Hyperfocal - Fact or Fiction

After seeing a couple of threads to this topic I decided to see for myself.<p>

 

Equipment : Nikon D70, Nikkor 35mm f2 AIS (at f22), JPEG Normal Medium<p>

Other parameters : 0.3 radius and 180% smart sharpening to both images<p>

 

Distances from camera:<p>to foreground - small tree, 1metre<p>midground church,

approx 800m<p>to background mountain, approx 5km<p>

 

<a href="http://www.photo.net/photo/5335813" >HERE</a> is the

image using hyperfocal point (I used the DOF scales. I hope it means the same

thing)

<p>

 

<a href="http://www.photo.net/photo/5335814" >HERE</a> is the

image using infinity focus<p>

 

<a href="http://www.photo.net/photo/5335811" >HERE</a> is a

comparison of the foregrounds<p>

 

<a href="http://www.photo.net/photo/5335812" >HERE</a> is a

comparison of the midgrounds<p>

 

<a href="http://www.photo.net/photo/5335810" >HERE</a> is a

comparison of the backgrounds<p>

 

I make no public conclusions as to what are and what are not facts. I just hope

that the files are sufficiently sized for viewing. I recommend all those who

are interested in this topic to do there own tests and make their own

conclusions. The exercise was worthwhile.

Your excellent test shows a couple things rather well. Depth of field is not an on/off or sharp/fuzzy thing but a gradual worsening of focus as you go away from the plane of sharp focus. When you start to see this degradation depends on several factors like degree of enlargement, viewing distance and personal standards. Lens DOF scales are engraved for film use and average print sizes viewed reasonably far off for their sizes. Those using a 1.6x crop camera will be enlarging the image more than average so they will benefit less from the DOF available. Pixel peeping on screen is like examining 20x30 prints at less than arms length. You can do it but any faults will be magnified. If we allow a little over a stop lost DOF to the crop factor and a couple stops to the desire to make a huge blow up, we see that the marked DOF scales are pretty useless unless read carefully. For my personal use, I apply a 2 stop fudge to marked DOF. That means I go by the DOF marks for f/4 and shoot at f/8. I never print over 12x18. If you are wanting to see nothing wrong on screen you might want to go to the f/2 marks. We come back to the sad fact that there is exactly one plane of best focus and you might just want to use it for whatever is important in the photo.

Maybe I am missing something here, but what exactly is this test showing ?

 

DOF for 35mm lens at f22 set to hyperfocal distance of 1.86m is between 0.94m and 111.15m.

If the lens is set to 1.9m, DOF stretches from 0.95m to infinity.

If the lens is set to 1.82m - DOF is between 0.93m and 47.54m.

 

Now please show me DOF scale on the lens which allows you to distinguish between setting the focus distance at 1.82, 1.86 and 1.92m. In other words: it is very easy to set the focus at a distance slightly shorter than hyperfocal, thus negatively affecting sharpness at mid and background (my guess it is what you are referring to). It pays to focus the lens at a distance larger than the hyperfocal distance, since it only marginally affects foreground.

 

Or is there something I am missing here ?

Andy, would you please reshoot the same scene at f/5.6 and f/8 and f/11.

 

Then we would know where diffraction sets in on that lens. The infinity focused mountain range crop clearly lacks sharpness because of diffraction.

 

Do use a tripod, please, too. Thanks!

Interesting test Andy! I second Doug's suggestion about shooting at a smaller aperture than was used to set the DOF...for extra insurance. As your tests show, hyperfocal focusing is a way to produce "acceptible sharpness" throughout a selected DOF, at a normal print-viewing distance. It doesn't magically make everything "tack-sharp" throughout the DOF!

 

Sincerely,

 

Dave

Somehow I knew that the many hours I spent doing something that I thought would be useful would turn out to be something that someone wants to attack me with.

 

Frank, do you think that anyone who knows anything about anything would NOT shoot it on a tripod and use a remote shutter release? And the only shot I would have to "redo" is for f5.6 since I also shot at f8, f11 and f16. If you want to pay me say 30 Euros an hour then I will happily redo all the shots, including the one that I "forgot" for f5.6.

 

The whole point of the exercise was to encourage everyone else to just go out and do it and find out for them selves. As I already said, it was a worthwhile exercise.

 

Lezek, if you can't see any subtle differences, then I am afraid you have missed something.

 

If anyone else wants to attack me, be forewarned. I am now seriously P*ssed off.

<i>If anyone else wants to attack me, be forewarned. I am now seriously P*ssed off.</i>

 

<p>Jeez - who pissed in your cornflakes this morning? All people are pointing out are flaws in your test. If you want to post something of consequence, then try to get it right. If not, no big deal, but you have no right to get your knickers in a twist just because people point out the flaws. And using f/22 on a D70 is a flaw for the purpose of this test.

