16-35 F4 and hyperfocal distance etc etc

I've tried to be good this week. I've been using ND grads to create better on camera images, I've been using my monopod and tripod. So far so good and I can definitely see the benefits of doing these things when I'm not being lazy.

Then I got onto my 16-35 which should be pretty sharp and certainly has been on occasion. I think the problem is with landscapes and where to focus, especially when you have close by foreground objects. So tonight I tried some tests on a landscape with a closest object 2.5m away. The hyperfocal distance comes up at around 1.75m at F11, however I get better overall sharpness focussing on the object 2.5m away. I'm really not sure why that is. Also, the scale on the lens doesn't indicate that distance so it's guesswork.

For landscapes without close foreground subject it's less of a problem and I generally just focus around 1/3 into the scene.

So what do you do in these situations?

Sent from my iPad using Tapatalk HD

Back in the day, our prime lenses had the aperture markings (lines to extend to the distance scale) so we could see and set hyperfocal easily. We gave that up for zoom lenses.

2.5m is half again past 1.75m, which is not an insignificant focusing error.

There are no sharp border lines in DOF, focus just gradually fades away until we hit the arbitrarily computed number, where we draw a line, but nothing is actually different on either side of that line. The farther you get from correct focus, the worse it gets. Saying, DOF is a limit, and saying precise focus is only possible at one distance, and at that distance, a point source will show as a point. If not at that distance, a point shows as a larger blurred circle (called Circle of Confusion, CoC). The farther, the bigger. The smaller your sensor, the more this circle will be enlarged to view it. The DOF number is arbitrary, since the CoC limit we choose is arbitrary, and also a standard viewing distance is assumed, typically is an 8x10 print (about 9x for 35mm film) viewed at 10 inches. DOF is simply about the magnification at which we view this blur. It is NOT a precise number, more a vague concept.

I don't know your sensor size and focal length (I assume D800 and 24mm ?) But that does say f/11 hyperfocal is 1.72m, thus the zone from 0.86m to 2763m (1.7 miles)... calculation based on vague assumptions (implying, maybe a little bit silly).

Hyperfocal assumes a "sharp enough" zone from half the hyperfocal distance to infinity. But if 1.7m, but if you have nothing to show at 0.86m, and you do want want to show 2.5m, absolutely nothing wrong with focusing at 2.5m, which then f/11 says 1 meter to infinity. Seems very adequate. You are not necessarily seeking the absolute maximum. You are trying to fit it to your situation, and less can be more.

If you back off to f/5.6 there, then calculator says 2.5m is 14% into that depth zone (about 1 to 7 meters). Dunno, I always imagined it was 1/3 for landscape scenes (and about 1/2 for closer distances, certainly for macro).

WayneF said:

I don't know your sensor size and focal length (I assume D800 and 24mm ?) But that does say f/11 hyperfocal is 1.72m, thus the zone from 0.86m to 2763m (1.7 miles)

Click to expand...

Correct on both counts Wayne. I think the reality is my expectations exceed physics which is common for me. Whilst I was playing I tried F8-F16. F16 was definitely the best even though some would point out that I was now suffering from diffraction!!



Sent from my iPad using Tapatalk HD

Geoffc said:

Correct on both counts Wayne. I think the reality is my expectations exceed physics which is common for me. Whilst I was playing I tried F8-F16. F16 was definitely the best even though some would point out that I was now suffering from diffraction!!

Click to expand...

Some would. But this concept of pixel size being smaller than the diffraction disk is one of my pet peeves. I think the comparison is stupid. I'm not arguing with you Geoff, I'm just arguing.

Best I can tell, the notion must imagine the diffraction disk is somehow perfectly centered on the pixel. But it completely overlooks that the digital pixels merely digitally sample the lens image (not unlike a scanner samples a paper print). The pixels merely try to reproduce the lens image the best that it can. Pixels do not create the image, the lens does. More and smaller pixels do have more resolution, used simply to reproduce the digital copy of the lens image, just a little better copy of what already exists. Whether that lens image is perfect, or somewhat less, the best job the pixels can do is to copy it well. If that includes diffraction, so be it... but it is good to reproduce what is actually there. Larger pixels (to hide diffraction in a less resolved copy?) certainly would be counterproductive to the resolution of that copy.

IMO, just dumb. Novice geeks making up psuedo-techie junk. No matter what your pixel size is, the diffraction disk is what it is. And the best digital can do is to reproduce it faithfully. Little compact cameras really suffer diffraction at anything past f/4, but that is not due to pixel size. It is their tiny sensor requiring very short lenses for a decent view, so (focal length / diameter) means a tiny aperture diameter. And the greater enlargement (of CoC) to see their tiny sensor image.

When appropriate, we make a big mistake avoiding f/16 (meaning, when warranted), and even f/22 for extreme circumstances (macro), and f/32 for telephoto lenses. Sometimes it can help more than the diffraction hurts. The diameter of f/32 on a lens 4x longer compares to a four stops less on the shorter lens anyway (Airy disk diffraction depends on the diameter, not the f/stop).

This is why our lenses have these settings on them. They can be useful.

I think we are on the same page with diffraction Wayne, I was being a little ironic. In my little test F16 actually looked the best to my eye and I was not aware of diffraction. Maybe I just don't know what I'm looking for so it's unable to spoil the image.


Sent from my iPad using Tapatalk HD