Hyperfocal distance - theory and practice

[northlondon43]

Hi everyone

I've just come back of my hols in Southern Africa. Whilst there I tried on a number of occasions to take landscape pictures using my understanding of the theory of the hyperfocal distance. The crude measure (IMO) is to focus roughly a third into the distance and take the photo. This didn't work in practice as my background was blurry. So I tried another two methods a) using a DoF calculator app on my phone - it allows you to input camera, focal length, aperture and subject distance b) focussing roughly 5ft away (read about this method in a photography book). In both cases I again was left with a blurry background. In the end I simply focussed on the far background and this seemed to work. I should mention I tried all methods using AF and MF.

Any thoughts on why only the latter method appeared to work?

Thanks

Adrian

[Mike Buckley]

To thoroughly understand the situation, we would need to know the accurate distance to the focus point, the farthest object and the nearest object. We would also need to know your aperture setting and your focal length.

The crude measure (IMO) is to focus roughly a third into the distance and take the photo. This didn't work in practice as my background was blurry.

That method can often work. However, certain combinations of focal length, aperture setting and distance to subject will achieve the largest depth of field if you focus halfway through the scene, not one-third of the way. That could have been your issue.

Another issue could have been that you simply didn't use an aperture that was small enough.

Another possibility that would apply in my case is that you didn't accurately determine what is halfway through the scene. The further the background objects are away from you, the more difficult it is to determine their distance. Many people tend to think they aren't as far away as they really are.

using a DoF calculator app on my phone - it allows you to input camera, focal length, aperture and subject distance

You could have inaccurately judged the distance. As I mentioned, I wouldn't be very good at doing that.

focussing roughly 5ft away (read about this method in a photography book).

That would work only with a relatively short focal length, a relatively small aperture and a relatively short distance to objects that are farthest away from you. Perhaps your situation didn't meet the correct combination of requirements.

I simply focused on the far background and this seemed to work.

In that situation, it also probably would have worked if you had focused about halfway through the scene. However, my guess is that you really focused on something closer than the farthest objects.

By the way, your post-processing software may reveal the focus point in the image. You might want to see if yours does. If it does, you might learn a lot about where you focused.

[northlondon43]

Thanks Mike...as you mention I should have said that my aperture was at f22, focal length 16mm. The nearest point was around 2m, furthest? Best guess would be a kilometer or two.

Adrian

[Mike Buckley]

For those who really want to delve into the details and perhaps offer more ideas, let us know whether you were using a full-frame sensor and, if not, what the crop factor is. A crop factor will change the effective focal length. It just now occurred to me that if you forgot to take a crop factor into account, that would have thrown off your use of the tables.

[northlondon43]

Mike
The app that I used takes the crop factor into account when you input the camera model in. Mine has a factor of 1.5.

[Glenn NK]

Thanks Mike...as you mention I should have said that my aperture was at f22, focal length 16mm. The nearest point was around 2m, furthest? Best guess would be a kilometer or two.

Adrian

With a lens FL of 16 mm at f/22, it's hard to get anything OUT of focus.

Put your camera make and settings into here:

http://www.dofmaster.com/dofjs.html

Even with a subject (focus) distance of 10 metres, the DOF range is 0.54 metres to infinity.

My first question is, "what is your shutter speed"?, followed by "are you using a tripod"? And (not being rude), if you aren't using a tripod, "why not"?

Typically my process is:

a) camera on tripod -always
b) remote release (or ten second timer if I forget the remote)
c) mirror lockup (or shoot in live view - but still use timer).

I often use 1/6 or 1/2 second shutter speeds with landscapes (with waves on water I try to keep the shutter at least 1/50 or faster - because I don't like blurry water, but that's another topic).

Glenn

[AB26]

Hi Glenn,

Thanks for the link!

[northlondon43]

Like AB26, thanks for the link Glenn; I've bookmarked it.

I'll have to look up the EXIF data to tell you the shutter speed but I'm sure it was fast relative to the focal length, e.g. 1/100th sec to 16mm focal length, to minimise the risk of camera shake. Even though I wasn't using a tripod (not always convenient to do when travellling in a car and you suddenly stop to take a photo) I do make sure that I have something sturdy to lean against.

