Sudden, seemingly erratic, jumps in depth of field

[peterB1966]

I got taught that the general rule of thumb is that if your zone of sharpness in front of the focal point is X, then the zone of sharpness behind will be 2X.

I realise there are instances where the DoF will suddenly jump to hyperfocal distance, but those tend to happen at the extreme ends.

But occasionally I will encounter an odd situation where the DoF does not do what I expected it to do. Take the following example:

Shooting with a Nikon D610 (CoC of 0.03) at 24mm, F4:

• at 1 meter, it is sharp .17 in front and .26 behind
• at 2 meters, it is sharp .58 in front and 1.4 behind
  
  - all good so far, this roughly conforms to the general rule.
  
  
• at 4 meters, it is sharp .1.8 in front but suddenly jumps to 19.3 behind
• at 5 meters and above it is infinite behind

Why does it suddenly change in that way? It seems erratic, and difficult to predict when in random, constantly changing situations like street photography? (I see a similar odd jump when changing the F-stop from 1 to 2 to 4).

 

[gone]

Rules? Well you can forget that right now, there aren't any. If your lens is clean and everything is in order, perhaps what you describe is it's "normal". Fortunately, they sell all manner of different lenses, so we get to pick the ones we prefer, and w/ the DOF that we prefer.

If you don't have a DOF preview, or even if you do, it also helps to make notes initially so that you know what's up. Every lens is going to have it's own DOF at it's different apertures and focal lengths anyway, they're not all designed and built alike. It's probably best to forget all those x's and y's thangs and just go by what things look like when you've made enough exposures to have a clear and concise data base. Me, I use the same lens almost all the time, so it's easy. I know very well what it is going to do before tripping the shutter.

 

[Dan Fromm]

OP, the general rule you cited is, as momus said in post #2 above, is false and stupid.

Learn about hyperfocal ... Learn how to read your lenses' depth of field scales, if present. Learn basic photography, i.e., how aperture setting affects DoF.

 

[Bronson Dugnutt]

DoF is not at all a linear function of subject distance, but rather quadratic: https://en.wikipedia.org/wiki/Depth_of_field

The len's hyperfocal distance is a boundary case in which the DoF tends to infinity and is a good next step in understanding optical equations relevant to photographers. Your rule of thumb is probably from a misapplication of a hyperfocal distance's consecutive depths of field: https://en.wikipedia.org/wiki/Hyperfocal_distance

The rule you were taught is only true for very limited areas of the function that approximate linear behavior. There's also a quadratic term in the DoF equation that accounts for the lens focal length, the impact of which should intuitively make sense if you compare the behavior of a 24mm vs 200mm lens with regards to DoF.

Not exactly rocket science if you grok basic algebra, but try fiddling with this calculator / graphing tool to get a better view of a given configuration's optical properties: https://www.pointsinfocus.com/tools/depth-of-field-and-equivalent-lens-calculator

 

[wiltw]

1. The Depth of Field grows progressively deeper as the subject distance gets farther away...there are no sudden 'jumps' in DOF.
2. And if you change FL while standing in the same spot, the DOF does change inversely proportional to FL...shorter FL yields deeper DOF, and longer FL yields shallower DOF
3. But if you change subject distance so that the subject is identical in size in the frame for all shots regardless of FL used, the resultant DOF is about THE SAME for all FL!!!

If you play with DOF calculator programs, all of the three points above will be proved to be true.

Point 1

• 25mm f/2.8 focused at 15' has DOF zone of 9.1'
• 25mm f/2.8 focused at 20' has DOF zone of 10.7'
• 25mm f/2.8 focused at 25' has DOF zone of 12'
• 25mm f/2.8 focused at 30' has DOF zone of 13'
• 25mm f/2.8 focused at 35' has DOF zone of 13.8'
• 25mm f/2.8 focused at 40' has DOF zone of 14.6'

Point 2

• 25mm f/2.8 focused at 20' has DOF zone of 144'
• 50mm f/2.8 focused at 20' has DOF zone of 9'
• 100mm f/2.8 focused at 20' has DOF zone of 2.2'
• 200mm f/2.8 focused at 20' has DOF zone of 0.53'

Point 3

• 25mm f/2.8 focused at 10' has DOF zone of 10.7'
• 50mm f/2.8 focused at 20' has DOF zone of 9.1'
• 100mm f/2.8 focused at 40' has DOF zone of 8.8'
• 200mm f/2.8 focused at 80' has DOF zone of 8.7'
• 400mm f/2.8 focused at 100' has DOF zone of 8.7'

 

[peterB1966]

Thanks for the feedback, everyone.

