Depth of Field ... is it curved? ... or linear?

Usually I shoot individuals, couples or a few people. Rarely groups. But I see myself quickly headed for large group portraits. The question came to mind tonight whether the field in focus (DOF) is a linear straight line or field perpendicular to the lens, or whether it's curved? Suppose a certain aperture offered a DOF 4 feet deep, would a group of 30 people standing side to side need to be arranged in an arc or a curve? Or could they be in a straight line perpendicular to the lens?

I've only done one image with that many people, but never gave the shape of the DOF much thought. If the area in focus happens to be curved, are there lenses designed to compensate, or are there references to know the difference, one lens from another?

Very good question. I'm keen to hear from others more knowledgeable than me but to kick it off, I understand that most lenses have a fairly flat plane of focus (and that lenses are generally designed to achieve this). This makes for good test chart results. For some lenses, however, the plane of focus has some degree of curvature towards the camera at the edges and corners. For some types of photography, especially landscapes, this is not all that important but for others it is. You have identified one of those situations when it can be important.

It is my understanding that usually it is only the macro lenses which are flat field corrected, but only at a certain range of distances!

Some lenses get criticized more for field curvature, but I think the reality is that -- other than the flat field macro distances -- most lenses have some degree of curvature, and

• the characteristic of curvature actually can CHANGE, particularly in zoom lens designs!
  
• Even for fixed focal length lens, the shape of the curvature can DIFFER at different focus distances, too.
  

Pippan wrote in post #18481258
Very good question. I'm keen to hear from others more knowledgeable than me but to kick it off, I understand that most lenses have a fairly flat plane of focus (and that lenses are generally designed to achieve this). This makes for good test chart results. For some lenses, however, the plane of focus has some degree of curvature towards the camera at the edges and corners. For some types of photography, especially landscapes, this is not all that important but for others it is. You have identified one of those situations when it can be important.

Here's the only big group shot I can recall taking. Two classes on a redwood hiking trail. I cropped to show it a bit closer. Not razor sharp, but good enough for the purpose. This was taken from farther than it seems. This is the center third from side to side of a portrait oriented image.

Tamron 24-70mm shot at 24mm ... ... ISO 1000 ... 1/100 ... aperture f/2.8

I don't have numbers, but I'm pretty sure most lenses have a curved focal plane. I don't think it is a significant issue in most situations, though. Photocopying flat originals would be an exception. Process lenses, enlarger lenses and some macro lenses are designed with a flat field for such applications.

In the olden days, we had projection lenses intentionally designed with curved fields so they would be better able to project color slides, because the film in those slides was often curved. So curvature depends on lens design.

Apart from the issue of curvature of the focal plane, we could ask if the depth of field is the same off-axis as on-axis. My guess is that it might be a bit wider off-axis because the off-angle aperture would appear narrower. So in your case when photographing a group, the members on the ends might be further from the focal plane, yet might still fall within the acceptable range of sharpness depending on how the DOF varies with the angle.

Image sharpness usually declines away from the center, and that might be a more important consideration than field curvature or off-axis DOF.

Wilt wrote in post #18481267
I have seen diagrams like in this article, http://tashley1.zenfol​io.com …ature---a-practical-guide which show the DOF as parallel to whatever curvature actually exists at the 'plane of focus'. See figure 4 for this!

That Ashley/Cicala/Zenfolio article is excellent. Roger's summation at the end is probably the most valuable:

"1. Anecdotal reports suggest that wide-angle rangefinder lenses exhibit more noticeable field curvature than SLR lenses. This may be because of differences in rangefinder focusing or because few rangefinder lenses are reverse telephoto design. (Reverse telephoto designs have more lens elements, giving the designer more opportunity to correct field curvature.)

2. The actual curvature is improved to some degree, and often changed in shape, by stopping down. Of course, increased depth of field helps mask field curvature.

3. Field curvature increases as the square (3rd order curvature) or 4th power (5th order curvature) of the distance from the center of the lens, so it naturally is more severe toward the edges of a full-frame image. Lens designers know this better than us, though, and use their tricks to minimize this – which results in more complex curvatures affecting the mid range, etc.

4. A given lens may exhibit different curvature at different focusing distances.

5. Taking a few controlled photographs and evaluating them with a bit of pixel-peeping will give you a good idea of the field curvature in your lens.

