sample should be immediately suspect: The reviewer has no idea where that lens falls in the performance distribution for that design, and for consumer-grade optics such as used for photography the distribution can be quite wide.

underlying that is the effect of balancing of all the aberrations, which for modern optics really must be viewed as Zernicke terms as the higher order aberrations are often balanced against and of similar magnitude as the residual third order aberrations.

if you really want to understand field curvature, and not just myths and legends, I highly recommend accepting the fate that it requires maths and follow the link I provided above. You will not gain a full understanding there, but it will shed light on the tools required to truly understand the phenomenon.

If however you are happy with myths and legends, then yes.. field curvature causes everything.

I don't know who you're talking too... I don't think field curvature causes anything.
And no one here has said field curvature causes everything: you need help, seriously...
And about the math and graphs ways of understanding what happens on photographs, that was not needed at all by Nadar or Atget or Cartier-Bresson or Frank or Winogrand.
I have no interest at all in what you feel is important, I'm sorry.
I live inside a different field called Photography.
And we all (photographers) understand field curvature very well.
Lens designers study those fields differently.
Enjoy your math and graphs, and also enjoy photography whenever you feel like doing it.

 

[Juan Valdenebro]

Why do you believe that lens design can tame diffraction? And which lens are you talking about? I ask because I'm not aware of any 35 mm lens, if that's what you meant, that is a double Gauss design. Many 50 - 60 mm lenses for 35 mm cameras are 6/4 double Gauss designs or variants, but 35 mm?

Do a search.

 

[Juan Valdenebro]

complete broken biscuits by op

Best anwser to this thread's subject.

 

[Juan Valdenebro]

In practice, getting an idea of the field curvature of one's lenses is easy. Take a picture of something textured and flat, stretching away from the camera. For close up, a newspaper or similar, for longer distance, flat lawn should work. At higher magnification, the plane or curve of focus will be obvious.
I doubt you will find much about the lenses' behaviour on film. Field curvature is different on digital due to sensor cover glass. Towards the corners of the sensor, light must travel for a longer distance through the cover glass, which has a different refractive index from air. No such effect with film.

Thank you, grain...
You're right.
I thought maybe some other forum members could have found situations where field curvature required, as in the case I explaned, an unusual f-stop.
Have a nice day.

 

[Nodda Duma]

I am not sure why you think you need to be insulting, especially to someone who is trying to be helpful.

Subject matter experts stop contributing to Photrio because of attitudes like that. In fact I wasn’t going to comment except Dan Fromm mentioned my name and I figured I’d try out of respect for him.

I was talking to you, because I’m confused about your original ask combined with your reticence to gain the knowledge needed to understand the phenomena. In any case, I won’t try to help further as it’s no skin off my back whether or not you gain an understanding.

Regards,
Jason

I don't know who you're talking too... I don't think field curvature causes anything.
And no one here has said field curvature causes everything: you need help, seriously...
And about the math and graphs ways of understanding what happens on photographs, that was not needed at all by Nadar or Atget or Cartier-Bresson or Frank or Winogrand.
I have no interest at all in what you feel is important, I'm sorry.
I live inside a different field called Photography.
And we all (photographers) understand field curvature very well.
Lens designers study those fields differently.
Enjoy your math and graphs, and also enjoy photography whenever you feel like doing it.

 

[grain elevator]

Thank you, grain...
You're right.
I thought maybe some other forum members could have found situations where field curvature required, as in the case I explaned, an unusual f-stop.
Have a nice day.

Oh yes, I have had that situation. A Jupiter-12 has crazy field curvature.
I think for reportage and portrait lenses, a field curvature of a radius equal to the object-camera distance would be ideal. This would eliminate what I think is called cosine error, then focus and recompose would never result in missed focus, at the expense of flat objects not being all in focus, but who shoots those at wide apertures? It's not going to happen nowadays, because test targets are flat...

