[FHPhotographer]

I'm trying to figure out whether I'm getting "soft" K20 images because of the camera or the new Tamron 17-50. I've done the FF/BF test, but frankly I don't know what I should be looking at. The following is at 35mm @ f/11Av and AF, on the tripod with 2sec delay. The same focus results at 50mm. Hand held the images are noticeably softer, but consistently so. This all seems reasonable to me for this lens, or am I missing something FF/BF?
Thanks,
Brian

---

Last edited by FHPhotographer; 12-18-2010 at 11:27 AM.

[jaieger]

To test for BF/FF, you need to use the widest aperture possible. This minimizes the depth of field and maximizes any inaccuracies - you should then able to clearly tell if it's focusing closer or further than intended, or if it's focussing correctly.

[Quicksand]

Also, I'd strongly recommend NOT using that particular chart.

There's WAY too much detail around the central focus target. It's impossible to know whether your camera locked focus on the central cross, or on the lines above or below it, or on the words to the left and right, or on the crosshatching everywhere else.

The autofocus sensor is considerably larger than the dots in the viewfinder, and it might not be exactly aligned with the dots either. So it's best to remove all doubt and have only a single small feature in the central focus zone.

There's a chart out there on the net somewhere that meets this description, and I'll post a link if I can find it.

EDIT:

Here's the one I have used: http://focustestchart.com/chart.html

[FHPhotographer]

To test for BF/FF, you need to use the widest aperture possible. This minimizes the depth of field and maximizes any inaccuracies - you should then able to clearly tell if it's focusing closer or further than intended, or if it's focussing correctly.

Here's the f/2.8 of the same shot,
Brian

[jolepp]

Seems pretty accurate to me.

[FHPhotographer]

Is this the chart you meant? Seems to get somewhat different results, and now showing BF ? And is this within a reasonable range or does that mean using the AF correction, and in which direction?
Brian

Also, I'd strongly recommend NOT using that particular chart.

There's WAY too much detail around the central focus target. It's impossible to know whether your camera locked focus on the central cross, or on the lines above or below it, or on the words to the left and right, or on the crosshatching everywhere else.

The autofocus sensor is considerably larger than the dots in the viewfinder, and it might not be exactly aligned with the dots either. So it's best to remove all doubt and have only a single small feature in the central focus zone.

There's a chart out there on the net somewhere that meets this description, and I'll post a link if I can find it.

EDIT:

Here's the one I have used: Nikon D70 Focus Chart

[pentup]

There's another method of checking focus accuracy, by using Live View and checking moire patterns on your computer display, described here:
AF microadjustment for the 1Ds mark III, 1D Mk3, 5D Mk2, 7D
(the idea is to adjust with Live View till the interference pattern appears strongest, then turn off live view and activate AF with the shutter button (or AF button, whatever) and see if the lens moves. )

I don't know if it's superior or inferior to other methods, but it's much quicker, and it avoids the issue of the sensor locking on to the wrong part of a busy, angled test chart.

[FHPhotographer]

pentup, well I tried it and (if I did it right) the lens/camera combo was spot on. Thanks for the link,
Brian

[ve2vfd]

I'm not a fan of charts... I by far prefer shooting perpendicular a flat target (like 3 similar boxes staggered)

A quick and dirty (but still better than most charts) method uses 3 staggered batteries, you aim for the middle one. This is one of the test I do with a lens when evaluating it in a store.

This is from my Tamron 17-50 (which needed -10 adjustment on my K20d but front focuses more than what I can adjust on my K5 and is unusable under F4 so it will be sent for service after the holidays)

This is with no adjustment on the K5:



This with -10 adjustment, better but it still front focuses:



I move the battery on the right to see where the real focus point is:



Be sure to evaluate your body with other lenses first by the way... you want to make sure that it's that lens thats front focusing and not the camera body. In my case my K5d is dead on with my other lenses.

Pat

[pentup]

pentup, well I tried it and (if I did it right) the lens/camera combo was spot on. Thanks for the link,
Brian

Glad to hear it; you're welcome, Brian. A caveat though:

What really drives me nuts is differences in AF between tungsten and other lighting, and differences in accuracy depending on the subject distance from camera -- neither of which is addressed by that moire test (but I do hope it is good for at least getting a baseline idea of the focusing performance) Past a few yards and I'm never sure what I'll get from my DA40, for example, even in daylight; I get tired of coping with different performance with different lenses and try to shoot manually as much as possible; good thing I don't shoot sports. Probably most of my problems with AF are due to inexperience and simply not liking AF very much. I hope you have much better luck!

[FHPhotographer]

ve2, thanks I'll try the "battery test" and see what that tells me. Perhaps the fault is in my expectations (or ability) rather than the equipment but I really do expect more from a top quality body/lens combination. And I certainly expect it to be equal to the quality of my former K20/Tamron 17-50. Part of the problem is there are too many variables: camera, lens and operator. If I choose to spend more to get more, I feel I shouldn't make the same compromises with a $1000 system that I would on a $300 pocket camera. In any case, thanks one and all for the good feedback, and if improvement can be found I'll report back,
Brian

[FHPhotographer]

Well, after much agonizing, I finally settled on a -7 AF correction and things improved, enough to see that the real problem lies in the general AF. Here are two new images, the first is AF and the second MF. Obviously, something's not right in the AF, either in the camera or the lens. So fix the camera or the lens?
Brian
AUTOFOCUS:

MANUAL FOCUS: