A Question about Front Focus and Hyperfocal Distance

[xpatUSA]

Near where I live, there's an abandoned farmhouse with rickety outbuildings on a hill, all about 100-200 yards from the road. I would probably use f/8 on a 1.7 crop camera. About 50mm focal length would take in the scene from that distance (haven't tried yet)

I've read a lot on the net by folks who get their knickers in a twist if their lens front-focuses even by inches!

On the other hand, DOF literature mentions the hyper-focal distance and gives a range over which stuff is "in focus". So if I plug the numbers from above into my up-dated calculator, I get:

Hyper-focal distance: about 19 meters

If I set my focus to more than 19m the DOF on my calculator goes berserk as one would expect with a simple calculation with no bounds.

If I set the focus to 18 meters, then the DOF is about 370m, enough for my purposes.

In view of the line above, why would I care about a few inches, or even a few feet, of front or back focus at that distance?

I'm saying that those inches or feet would be within the Acceptable Cone of Confusion and would be acceptably "in focus" in terms of DOF.

Of course, I could just focus on the buildings themselves - but then I would I lose some foreground detail would I not (it is after all a scene with foreground detail with grass, stones, shrubbery and stuff)?

Help would appreciated!

[Shadowman]

Hi Ted,

Perhaps those that complain typically shoot wide open, however from the front/back focus examples I've seen the problem occurs when the focus happens on an object other than the intended subject and the actual subject appears soft. I had a similar incident happen to me but I don't think it was actually front/back focus. My focus indicator was locked dead on my subjects eyes yet she appeared soft and other individuals who were almost parallel to my subject had adequate sharpness; I attributed the softness to three things: my subject was moving, the other individuals weren't and I was using 3D tracking.

[Stagecoach]

In view of the above, why would I care about a few inches, or even a few feet, of front or back focus at that distance?

I doubt anyone would care at the shooting scenario you mention.

If you shoot with a long lens at wide apertures, front and rear focusing errors can be a significant disadvantage to your results.

[xpatUSA]

Hi Ted,

Perhaps those that complain typically shoot wide open, however from the front/back focus examples I've seen the problem occurs when the focus happens on an object other than the intended subject and the actual subject appears soft. I had a similar incident happen to me but I don't think it was actually front/back focus. My focus indicator was locked dead on my subjects eyes yet she appeared soft and other individuals who were almost parallel to my subject had adequate sharpness; I attributed the softness to three things: my subject was moving, the other individuals weren't and I was using 3D tracking.

Certainly food for thought, thanks John.

A certain manufacturer sells a USB device for their SA mount lenses which adjusts AF for 4 different apertures and (zooms) focal lengths; I know little about that side of things being "always in manual".

[xpatUSA]

I doubt anyone would care at the shooting scenario you mention.

That's what I was thinking, thanks Grahame.

If you shoot with a long lens at wide apertures, front and rear focusing errors can be a significant disadvantage to your results.

Good point.

[pschlute]

I doubt anyone would care at the shooting scenario you mention.

If you shoot with a long lens at wide apertures, front and rear focusing errors can be a significant disadvantage to your results.

Or indeed shoot with a shorter lens at close focusing distance it can have a major impact. With portraits for example getting the eyes sharp is paramount generally, so a FF/BF issue with the camera can ruin a shoot. One inch of FF/BF would be hopeless. Luckily most DSLRs today have an in-camera adjustment that can be done with multiple lens combinations to correct for any autofocus deficiencies.

As for Ted's barnhouse picture, the old fashioned DOF scales that all lenses used to have prominently etched on them would be ideal for establishing the best focus point. Sadly they seem to be an afterthought on many lenses today.

[Stagecoach]

Or indeed shoot with a shorter lens at close focusing distance it can have a major impact. With portraits for example getting the eyes sharp is paramount generally,

Exactly.

A head shot taking up 80% of the frame (portrait orientation) shot with a 1.6 crop camera at f/8 only has a DoF of 52mm, whether using a 75mm or 300mm lens. As you say, one inch out is significant.

Whilst I do not generally do portraits I tuned my long lens specifically to deal with capturing both stalk eyes on crabs the best I could

[pschlute]

Whilst I do not generally do portraits I tuned my long lens specifically to deal with capturing both stalk eyes on crabs the best I could

I do dog portraits for myself and friends. The long snout often means I have to use f11 if I want to get eyes and nose both acceptably sharp.