I'm still not sure just exactly what the exercise showed, myself, or why you'd be seriously PO'd about it, either.

 

DOF scales on a lens are based on someone's estimation of average image size, viewing distance, definition of what is or isn't sharp, etc. If you blow your pictures all up to poster size, those DOF scales aren't going to hold. If you reduce all your shots to postage stamp size, then those scales aren't going to hold. If you do the equivalent by zooming in digitally, those scales aren't going to hold. If you're doing digital sharpening, I have no idea what that does to a DOF scale. If you're pickier than the average photographer about sharpness, those scales aren't going to hold. If the DOF scale on a lens was intended for a full-frame image, it won't be the same for a reduced sensor- as the magnification in the enlarging process is changed.

 

It looks to me like there is some, but not a lot, of difference between the two focus points at infinity. There is somewhat more difference in the close-ups. But all in all, it looks like hyperfocal gave you a better balance of focus through the depth of the scene than infinity focusing. If that's what you're trying to prove, fine, you did it. If you were expecting infinity to be identically sharp using hyperfocal focusing as it was using infinity focusing, then fine, you've shown it isn't.

 

If you get pickier about what's in focus and what isn't, it'll simply move the nearest-focus point and the hyperfocal focusing point out a fair bit. The concept will still work, just the numbers will change. In this particular shot, it will mean either putting up with out-of-focus elements in the foreground, or else recomposing the image so there is nothing near the camera.

 

In the past, I've seen people go into great detail about their theories of DOF, especially pertaining to view cameras. While I don't remember details, one of the conclusions was that normal DOF tables are just entirely too liberal. I suspect that's simply the case of lens manufacturers trying to deduce what the "average" shooter does, versus the exacting shooter.

Richard, do know the history between me and Frank? Do a search.

 

As for any flaws, I didn't submit a PHD thesis for people to dissect. And I will say it again, the purpose of the post is to encourage those interested in landscape photography to find out for themselves what they prefer.

 

If anything I was trying to show (what Frank is a big fan of) that focusing at infinity is an excellent method. I deliberately used a very short foreground anchor to make it hard for the image focused at infinity, but it performed exceeding well.

 

So Richard, this time I will leave it at this.

Andy, Andy, you keep grudges a long time; you must have a good memory. I do not remember ever interacting with you here:

 

Andy just wrote: "Richard, do know the history between me and Frank? Do a search." So what did I do to offend you ? I am not aware of anything personal. Sorry, if I did.

 

But f/22 is the giveaway for problems with diffraction and sharpness. And how did those f/8 examples look, anyway, regardless of your hate of me?

 

This is applied science, right? Can we stay objective here? I wish we all could.

Here are the links<p>

<a href="http://www.photo.net/photo/5336814" >f8 Infinity</a> <p>

 

<a href="http://www.photo.net/photo/5336817" >f11 Infinity</a> <p>

 

<a href="http://www.photo.net/photo/5336816" >f11 hyperfocal</a> <p>

<a href="http://www.photo.net/photo/5336819" >f16 Infinity</a> <p>

<a href="http://www.photo.net/photo/5336818" >f16 hyperfocal</a> <p>

 

You got me with f8 at hyperfocal. I admit I don't have it. After taking it f8 for infinity focus, I checked the hyperfocal distances, and at 7.7 metres it made no sense to me to make a comparison. The tripod - Manfrotto 3021 PROB was already setup at 1 metre to the foreground anchor. OK?

Andy, thanks for sharing. Whether or not anyone agrees with your conclusions, they should appreciate the effort that went into it.<P>

 

And before we reject anything, I would suggest that we all go to the virtual library and read up on exactly what Andy has tested. Sorry, Andy, but you aren't covering any new ground here.<P>

 

Harold M. Merklinger wrote a book on exactly this subject (infinity vs. hyperfocal) called <U>The Ins and Outs of Focus</U> back in 1990. The book is out of print, but he has been generous enough to prepare a PDF version which he offers online. (He is willing to accept a small gratuity if we find it useful.) In the book, Merklinger not only presents sample photos, but mathematically derives the results to be expected for various subject distances when the lens is focused at either the hyperfocal distance or at infinity. His results and conclusions are certain to amaze some of the folks here, though perhaps not Andy.<P>

 

Here's the link to <a href="http://www.trenholm.org/hmmerk/download.html">the download page for <U>The Ins and Outs of Focus.</U></a> The actual link to the PDF file is in the last paragraph on the download page.<BR>

<P>

If that isn't enough, read a few of Merklinger's articles that were published in <I>Shutterbug</I> over the years. In part II, he specifically discusses using infinity focus over hyperfocal:<P>