I apologise for using the term "rule" - by qualifying with the word "general" I wanted to give the impression it was as much a "rule of thirds" i.e. not at all, but of some use as a starting point. Nonetheless, I'm surprised none of you are familiar with the newbie explanation about front focal area being half the range of the back focal area, just to help people understand where to focus for a landscape, i.e. 1/3rd up on the picture plane. Not hard and fast, but a good start for a beginner!

But of course there are exceptions that I haven't quite been able to wrap my head around in a way that I can use to predict focal behaviour. The suggestion that it is down to quadratic progression makes sense, and may well provide me with the handle that I was looking for.

 

[Nodda Duma]

Here is a sticky thread I created providing some details on depth of field for portrait distances and different formats / focal lengths. The explicit equations to determine DoF are not found in there, because they are difficult to type in the forum. If desired, I can try tho.

https://www.photrio.com/forum/threads/charts-of-depth-of-field-vs-focal-length-scaling-f.169727/

 

[reddesert]

I'm not sure how you are deriving the "sudden" jumps - are you talking about a depth-of-field indicator, or from an online calculator, or DoF evaluated on an image?

As other people have said, DoF doesn't jump suddenly, but it is not linear. At the hyperfocal distance, DoF goes all the way to infinity, but this is a limiting case; the size of an object at infinity would be infinitesimally small on the detector. This is a clue that DoF is related to object magnification, you can look at Nodda Duma's thread for more information.

Perhaps the most hands-on way to look at the non-linearity is to take a lens with a conventional focus scale and DoF marks (such as a Nikon or other manual focus lens). You can see that the focus distance changes very non-linearly with the turn of the focus ring, but the DoF marks are evenly spaced.

 

[wiltw]

Nonetheless, I'm surprised none of you are familiar with the newbie explanation about front focal area being half the range of the back focal area, just to help people understand where to focus for a landscape, i.e. 1/3rd up on the picture plane. Not hard and fast, but a good start for a beginner!

The realiity is your 'rule' is only accurate at ONE SPECIFIC distance...every other subject distance has a front:rear split between 50:50 (at macro distances) to 0.5:99.5 at very long focus subject distances. And the split betrween those extremes is, again, a CONTINUOUSLY changing split.
This for 100mm lens at f/4...

...and your 33:66 split only applies at about 69m subject distance, and in this case ONLY for 100mm f/4
Change lenses or change apertures and the 33:66 split moves to a different distance.

 

[peterB1966]

Here is a sticky thread I created providing some details on depth of field for portrait distances and different formats / focal lengths. The explicit equations to determine DoF are not found in there, because they are difficult to type in the forum. If desired, I can try tho.

https://www.photrio.com/forum/threads/charts-of-depth-of-field-vs-focal-length-scaling-f.169727/

Yeeew, some serious weekend reading! Much appreciated.

 

[peterB1966]

I'm not sure how you are deriving the "sudden" jumps - are you talking about a depth-of-field indicator, or from an online calculator, or DoF evaluated on an image?

I guess it was a lazy description :-/ What I meant was sometimes when I am setting up a shot and expecting the rear and fore focal area to confirm to roughly 1/3 - 2/3 split, and it turns out the rear is way more, it appears in that isolated moment to be an erratic jump.

It is now clear from the other input that it is not an uncharacteristic jump, but part of a progression that I would have noticed if I had not been setting the shot in isolation.

Thanks for the input.

The realiity is your 'rule' is only accurate at ONE SPECIFIC distance...every other subject distance has a front:rear split between 50:50 (at macro distances) to 0.5:99.5 at very long focus subject distances. And the split betrween those extremes is, again, a CONTINUOUSLY changing split.
This for 100mm lens at f/4...

Thanks - I think that makes it really clear.

 

[Nodda Duma]

Yeeew, some serious weekend reading! Much appreciated.

Good for when you have trouble falling asleep.

 

[MattKing]

Good for when you have trouble falling asleep.

I'd be concerned though about the dreams that might ensue...