Hyperfocal distance

[aprotosimaki]

This may sound like an odd question but it is one that has bugged me for some time. Here is an example scenario.

Using a 5d2 and a 24-105L lens, I want to use the hyperfocal distance for 35mm at f/8. I know that it is approximately 17 feet.

So it seems to me that I have two choices.

1. Using the lens' distance markings I can manually set the focus distance to be 17 feet and take the picture. This is tricky because the markings are a little off but assume that I have set it accurately.

2. Using an object that I know is measured to be 17 feet from the lens I can use AF, focus on it and take the picture.

Should there be any difference in acceptable sharpness or is it the same thing? I guess the easiest way for me to tell is to test it out. I am just not very good at determining distance by eye so the AF approach runs the risk of mistake but at least I am guaranteed that there will be some plane of sharpness, even if I focus on an object not exactly at the hyperfocal distance.

[neuroanatomist]

Honeslty it will likely make no difference.

[awinphoto]

The theory/concept of hyperfocal distance is to get the most possible infocus range possible, assuming you have foreground and background elements you wish to capture...  Applying the technical rule of 1/3 in front of the aubject and 2/3 in back of the subject will be infocus, shooting farther out in the distance to include the 2/3 in focus reaching into infinity and 1/3 in front of your subject in focus, then you extend your in-focus range/capabilities of your image.  Odds are, in practice, you may or may not tell any difference in your overall image unless printed at full resolution or bigger, and even then it may be a wash, but if you are shooting with no real foreground elements or midground elements you need in focus as a focal point, odds are you wont notice.  A good rule of thumb if you want something in the mid ground to fore ground in focus as well as the background, put the main subject at the hyper focal distance point, focus on them and let everything else fall into play. 

[aprotosimaki]

Thanks for the replies. I tried setting the hyperfocal distance manually and then using AF, and as you both say no discernible difference between the two approaches. For the 24-105L, I find F8 to be optimal, i.e, 35mm@f/8 focusing at 17 feet. Nice depth of field with acceptable sharpness. 

[awinphoto]

Thanks for the replies. I tried setting the hyperfocal distance manually and then using AF, and as you both say no discernible difference between the two approaches. For the 24-105L, I find F8 to be optimal, i.e, 35mm@f/8 focusing at 17 feet. Nice depth of field with acceptable sharpness.

It's not even so much that there's no discernible difference, but for lets say 90% of all landscape shots taken at infinity, you probably wont tell a difference... now lets say you have a tree or a person (subject) that you want taken in crisp focus as well as the background and you want them to be somewhat sizeable in the shot and not too small, you can place that person, subject, element in relation to the camera at it's "hyper focal" distance and have both be in focus were before if shot at infinity, depending on the lens focal length, may or may not have been in focus anyways.  Some lenses like ultra wide angles where the focus scale is incredibly short from it's closest focus to the infinity point, you probably would not tell any difference... on Primes for instance where they have a pretty sizable focus scale and doesn't take much for an object to be infocus or out of focus, it may be more noticeable.  General rule of thumb, dont sweat the small stuff... if i situation arises and something in the foreground/midground is OOF and you're wanting as much in focus as possible including that something, then get your phone, computer, whatever, and look up the hyperfocal distance and apply that then.   

[briansquibb]

My rule of thimb method is to have the camera on f/11 and 5ft up and level. By focussing at the nearest point visible on the ground (manually) then I get approx the hyperfocal distance

It is a rough approximation but in a hurry and without a laser rule it is close enough 

[lol]

Hyperfocal distance is about getting as much as you can *acceptably* sharp. You need to decide what is acceptable. For example, if you want pixel level sharpness, that will be very different than if you were doing a small print.

The 1/3 in front, 2/3 behind is a long established myth. You get close to equal distribution of acceptable focus around the focal plane at close distances, and at greater distances the region behind will approach infinity. Sure there is some point where that applies, but you're unlikely to be on it. Having said that, as a rule of thumb, it might be close enough to be useful.

If you need depth of field, my tip would be to stop down as far as you dare if you can afford the shutter time. Just watch out for diffraction softening if taken too far.