<a href="http://www.trenholm.org/hmmerk/SHBG01.pdf">"Adjusting Depth of Field", <I>Shutterbug</I> Vol 20 No. 12, October 1991.</a><BR>

<a href="http://www.trenholm.org/hmmerk/SHBG02.pdf">"Adjusting Depth of Field Part II", <I>Shutterbug</I> Vol 21 No. 7, May 1992.</a><BR>

<a href="http://www.trenholm.org/hmmerk/SHBG03.pdf">"Adjusting Depth of Field Part III", <I>Shutterbug</I> Vol 21 No. 8, June 1992.</a><BR>

<a href="http://www.trenholm.org/hmmerk/SHBG04.pdf">"Adjusting Depth of Field Part IV", <I>Shutterbug</I> Vol 21 No. 9, July 1992.</a><P>

These links and a bunch of other good stuff may also be found <a href="http://www.trenholm.org/hmmerk/HMbook14.html">here, on Harold Merklinger's web page.</a>

Thanks for that, Alan.

 

A "short cut" to copeing with Hyperfocal, DOF issues is to remember (and the author of "Ins and Outs..." mentions this) that the standard metrics for DOF, HyperFocal have no concern for specific subject matter, composition. The brain seeks things in real life to find detail. So when you are looking at that shrub 7 feet away, then to the horizon, the normal human eye will see each in critical focus. The fixed plane camera will likely not.

 

In respect to Mr. Aungthwin's effort - it does illustrate quite well how the metrics make no allowance for subject and human expectations.<div></div>

Looking at the two f/11 pics, I can count the number of windows in each of the houses; I see the trees clearer at the far hill side when at infty focus. The unsharp fir tree does not bother me as much as the tree soup on the slopes and the vanished windows do at the hyperfocal distance pic and f/11. Infinity focus wins, or so it seems here.

 

At f/16, the objects far away have already become soupy and so so in either focus mode.

 

Of course much depends on the lens's resolution power. I do not know whether the 35/2 AIS is a champ at resolving. I would have preferred to use the 28/2 AIS as the test lens and it resolves best near f/4.

 

So, for deep and clear far object, I would use wider openings and infinity focus. But then why do I want to have the illusion of houses with windows and trees with demarcations 1 mile or two away in landscape pictures. Pinholes do give wonderfully moody pics, too ...

 

In the end it is a personal, arty type decision which rules: sharpness at the end of the view or first object focus.

Andy - yes, I see differences: with hyperfocal setting, the foreground is a tad sharper, the reverse holds for midground and background.

 

What I was trying to say is that you can't set hyperfocal distance reliably on a 35mm lens using the distance markings on the lens - you would need to take a measuring tape, set some object on hyperfocal distance and focus on it. The accuracy of distance markings on the lens simply is not good enough - which I tried to illustrate by showing differences in DOF with setting the focusing distance 4cm below and above the hyperfocal distance.

 

It is usually better to set the lens to more than hyperfocal distance to avoid loss of sharpness in the background.

 

If you managed to set the lens precisely to hyperfocal distance using the markings on the lens, with the accuracy of 4cm - then kudos to you, because I can't do that (but then I suspect neither can you).

 

Still, thanks for sharing.

I think we look to the numbers to try and obtain an understanding of the limits of the equipment. This is a good thing. It generally helps. But practicality comes into the equations. Closing down a couple of stops for comfort...and then getting back to the image we are trying to make. Maybe rocking your focus to gain the best at that depth of field. Using our eyes, along with the calculations and the camera.

 

And smiling.

Thank you everyone for taking the time for participating in this post.

<p>

I guess on reflection I went temporarily insane and therefore apologize to the forum members. All I can say is that sometimes it happens. I promise I won't do it again. Well, at least until next year. :)<p>

 

It is ironic that I started the thread based on a recommendation made by Frank on Mr Harold M. Merklinger's articles which I have read, but appreciate the extra links that have been provided.<p>

 

I will finish by inviting you to look at <a href="http://www.photo.net/photo/5339182" >THIS</a>

and <a

href="http://www.photo.net/photo/5339174" >THAT</a>

. (best viewed larger)<p>

 

They were taken this morning (yes, I actually do more than to take pictures of irrelevant looking dwarf-sized pines as my foreground).<p>

 

One of the images was f9 at hyperfocal, and the other is f13 at infinity. By now we should all know which is which, but as most will probably agree, it is more of a personal choice, and to stress that point I have post processed them slighly differently.<p>

 

When I showed the wife and asked what she thought, she said (of the foreground) "Do you realise that they look like bear foot prints?"<p>

 

Hyperfocal and Infinity very quickly took a back seat.

Alan, thanks for the links- I think the Shutterbug articles are probably what I remember reading in time past.