what is your tip when using manual focus under f2.0?

Maybe a tad soft, but not unacceptably so for max aperture IMO (I haven't done any sharpening to the shot at all).

100%




(Sorry! Totally hijacking this thread :blush

It's very difficult to judge properly on a small image, but if you want a proper answer you can e-mail me a full sized picture to have a look at.

Just PM me and we'll take it from there.

This shows what seems to be a massive curvature of field and
what nonsense DOF calcs are:

Lens: Kiron 28-105mm Camera K-5 200ISO
All at f/8 1/200th sec.

All shots are straight Jpgs off the camera with no PP and all
were captured by pressing the PrtSc button with the screen set to
800 x 600. Viewer = AcDSee

LHS of pic: Focused on bridge (40-50m away) by eye also green signal was on.
RHS of pic: Focused on Infinity

The left and right sides of the shots were almost identical so the
lens seems centered OK.

The centre of bridge shots are both acceptable and only included
for comparison with the right side of the bridge.

From DOF calcs everything should be in focus, but it isn't

1). Centre of Bridge




2). RHS of Bridge




Example of where DOF just isn't working.
Focused on centre of bridge at f/8 .... Far side and trees are OOF
(May not show up too good on this 800 wide pic.)
I would probably be better setting the focus at infinity.




The landscape solution with 24/28mm lenses seems to be start with
the lens set at infinity and take shots moving the focus ring by
tiny amounts back to about 15m and see which is the sharpest.

f/11 will help with the edge, but with most lenses the centre will
be worse than f/8..... maybe the half stop between the two would be
the best compromise.

A Sigma 24mm f/2.8 (recent purchase) gave similar results, but was
acceptable with the focus mark moved about 2mm back off infinity.

You say an example where DOF "doesn't work" but DOF always works. It is what it is, an unavoidable result of optical laws.

I would seriously look at your lenses if you can't make a landscape sharp front to back with a wide angle lens.

That's a great little tip - thanks !

For anyone interested in properly focusing wide angle lenses here's the link again that MattMatic posted some time back. It quite clearly shows that DOF calcs aren't reliable (possibly because the assumption that the lenses are perfect and give the same resolution at every aperture) In this case the maximum "SHARP DOF" was found to be at f/5.6 link
Click on the images to see them larger.

This article "Depth of Field Revisited" by Harold M. Merklinger from 1992 by an author who has written a book on DOF suggests focusing on the furthest point that you want in focus (could be infinity) and using an aperture equal in diam to the smallest detail you wish to resolve. There are quite complex reasons for this (not included in the article). link

I can see that there is a clear misunderstanding here about what the first article is saying, which is not new in any respect.

DOF increases with using smaller apertures, but it is also true that the resolution of a lens falls off either side of an optimum aperture, usually about two stops down from open aperture. So a good f2 lens might peak at f4 or f5.6.

There a a couple of consequences here. True, wider apertures might give sharper images, so stopping down to f22 will result in a less sharp image. But this has nothing to do with whether the lens is in focus or not. This relates to the ability of the lens to render those in focus areas sharply.

This we have a balance to achieve between using the "best" aperture for sharpness and the "best" aperture for DOF. To maximise DOF and with a good lens, it';s clear from tests that pereformance can hold up well to f11 and maybe even f16 and those are the apertures I would choose. It works well.

A further caveat is that less expensive zooms aften break the peaking rule and can actually improve down to f16. This is the sign of a less good lens, but it does at least unify our "best" points.

It's very easy to get bogged down by all this and that's why I try to talk about simple "rules of thumb" that we can use and find OK, most of the time. The more pedantic we are, the less true anything becomes.

I'm going to shamelessly steal that for a sig.

As I have said before, the example here of the 15mm f4 shows why I flogged mine (at a profit ). At open aperture it's not very sharp. At f8 it matches the kit lens, albeit with less distortion.

One of the characteristics of this lens is that if you focus it on, say, a brick wall, the edges will be sharper than the centre. I was rather unhappy with my example and swapped it for another which was exactly the same. This accorded with all the reviews I read too.

A lovely lens in every way except it is optically not that good. Put the 12-24mm on at 15mm and see the difference at f4. Not that the zoom is that great at the edges wide open!

I loved my little15mm but found I was using the 12-24 for preference because it was so much sharper. Were I richer, I'd have kept both.

My DA 14mm/2.8 is a lot better than what I've seen above in pentaxophile's copy lemmy

The photozone image sample at f4 is a lot better, so it appears I have a bad sample. I've never had much luck with pentax qc, my kit lenses have also been poor. I like the characteristics of the 15mm in other respects so I'll try at least one more copy. Not really interested in the 14mm, not really compact enough compared to my sigma 10-20. This experience makes me wonder about the nex + 16mm pancake, which is cheaper as a package than the 15mm alone.

That would make sense if you like the other attributes of the 15/4.0.

I would guess you have a bad copy. It was easy to see, even at the size you posted and regardless of PP that something wasn't right with that image. My little Sony compact with it's tiny lens and tiny sensortakes better pics

Good luck in getting a decent copy.