6. The lens’ MTF chart, particularly if it shows a ‘dip’ in the mid range or widely separated sagittal and tangential MTF, may hint that a lens has field curvature.

Now the truth is that the majority of lenses these days exhibit very mild field curvature, if any. But there are certainly a few that do and those few tend to be superb wide-aperture prime lenses. If you have one or two of those, taking a few minutes to evaluate the pattern of its field curvature (and I still prefer the term ‘area of best focus’) can improve your photography."

mdvaden wrote in post #18481251
Usually I shoot individuals, couples or a few people. Rarely groups. But I see myself quickly headed for large group portraits. The question came to mind tonight whether the field in focus (DOF) is a linear straight line or field perpendicular to the lens, or whether it's curved? Suppose a certain aperture offered a DOF 4 feet deep, would a group of 30 people standing side to side need to be arranged in an arc or a curve? Or could they be in a straight line perpendicular to the lens?

I've only done one image with that many people, but never gave the shape of the DOF much thought. If the area in focus happens to be curved, are there lenses designed to compensate, or are there references to know the difference, one lens from another?

At the scale you're shooting and using, it doesn't matter if it's linear/curved. For the intent and purpose at this scale, consider it linear.

You will never need to ask a group of 30 people to stand in an arc for depth of field purposes at these scales with today's equipment.

Very best,

MalVeauX wrote in post #18481352
At the scale you're shooting and using, it doesn't matter if it's linear/curved. For the intent and purpose at this scale, consider it linear.

You will never need to ask a group of 30 people to stand in an arc for depth of field purposes at these scales with today's equipment.

Very best,

So in other words, when I end up at a wedding needing to photograph 15 to 30 people, just line them up and don't get too concerned with curves or lines.

mdvaden wrote in post #18481419
So in other words, when I end up at a wedding needing to photograph 15 to 30 people, just line them up and don't get too concerned with curves or lines.

True.

Just go outside and shoot a brick wall. Lens testers do it all the time with the expectation that the plane of focus will be flat, and with modern lenses it almost always is for practical purposes.

So just go outside and shoot a brick wall. Then you'll know for your lens.

mdvaden wrote in post #18481419
So in other words, when I end up at a wedding needing to photograph 15 to 30 people, just line them up and don't get too concerned with curves or lines.

Correct.

Is it curved? Yes. But, it's at a scale that you will never notice and cannot measure visually within the photograph. It would have to be measured at a scale so small and precise that it wouldn't even be reasonable to talk about in this application. Even the curvature isn't consistent. Again, at this scale, you'd never know.

At the scale we're working at, even at the pixel level here (which is still a large scale frankly!), it can be considered linear or flat, because it's indistinguishable from a curved field. So don't be concerned about people being within depth of field from the perspective of that depth of field area representing a linear beginning or fall off area, or not.

Very best,

RDKirk wrote in post #18481436
So just go outside and shoot a brick wall. Then you'll know for your lens.

Just don't use this particular wall...

http://theheartthrills​.files.wordpress.com/2​013/01/mit_chapel-2.jpg

And when you test use a focus point on the edge of the frame so there is no confusion with optical characteristics of the lens. DoF is inversely proportional to magnification so some of the field curvature focus change is taken care of automatically.

RDKirk wrote in post #18481436
True.

Just go outside and shoot a brick wall. Lens testers do it all the time with the expectation that the plane of focus will be flat, and with modern lenses it almost always is for practical purposes.

So just go outside and shoot a brick wall. Then you'll know for your lens.

J Michael wrote in post #18483775
And when you test use a focus point on the edge of the frame so there is no confusion with optical characteristics of the lens. DoF is inversely proportional to magnification so some of the field curvature focus change is taken care of automatically.

Test against a flat brick wall at the distances and apertures you'd use for groups.

Curved or flat? Shoot 10mm at f/8. It will just be infinite.

It's definitely curved for many lenses, particularly wide-angle. However, I have only ever noticed it when trying to eek out the absolute maximum DOF for a given aperture (e.g. focus at 5ft at f11 and then get variations in sharpness at infinity). If you are focusing on the middle of the group at any reasonable aperture, you should be fine. I doubt 15-30 people could stand in a straight enough perpendicular line to make a difference, you'll need to stop down enough to take care of that first.