[Stagecoach]

Adrian, any chance of posting an example with shooting data ?

[Mike Buckley]

Adrian: The information that you have provided is very perplexing. So, I agree that being able to see a problematic image that still contains the EXIF information would be helpful, especially if you have software that determines the focus point. When you post an image, some of us may have software that will determine the focus point.

[Glenn NK]

On the same site (plus a lot of other good stuff):

http://www.dofmaster.com/iphone.html for you I-phone guys

http://www.dofmaster.com/custom3.html for Palm guys

I have DL'd this and printed out tables which I carry (but admittedly rarely use): http://www.dofmaster.com/charts.html

I have some other links, but I'm on a laptop that doesn't have all my bookmarks - home tomorrow.

Glenn

http://www.dofmaster.com/charts.html

[northlondon43]

Sorry for the tardy response but been very busy this weekend.
I now realise that I deleted all the 'out of focus' photos from my card as I went along so I don't have an example to show you....apologies.

[ajohnw]

If you look towards the end of the camera list on the dofmaster site you will see circles of confusion listed. Often this is never mentioned in relationship to hyperfocal distance. Instead a circle of confusion of 0.030mm on the sensor is assumed and scaled according to actual sensor size - smaller for smaller sensors. Taking an A4 print as an exampe a 35mm sensor has a diagonal of 43mm, an A4 sheet 351mm so 8.16 enlargement at least would be needed. At this point the 0.030 circle of confusion would be nearly 1/4mm diameter. Clearly visible to some one with decent eyesight at 250mm. Oddly the limit for that on the image as viewed is about 0.030mm more or leas matched by pc screens. If the image was enlarged to A3 size and cropped to A4 the circles of confusion would be correspondingly bigger. This is why the calculator includes circles of confusion sizes. What is needed really depends on final enlargement and to some extent detail in the photo that needs to be clear. Take a photograph including an object in sharp focus and the blur due to the size of the circle of confusion assumed will be there but may not be noticed unless the image is examined closely all over. That really is what the hyperfocal distance is all about.

So best bet on a scene to the horizon is to focus just too the horizon plus depth of field preview or do some sums related to what's needed which in practice is more complicated than depth of field calculators often imply. I did find one some where that included print size and an option for 20/20 vision or some one with poorer vision because print sizes came out rather small. The other option is to increase the final viewing distance. Fortunately jpg compression takes care of some of these problems. Best illustrated on compacts. Take a photo of something and it will be clear, zoom into the detail in the background and it will rapidly get blurred to such an extent you will wonder why the picture looks ok. The jpg engine and sharpening etc looks after that. I once compared a 2mp early canon digital ixus with something much newer and a lot more pixels this way and the 2mp camera won easily. Both by canon - the newer one was a powershot. I switched make. Not that there was much point in doing that really. All are virtually the same.

[Mike Buckley]

Considering the impact that circles of confusion have on my mind, I have always felt that perhaps nothing is so aptly named.

[ajohnw]

LOL - I think they are a good example of how the web damages the mind.

Personally I think there may be a case for bringing out a digital camera with a focus and aperture ring and easily interchangeable focusing screens. Might be able to really see what's going on then.

[Glenn NK]

Like AB26, thanks for the link Glenn; I've bookmarked it.

I'll have to look up the EXIF data to tell you the shutter speed but I'm sure it was fast relative to the focal length, e.g. 1/100th sec to 16mm focal length, to minimise the risk of camera shake. Even though I wasn't using a tripod (not always convenient to do when travellling in a car and you suddenly stop to take a photo) I do make sure that I have something sturdy to lean against.

There is more to eliminating vibrations than using a fast shutter speed.

The first link is to my drop box site and it's a pdf file you can DL and save:

https://www.box.com/s/q1yj0jxecmudpd9i4vbc

This second link gets to the point very quickly:

http://www.luminous-landscape.com/tu...ror-lock.shtml

My first choice is to use mirror lock-up with a remote release; second choice is to use live view with the timer. I don't feel that the two second timer is quite long enough, so usually use ten seconds.

Glenn