Hyperfocal issues

I have taken a few shots lately using the hyperfocal method on the 17-40L and 5Dc and the results have been less that satisfactory. So today I decided to test out hyperfocal vs infinity, I take it the ¬ on the distance window is infinity on the 17-40L. The hyperfocal distance was worked out with a phone app for each aperture tested. I had a piece of white paper with a black cross in the middle as my target in good light, and measured out the hyperfocal distances with a tape measure then focused on the black cross. I then took shots at f/8, f/11 and f/14 as I figured these are my most popular apertures when I want to take landscape shots. One shot was taken at the hyperfocal distance, and one set at the ¬ line on the distance window on the lens. I could use some advice as to why they are turning out this way. Here are the examples. It's pretty obvious the infinity shots are much sharper but why? These are 100% crops were from just off center to show the tree. All shots at 17mm.

Hyperfocal f/8

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

DOF calculations are guidelines. When it says that you have 10m of DOF it doesn't mean that suddenly at 10m everything starts to blur. You get blur as soon as you start to move away from the focal plane. The guidelines are based on when that blur becomes significant based on some assumptions. There are some physical elements involved WRT microsensor size etc (essentially related to the media recording the image, it has limits on what it can record) and some subjective elements (print size, viewing distance etc).
Your standards may differ.

I think you're moving the goalposts a little by cropping the image. Hyperfocal distance for 17mm at any reasonable f-stop is going to be a couple of metres at the most. There's no foreground detail in your shots so the infinity focussed ones will always appear better as all the detail is towards infinity. If you included an object somewhere near the hyperfocal distance then it would look well OOF on the infinity focussed shots but fine on the HD focussed shots.

Bob

You're seeing this because the details of your scene are nearer to infinity than to the hyperfocal distance. When you have no foreground details, you're much better off just focusing where the details are rather than using HFD.

HFD is more useful when trying to keep a foreground element in focus and still retain decent details in the background. Even then, I find it's better to set HFD for one stop larger than you are using to keep the horizon in focus. For example, if I'm shooting at f/11, Ill generally use the HFD for f8 as my starting point before going into Live View.

An issue affecting both DOF and hyperfocal distance, that most beginners fail to understand, are the variables which affect DOF calculations...

1. The size (degree of magnification of the original image) of the print makes a difference
   
2. The distance at which one views the print/image makes a difference
   
3. The use of 'manufacturer standard' vs. '20/20 normal human visual acuity' -- which optometrists try to achieve -- makes a difference
   

For APS-C camera with 50mm lens at f/4, for example...

• 'manufacturer standard' visual acuity says that when focused at hyperfocal distance of 102', the zone from 51'-15177' will appear to be sufficiently sharp that the brain says 'in focus' even though only the focus plane is perfectly in focus
  
• 'manufacturer standard' visual acuity says that when focused at hyperfocal distance of 308', the zone from 51'-15177' will appear to be sufficiently sharp that the brain says 'in focus' even though only the focus plane is perfectly in focus
  

For the FF/135 format camera with 80mm f/4,...

• 'manufacturer standard' visual acuity says that when focused at hyperfocal distance of 164', the zone from 82'-88945'will appear to be sufficiently sharp that the brain says 'in focus' even though only the focus plane is perfectly in focus
  
• 'manufacturer standard' visual acuity says that when focused at hyperfocal distance of 492', the zone from 246'-276964' will appear to be sufficiently sharp that the brain says 'in focus' even though only the focus plane is perfectly in focus

There is the same lack of detail as the tree comparisons in all the the frames in the foreground, I just cropped as I thought the tree showed it better. Looking at the screen fit view
rather than 1:1 the difference is noticeable between the two. I am starting to think I may be getting wrapped up in the theory and not getting enough practise....

JohnPh wrote in post #15379269
There is the same lack of detail as the tree comparisons in all the the frames in the foreground, I just cropped as I thought the tree showed it better. Looking at the screen fit view
rather than 1:1 the difference is noticeable between the two. I am starting to think I may be getting wrapped up in the theory and not getting enough practise....

Congratulations, you have taken the first step to understanding how to optimize sharpness in your images (not being sarcastic). It's neither as simple as stopping down nor using hyperfocal distance. It completely depends on your scene, what you want the sharpest and how non-sharp the rest of the image can be.

Also you will now understand the power of focus stacking, even at f11 or f14.

Of course, one could say that if you were just looking at uncropped images on your monitor or printing no more than 8.5x11, you would not actually see any appreciable difference in the images you posted.

ejenner wrote in post #15379880
Of course, one could say that if you were just looking at uncropped images on your monitor or printing no more than 8.5x11, you would not actually see any appreciable difference in the images you posted.

This is the key point. There are assumptions in those depth of field calculations regarding the size that you will be viewing the image. You broke those assumptions by pixel peeping. The definition used in the DOF is "the depth that is acceptably sharp", and what is acceptable depends on many things, but especially the size of the final image (combined with the distance you view it from).

I use the 17-40 a lot, mostly at 17mm, and I like to make large prints. When I need foreground and infinity both sharp, I will shoot multiple images focused at a variety of distances, so that I will later have choices as to which focal point gives me the best compromise between foreground and infinity sharpness and detail for that particular scene, as content will also play in to the choice. Doing this for a while will give you a good feel for where to focus for good results.

I find myself using f/11 when possible, and if I want a particular foreground element to be sharp, it often works out that if I focus with the center point on that element, then recompose so that the focus point is effectively farther than the element now is from the sensor, that works out to be a good compromise.

John
Thanks for posting. Good to know.

Mike

Thank you all for your replies, makes much more sense now