 

[MattKing]

I don't know who you're talking too... I don't think field curvature causes anything.
And no one here has said field curvature causes everything: you need help, seriously...
And about the math and graphs ways of understanding what happens on photographs, that was not needed at all by Nadar or Atget or Cartier-Bresson or Frank or Winogrand.
I have no interest at all in what you feel is important, I'm sorry.
I live inside a different field called Photography.
And we all (photographers) understand field curvature very well.
Lens designers study those fields differently.
Enjoy your math and graphs, and also enjoy photography whenever you feel like doing it.

In case you weren't aware, Nodda Dumma designs lenses for part (most?) of his living.
In the guise of Jason Lane, he also manufactures and sells to customers dry plates.
In both roles, he is likely to have lots of useful knowledge about how lenses exhibit field curvature, and how lens designers and manufacturers deal with it.

 

[Dan Fromm]

Do a search.

I'm sorry, that's an inappropriate answer.

 

[Sirius Glass]

Yes, Cicala's/lensrentals MTF graphs have been around for some years, but despite their multiple and possibly accurate graphs, I never found real photographs showing real life differences between the most used rangefinder 35s by Leitz/Leica/Zeiss/Canon/Nikon/Voigtlander in a comparative or comprehensive way. Thank you.

If a company provides MTF curves, each uses its own standard so the cannot be compared.

 

[Dan Fromm]

If a company provides MTF curves, each uses its own standard so the cannot be compared.

Why do you believe that? Its true that they don't all measure MTF at the same spatial frequencies or apertures but that doesn't mean one can't base decisions on them.

 

[warden]

I'm interested in learning about field curvature as we find it in 35mm lenses for 35mm rangefinders... Articles, links, books, and of course general comments from forum members...
I just thought there was an optics/lens design forum, but maybe I'm confused and it was not here at Photrio...
Thanks.

Do a search.

 

[Oren Grad]

Why do you believe that? Its true that they don't all measure MTF at the same spatial frequencies or apertures but that doesn't mean one can't base decisions on them.

Some show calculated MTF without diffraction... some show calculated MTF with diffraction... some show measured MTF. Caveat lector.

 

[Dan Fromm]

Oren, not to quarrel with you, but I'd love to have a reference for published MTFs with diffraction. Understand, I've ignored published MTFs for small format lenses, have looked only at published MTFs for LF lenses.

 

[Oren Grad]

Oren, not to quarrel with you, but I'd love to have a reference for published MTFs with diffraction.

Fair question!

https://www.sigma-global.com/en/glossary/mtf-chart-guide/

Example - scroll down to see both sets:

https://www.sigma-global.com/en/lenses/s014_150_600_5_63/#performance

 

[Dan Fromm]

Fair question!

https://www.sigma-global.com/en/glossary/mtf-chart-guide/

Thanks for the link. Text in it says:

The readings at 10 lines per millimeter measure the lens’s contrast ability ( red lines), repeating fine parallel lines spaced at 30 lines per millimeter measure the lens’s sharpness ability (green lines), when the aperture is wide open.

Huh? I thought that both report subject contrast/image contrast by distance off-axis at, respectively, very low resolution and at somewhat higher resolution, both at full aperture. Possibly a poor translation. Yes, I understand that it is an example, not a chart for a specific or at any rate identifiable lens.

I took a casual look at a few identifiable lenses' MTF charts. All were at full aperture. I'm interested in 105 mm +/- macro lenses so paid attention to this https://www.sigma-global.com/en/lenses/a020_105_28/ one. Wide open at unspecified magnification "diffraction" and "geometrical" MTFs are very close but not quite identical. This surprised me, especially since they measured and calculated at fairly low spatial frequencies (10 and 40 cy/mm). Since macro lenses are often used stopped down somewhat so I don't see how the published MTF chart for the lens in question can guide a decision about whether to stick with my good old 105/2.8 MicroNikkor AIS or replace it with Sigma's new wonder lens.

This may be an example of marketing to the digidiots who value lens' performance wide open much more than performance in what I see as real world use.

The LF lenses' MTF charts I'm familiar with are presented for a number of apertures, focused distances and, usually, for more spatial frequencies. They're potentially more informative than Sigma's marketing fluff.

