field curvature on AR Hex 57/1.4 shown

[ampguy]

Field curvature is when the edges are not exactly focused at the point the center is. Here's a good example with the NEX focused on the center area, on a perfectly straight string.

By the nut, and the 2nd fret, you can see the softness. This was taken wide open at 1.4, it's possible that the curvature would behave differently or be masked stopped down.

Now if I had focused on the nut or the 2nd fret, with same composition, someone might evaluate the lens as having sharp corners, but I didn't quite focus right 😉

Other lenses, that were designed for this, include the Leica 50 pre-asph summilux.

 

[former member]

So the interesting question: did the lens designer introduce this aberration knowing it would be offset by the bow of the film sitting in a Konica camera? Film is not as flat as a sensor.

I have this lens on my FT-1 now.

 

[ampguy]

Hi Brian

Hi Brian

Possibly only the designer really knows, but my guess is that it wasn't put in for film curvature because the nex sensor area shown is not including the extreme edges.

If you have an extra frame next time you run a roll through your SLR FF, maybe you could check wide open and min distance.

So the interesting question: did the lens designer introduce this aberration knowing it would be offset by the bow of the film sitting in a Konica camera? Film is not as flat as a sensor.

I have this lens on my FT-1 now.

 

[former member]

I read in an older Optical Design manual that optical engineers would introduce a bit of field curvature into the lens design to negate astigmatism. Looking at the lack of astigmatism in Sonnar formula lenses, I tend to believe it.

 

[rxmd]

My first question would be how you made sure that the strings of the bass guitar and the NEX sensor plane were exactly parallel.

 

[ampguy]

I don't know how much of it was intentional, but once learned that it's there, you can use it to focus more sharply on subjects, or groups of subjects that are convex from the lens.

For example, if you're shooting a group of family members, you don't have to ask them to all be at the same distance from lens, the folks on the edges can stay hanging back a bit, and you can shoot wider open and sharper.

Dick Gilcreast has a landscape example with the 50 lux, in a very old Viewfinder article.

 

[ampguy]

they probably arent

they probably arent

exactly. But the string is straight and I'm showing it going out of focus at both edges. How do I know the string is straight? because it is 34" under tension from nut to bridge, without touching any frets.

The fretboard is not perfectly flat, but essentially. In that when the string is fretted at the first fret, and the 15+fret, there is about 0.10mm of relief in the middle if you put a straight edge across the frets.

My first question would be how you made sure that the strings of the bass guitar and the NEX sensor plane were exactly parallel.

 

[rxmd]

exactly. But the string is straight and I'm showing it going out of focus at both edges.

The question is not whether the strings are straight; for all practical purposes they are.

However, you don't know whether it goes out of focus in the same direction on both ends (because of curvature of field), or whether it goes out of focus forward on one end and backward on another. The latter case would not be due to curvature of field, but due to extremely low DOF close up and wide open when they aren't exactly parallel.

I don't think there is enough detail in the scene to distinguish between those two cases. The diagonal frets at various angles, however, make me suspicious that it isn't parallel. You can't distinguish between the two sensibly if you just sit in front of your bass guitar and shoot the strings close up. With that kind of scene, unless you make sure that the field you're photographing is actually parallel to the sensor plane, there is no obvious way to demonstrate curvature of field.

 

[ampguy]

These are good points

These are good points

Actually if I did this test on an RF, I'd have to take 20 or so photos and probably need a tripod.

However, with the NEX, I am able to just hold my Nex in place, and slightly focus a bit more towards infinity, to get either the higher (nut), or lower (2nd fret) in focus, while holding the Nex steady.

I apologize if I should have posted this in the Nex forum instead of here to avoid confusions.

The lens was originally made to fit on an SLR, btw.

The question is not whether the strings are straight; for all practical purposes they are.

However, you don't know whether it goes out of focus in the same direction on both ends (because of curvature of field), or whether it goes out of focus forward on one end and backward on another. The latter case would not be due to curvature of field, but due to extremely low DOF close up and wide open when they aren't exactly parallel.

I don't think there is enough detail in the scene to distinguish between those two cases. The diagonal frets at various angles, however, make me suspicious that it isn't parallel. You can't distinguish between the two sensibly if you just sit in front of your bass guitar and shoot the strings close up. With that kind of scene, unless you make sure that the field you're photographing is actually parallel to the sensor plane, there is no obvious way to demonstrate curvature of field.

 

[sanmich]

Ted

I was thinking about this a couple of days ago:
We do a resolution test, and find that the corners are softer than the center.
The question is now:
Is this curvature of field or simple IQ drop?
Curvature of field would mean there is a plane, closer to the camera, that is sharp (as sharp as the center?).
This is a specially interesting question since some of field curvature would be good in order to compensate for recomposing. In that respect, the ideal lens would have a spherical plane of focus (??)
With all this in mind, I have reached the conclusion, that the amateur newspaper shots lens testing I'm doing from time to time may be almost completely irrelevant when the resolution is measured in the corners.
The only thing measureable would be stopped down performance where the curvature of field would be covered by the DOF, and we would need to know how much curvature of field is expected in order to know at what stop the lens is showing its true abilities in the corners.
Confused??

Not as much as I am... 😕 🙄

 

[sanmich]

there is no obvious way to demonstrate curvature of field.

Actually, one could use the proverbial newspaper on a wall, bracket on the focus and see if the corners get sharper while the center looses sharpness.

Not an enjoyable way to spend an afternoon if you use film...

 

[ampguy]

Sure, you could use a newspaper taped to the wall to show the same thing, has already been done for many lenses. Just pointing out that with some lenses, it doesn't show corner softness for that lens, it shows that they didn't focus that lens for the corners.

With RF's, it's not possible to focus in the corners, you have to reframe. With the NEX and full groundglass type focusing mechanisms, you can see the whole image.