[awinphoto]

Hyperfocal distance is about getting as much as you can *acceptably* sharp. You need to decide what is acceptable. For example, if you want pixel level sharpness, that will be very different than if you were doing a small print.

The 1/3 in front, 2/3 behind is a long established myth. You get close to equal distribution of acceptable focus around the focal plane at close distances, and at greater distances the region behind will approach infinity. Sure there is some point where that applies, but you're unlikely to be on it. Having said that, as a rule of thumb, it might be close enough to be useful.

If you need depth of field, my tip would be to stop down as far as you dare if you can afford the shutter time. Just watch out for diffraction softening if taken too far.

I shoot product photography for a living and can confirm 1/3 in front, 2/3 in back is indeed quite accurate... test it when you focus on items, especially small items... live view focus it, and have items in front and in back... press the DOF button... you will see more behind the focus plane fall into focus than in front.  But that's neither here or there.  Stopping down, even on 5d's, beyond F11 you are introducing diffraction.  F8 is a good compromise...  If there is something nearer to the camera that NEEDS to be in focus, focus on that or make sure it's at the hyperfocal distance... the rest will take care of itself. 

[hkhan]

There used to be a mode called Depth of field AE mode (DEP) in discontinued Canon cameras. I do not know why it was discontinued. It should be introduced again. Depth of field AE mode (DEP) allows you to select two focus points then camera takes a photo keeping everything in focus between these user selected points. Why we keep hyperfocal distance charts and manually set the distance on the lens for the aperture value, if these things can be accurately calculated by the camera’s processor.

[marekjoz]

There used to be a mode called Depth of field AE mode (DEP) in discontinued Canon cameras. I do not know why it was discontinued. It should be introduced again. Depth of field AE mode (DEP) allows you to select two focus points then camera takes a photo keeping everything in focus between these user selected points. Why we keep hyperfocal distance charts and manually set the distance on the lens for the aperture value, if these things can be accurately calculated by the camera’s processor.

Magic Lantern calculates DOF and displays it. Very useful. Sometimes there's an error when lens doesn't communicate properly to the camera (or there is a bug in the ML itself) values currently set: AF distance and focal. There is also another funny thing - values for current AF distance and focal in zooms are not communicated constantly with 1 value unit precision, ie for focal: 55mm, 56mm, 57...etc but rather by some undetermined  incremental values (not even linked with some logarithmic scale) and dependent on lens, ie: by from 2 to 5 mm in case of EF 24-105. No matter how gently you move the ring you're not able to set anything between 24 and 28 mm in this case. Distance on a  ring between reported 24 and 28 is something like 1 mm. Is it important? Calculated DOF for 24mm, f8 on 5d2 at 1m distance to target is 97cm and for 28 mm - 65cm. At 2m distance to target is 10m for 24mm and 3,8m for 28mm respectively.  It's a quite big difference and it maybe  the reason, that it's no more present in cameras and such calculations are not presented at all.

BTW: I rather consider DOF calulated than hyperfocal calculated. DOF understood as distance between the closest and farest points considered as being sharp on specific camera viewed from the average distance.
The rule 1/3 - 2/3 is not proper for all the focal and distance to target combinations.

[hkhan]

an interesting thread

http://forums.dpreview.com/forums/readflat.asp?forum=1032&message=41368427&changemode=1

[bycostello]

something that mystifies me a bit, but have an app on the phone to work it out for me...

[Marsu42]

There used to be a mode called Depth of field AE mode (DEP) in discontinued Canon cameras. I do not know why it was discontinued. It should be introduced again.

I had this on my legacy eos bodies, and looking back it really sounds like a good idea. But thinking about it, I strangely almost never used it because I usually shot with a large dof around f8 at these times and used the dof preview button to see what's in focus.

Magic Lantern calculates DOF and displays it. Very useful. Sometimes there's an error when lens doesn't communicate properly to the camera (or there is a bug in the ML itself) values currently set

You already wrote about some drawbacks of the ml implementation, another is that some (older) lenses simply don't give the camera the required distance information.