Cheers,

Dan

 

[Oren Grad]

Dan, I agree with you that the Sigma data raise as many questions as they answer. But I think that underscores the point that published MTF data for small-format lenses are a swamp; use for comparison purposes at your peril.

Tangentially and FWIW, the 70mm Sigma Art macro is superb, at least for my current use of it on a Sony A7RIV for high-resolution film copy work.

 

[Dan Fromm]

Dan, I agree with you that the Sigma data raise as many questions as they answer. But I think that underscores the point that published MTF data for small-format lenses are a swamp; use for comparison purposes at your peril.

Tangentially and FWIW, the 70mm Sigma Art macro is superb, at least for my current use of it on a Sony A7RIV for high-resolution film copy work.

Oren, thanks for the reply. I hadn't been aware that I had a good reason for avoiding MTF data for small format lenses.

Thanks also for the news about the 70 mm Sigma Art macro.

 

[Kodachromeguy]

Here is an example of major curvature: my 1949 Leitz 5cm ƒ/2 Summitar. This is an alley near Asan Tol in Kathmandu. Note that the house in the distance at the end of the alley is in focus, and so is the sign on the left and the water bottled on the lower right. This was TMax 400 film, f-stop not recorded, but probably ƒ/5.6 or so. The Summitar is a good alley lens

In contrast, the 50mm ƒ/3.5 Color-Skopar (Tessar-type) on my Vito BL is much flatter field. I'll need to take some brick wall photos....

 

[reddesert]

In practice, getting an idea of the field curvature of one's lenses is easy. Take a picture of something textured and flat, stretching away from the camera. For close up, a newspaper or similar, for longer distance, flat lawn should work. At higher magnification, the plane or curve of focus will be obvious.

Cicala also recommends this field-of-grass test. I have not actually tried it.

I doubt you will find much about the lenses' behaviour on film. Field curvature is different on digital due to sensor cover glass. Towards the corners of the sensor, light must travel for a longer distance through the cover glass, which has a different refractive index from air. No such effect with film.

To first order at least, I don't think the glass changes field curvature. Roughly speaking a 1mm thickness of glass replaces a 1.5mm thickness of air (for a refractive index of 1.5). Off-axis, the beam was traveling through a greater amount of air (1/cosine of the off-axis ray angle) and now it's traveling through the 1/cosine * thickness of glass.

I am pretty sure that the off-axis beam traveling through the cover glass plate at an angle does cause astigmatism. This might be why some people report that non-retrofocus wide angles designed for film have problems in the corners on FF digital. A retrofocus wide angle shouldn't suffer much of a problem because its ray angle is less steep in the corners. I wrote about this conjecture here once: https://www.photrio.com/forum/threa...n-lenses-on-film-cameras.169681/#post-2207803

 

[darinwc]

Do a search.

For all but the really dumb questions, this is the most asshat answer ever. This is a discussion forum right??? right???????
If everyone 'just did a search', this forum would not exist.

I think the OP question was valid and asking for advice and possibly reference material to learn about a technical subject.
Contrary to popular belief, web searches can return pretty superficial or incoherent responses to questions that have common words or phrases.

 

[Mike Lopez]

For all but the really dumb questions, this is the most asshat answer ever. This is a discussion forum right??? right???????
If everyone 'just did a search', this forum would not exist.

I think the OP question was valid and asking for advice and possibly reference material to learn about a technical subject.
Contrary to popular belief, web searches can return pretty superficial or incoherent responses to questions that have common words or phrases.

Um, I'm pretty sure that was posted in response to the shit stance that Juan took earlier in the thread. Read through the entire thread and you'll see where the asshattery began. (Hint: start with post #28)

 

[grat]

I think the OP question was valid and asking for advice and possibly reference material to learn about a technical subject.

Ordinarily, I would agree with you-- but when a bona-fide optical expert who's pretty well respected around here offered explanations (and reference material, if I recall), the OP became somewhat antagonistic.

 

[warden]

For all but the really dumb questions, this is the most asshat answer ever. This is a discussion forum right??? right???????

I think as Mike said you might have missed the context of that particular post. I try to be helpful around these parts and perhaps you'll see that if you review my posting history.