[Manfred M]

The advice I was given when I first got into "serious photography" in the old film days was to add one more stop to whatever the hyperfocal or depth of field tables suggested as a "factor of safety". It's a rule I still apply today.

In those days, having that "rule" did two things for the photographer:

1. We were generally "eyeballing" distances and it compensated for us guessing the distance wrong; and

2. It took care of any "errors" in the DoF markings inscribed on our lenses, focusing errors on the part of the photographer and the liberties some lens manufacturers made when it came to the CoC that should have been used in determining where those lens markings were actually inscribed on the lens body.

I find that this still works well for me today, in spite of the fact that when I print, I tend to use much larger paper sizes than I use today and the screens we use are significantly larger than the 8" x 10" "default" that likely went into the lens maker's decisions on DoF markings.

Today we seem to be so data driven in a lot of our photographic efforts that we fail to grasp that modern photographic lenses are complex beasts that are not all identical and certainly NOT the lens that some lens review lab has tested. Assembling a number of parts, assuming that the parts are actually "in spec" means there will be ± numbers on every part and material that go into the lens assembly. This will be particularly true for zoom lenses and lenses (and camera bodies) that have "floating" elements used for image stabilization purposes. Parts that move will never align as perfectly as parts that are fixed.

I suspect that this is one of the reasons that fixed focal length lenses are perceived to be sharper than zoom lenses; over and above the inherent advantage of having less variables to worry about when creating the lens design. They also tend to have fewer parts and far less complex internal movements. All these factors should generally result in a sharper final image.

I have seen a number of complaints about lenses that front focus or back focus as well. Again, in my experience, this does happen, but it probably less common than most people think. I have definitely seen cases where people have come back with purported front / back focus issues that were actually camera shake or incorrect focus point (usually multiple focus points used and the camera made the wrong call on where to focus). Put the camera on a sturdy tripod and use a single focus point and the problem seems to go away in most cases.

The final comment I want to make on the subject is that DoF and hyperfocal selections are reliant on the individual scene or composition as much as the mathematics associated with the CoC calculation. There are scenes were a certain amount of foreground softness seems to work, because we don't notice it because of the how the various parts of the image draw our attention to the subject. In other images the perceived softness just pulls our attention away from our centre of interest / subject and the shot does not work. No amount of physics or math will ever compensate for how the human visual system evaluates an image.

[TonyW]

I suspect that the concern is unnecessary. I used the CiC DoF calculator and got the following results (although for a crop factor of 1.6, 1.7 not being available). At f/8, focus distance 150 metres, focal length 50mm, everything from 14.16m to infinity is acceptably sharp. This means that unless the foreground is really close, you could just focus on the building.

You have to focus quite close to make the closest point of the DoF much closer which puts a good sharpness of the building at risk. If 14 metres is not close enough to get the foreground sharp enough, you could perhaps walk back a few metres if that doesn't upset your perspective.

[xpatUSA]

I suspect that the concern is unnecessary. I used the CiC DoF calculator and got the following results (although for a crop factor of 1.6, 1.7 not being available). At f/8, focus distance 150 metres, focal length 50mm, everything from 14.16m to infinity is acceptably sharp. This means that unless the foreground is really close, you could just focus on the building.

You have to focus quite close to make the closest point of the DoF much closer which puts a good sharpness of the building at risk. If 14 metres is not close enough to get the foreground sharp enough, you could perhaps walk back a few metres if that doesn't upset your perspective.

All good advice Tony, thank you . . I think I'll stay on the building side of the narrow Texas dirt road, though - and stay between the Jeep and the fence.

[William W]

. . . why would I care about a few inches, or even a few feet, of front or back focus at that distance?

I agree with those who suggest that you shouldn’t worry – for that particular shooting scenario.

***

. . . I've read a lot on the net by folks who get their knickers in a twist if their lens front-focuses even by inches!

There’s a lot of knickers twisting on the www, much of which is unnecessary.

I have mentioned the name Nadine Ohara before on this forum; she was a very experienced Wedding Photographer and pertinent to this commentary one of the most level headed and calm photographers I have known.

By happenstance from around 2004 until around 2012 we used almost the same kit of cameras and lenses.

One difference in our kits was that she used a (fast) Sigma Wide Prime and I used a Canon.

One Wedding Photography use of a (Wide) Prime Lens with a Close Focusing Distance is for Available Light shots like this:



Nadine’s Sigma Lens, Back Focused approximately 1 inch (20mm) at the Lens’s closest focusing distance – she simply made a physical adjustment to account for this error when she was making these types of images.

