which lens to buy?

[Greg Beetham]

It almost looks like the camera did a DOF stack of a few frames agorabasta, my attention was drawn to the little chain on the brolly which I think is much closer even than the paving at the bottom, if the focus point was so far away then you would expect that the little chain would be more out of focus than it is, even at f9 instinctively, in actual practise though I might be wrong because 43mm is telephoto on APS-C (65 ish) and would compress things quite a bit…logically, one would think
Greg

[agorabasta]

Greg, that little chain was in the plane no farther than 1m away. And it's not sharp enough, but still very far from being appropriately blurred.

[Greg Beetham]

I just took a photo (camera JPEG) with as similar characteristics as I could manage with the A100 and KM18-200. The sign on the left is about four feet away and the focus point was the second pot plant in the row in front of the shop, I estimated to be about 25 metres away @ f9. Too me it looks like the sign is more blurred than the little chain was, hard too say, and the road is maybe similar to the paving in focus perhaps.
Greg
Ps. the lens extension turned out to be 45mm according to the EXIF, had to guess that, the only one close on the barrel was 50mm.

DSC03651.jpg (220.38 KiB) Viewed 3970 times

[agorabasta]

Greg, there are two comments - in your sample the close pole is as blurred in a full resized sample as a similarly close one in my 100% crop above, and then in my sample the camera was only 85cm above the ground, I was sitting with my right elbow resting on a cafe table, so the pavement is much closer.

[Greg Beetham]

Hmm, so it’s not as good as it could be then agorabasta, for comparison purposes too see if there is any anomaly with DOF, I still think there is actually, your shot has the DOF starting much sooner then mine I think, that’s the impression I get anyway. It’s almost as if the format was smaller in your case than mine, but they are both supposed to be APS-C aren’t they. I guess the only way to find out is to do a measured distance shot of something; something of known width and both using the same length lens…but what? that is the question.
Greg

[agorabasta]

Greg, we could shoot some long passage with tiled floor. Just measure tile size and have some close tile marked somehow.

Meanwhile, here is another sample at 18mm f/8. As expected, the foreground is sharp, as is the farthest background. But there's a thick chunk of blurred space right behind the focus point.

DSC00419.jpg

(219.75 KiB) Downloaded 2247 times

Now the central crop, having the focus point in its lower left corner and quite sharp tree tops in the far background seen just above the restaurant roof.

DSC00419-2.jpg

(189.91 KiB) Downloaded 2318 times

And now a crop to the right showing a piece of bridge stonework and behind that some trees with a house in the closer background - all's quite blurred.

DSC00419-3.jpg

(212.61 KiB) Downloaded 2318 times

Actually, that's a picture with very much the same problem David described with his a77+1680.

[Greg Beetham]

The impression I get from that photo is there is much more sharpness in the DOF (and more of it) in front of the focus point than behind the focus point, and that focus point is reasonably deep into the scene; it’s like the actual focus point was located much closer than the one you used....if you get what I mean.
Greg

[agorabasta]

Sure I get you Greg, and I also thought at first that somehow the CDAF is preferring to frontfocus. But the problem is that the PDAF with an adapter does the same with totally different lenses as well. And most importantly, the careful manual focus produces the very same results - the best focus falls near the farther border of apparent DoF.

And there's absolutely no explanation to the DoF splitting in two ranges with the second DoF range seen here at infinity.

[Greg Beetham]

I think the roof of the restaurant is sharper than the background treetops, but yes the background treetops do look sharper in that particular area than elsewhere, maybe they are a tad closer just there? I mean if the ridge is receding away to the right as it gets higher it might explain the difference.
Greg