Detail of Front and Back Focusing is something we certainly did discuss. It is not a big issue, nor one to make a lot of noise about – it is more about understanding the LIMITS of our tools (and ourselves) and knowing when those limits will be a concern and also having a remedy for those situations.

The illustration is to underscore those comments above, which mention the reasons for when concern about Front/Back Focusing Issues are relevant-

> close focusing distance
> lens used near wide open
> a particular element of the scene which needs to be in sharp focus


All Images © AJ Group Pty Ltd Aust 1996~2018 WMW 1965~1996
WW

[William W]

. . . Today we seem to be so data driven in a lot of our photographic efforts that we fail to grasp that modern photographic lenses are complex beasts that are not all identical and certainly NOT the lens that some lens review lab has tested . . .

plus

. . . I have seen a number of complaints about lenses that front focus or back focus as well. Again, in my experience, this does happen, but it probably less common than most people think. I have definitely seen cases where people have come back with purported front / back focus issues that were actually camera shake or incorrect focus point (usually multiple focus points used and the camera made the wrong call on where to focus). Put the camera on a sturdy tripod and use a single focus point and the problem seems to go away in most cases

Data driven + denial of user error + erroneous testing procedures =

repetitive mistakes, made in a sea of ignorance, within a poor learning environment.



Thank the Lord and all the Arch Angels - (or whatever your particular fancy of phrase) - for those internet forums comprising level-headed people, who provide solid, well founded, knowledgeable and sincere advice.

WW

[george013]

Agree with Graham, post 3.
Just adding a thought.
I just don't understand the use of the hyperfocal distance in this issue. Even if you know that hyperfocal distance you must be able to set focus to that distance. I don't know how one can set focus to a distance of 19m. Most lenses have an distance info scale but that's quite rough.
And what about when that house is just in front of several trees or another subject?

George

[Stagecoach]

Even if you know that hyperfocal distance you must be able to set focus to that distance. I don't know how one can set focus to a distance of 19m.

With the detail of, and measurement accuracy of Google Earth there's a good chance you can pre-plan exactly where you need to focus for your subject if using the hyperfocal method for Ted's example

[george013]

With the detail of, and measurement accuracy of Google Earth there's a good chance you can pre-plan exactly where you need to focus for your subject if using the hyperfocal method for Ted's example

That's what one calls spontaneously shooting.

Don't rely on Google to much. Viewing my house it says copyright 2018, yes, now already. I know those maps are minimal 5 years old.

George

[pschlute]

plus




Data driven + denial of user error + erroneous testing procedures

If I may add one more : expectation that the same camera/lens when using the same autofocus point pointed at the same subject will always focus the same !

When i am calibrating a camera/lens for FF/BF I always use a "best out of 10" test. Relying on a single attempt to AF is just not accurate

[William W]

. . . I don't know how one can set focus to a distance of 19m. Most lenses have an distance info scale but that's quite rough. . .

It is a tad short (one large step short) of the distance between wickets. Just the spot where you'd like to drop a perfect Yorker.

If you have never played cricket and/or cannot judge distances reasonably accurately, then a tape measure is a good bit of equipment to have in the camera bag. It is not rocket science. If you estimate 19mtrs and it is really 20 mtrs that will be OK - the point is knowing the limits. Better to be 20mtrs than 18 mtrs so if in doubt add one big step.

Also knowing what is the length of your "standard step" is a useful skill to acquire.

None of the above need 'google', nor batteries, nor a serviced 'uplink area' . . . yes there are still many areas in this wide world which do not have internet and mobile phone access.

WW

[xpatUSA]

Also knowing what is the length of your "standard step" is a useful skill to acquire.

At my age, Bill, my "standard step" has become variable.

I do have a 100ft surveyor's tape, though.

[xpatUSA]

Agree with Graham, post 3.
Just adding a thought.
I just don't understand the use of the hyperfocal distance in this issue. Even if you know that hyperfocal distance you must be able to set focus to that distance. I don't know how one can set focus to a distance of 19m. Most lenses have an distance info scale but that's quite rough.
And what about when that house is just in front of several trees or another subject?

George

"And what about when that house is just in front of several trees or another subject?"

Oddly enough there are a good few deciduous shade trees around the buildings, but all are within the 370m DOF that I calculated in the OP.