[Akcelik]

i need help!

i purchased the K10D for the big bright VF for easy focusing. this is my first SLR.
i am mostly using the K28/3.5. (BTW do i need a KatzEye split prism focus screen for a f3.5 DOF???)

please help me (the noob) to determine how i might know if i have a miss aligned sensor.

it does not matter if i use focus assist or the naked eye, i am always getting Out Of Focus shots Front Focus & Back Focus.

i also use a cheap AF lens Sigma 28-80 for testing Auto Focus and that also gives OOF shots.

i have corrected calibrated the diopter and even went through the lengths of turning of AF assist green and red light and the beep and the LCD so nothing would distract.

for example..

i focus on the green led it should look like this..
http://farm1.static.flickr.com/149/413586685_c83b9a4efb_b.jpg

but most of the time i get this..
http://farm1.static.flickr.com/172/413586142_625ba42596_b.jpg

the green stub on the leaf with the arrow pointing should be..
http://farm1.static.flickr.com/170/413584861_bc4fa1def9_b.jpg

but i always get this result..
http://farm1.static.flickr.com/188/413585607_f230880883_b.jpg

when its clear in VF that does not translate what i see on the computer monitor.

how do i determine if i have a misaligned sensor?

[Adam]

Before you worry about a misaligned sensor, check the SR is either turned off or manually set to the corrct focal length (28mm).

[blende8]

Did you use a tripod?
The DOF looks very slim, if you move only very slightly, the result could be like this.

[Akcelik]

Before you worry about a misaligned sensor, check the SR is either turned off or manually set to the corrct focal length (28mm).

SR is on and set @ 28mm focal length. is that right? for a 28mm focal length that = FOV 42mm?

[Akcelik]

Did you use a tripod?
The DOF looks very slim, if you move only very slightly, the result could be like this.

no tripod. but its a 28/3.5! thats 3.5 DOF.

[jfdavis58]

You need a better understanding of Depth of Field. Google 'fCalc'; there is a small applet (free) that gives a whole lot of photo information, much of it related to your photography.

For grins and chuckles I built a large spreadsheet that does a similar set of tasks. Here is a small piece of the output showing the DoF for 28mm lens. Aperture across top, distance to subject down the left side (I work in inches and feet, so the strange distance numbers are cm converted from inches--starting at 8 inches to subject).

The pink area is your total DoF in cm-very small; it's NOT constant!

[Lowell Goudge]

something to think about.

Depth of field is somewhat subjective. There is only one point of absolute perfect focus and the depth of field is a range of focus errors where the resulting image quality is deemed acceptable.

I do not know the historical background of the definition of "acceptable"

[jfdavis58]

I do not know the historical background of the definition of "acceptable"

Too bad-it's mind blowing!

It's based on something like this: the size of a defect spot on the film/sensor plane that will show on the normal print at normal viewing distance as a distinct ring (circle in circle of confusion) [as opposed to a dot].

Then there is some math to relate that size to two distances; one in front and one in back of the normal focusing distance-the DoF.

Of course you can factor all manner of 'extras' into the situation as J. Sachs has done especially for digital. Ending, ultimately with the size of two pixels as the 'spot defect size'; he calls this 'resolution'.

Of course that 'resolution' is not the same 'resolution' found in tests like Imatest by N. Koran. ...

[Lowell Goudge]

Too bad-it's mind blowing!

It's based on something like this: the size of a defect spot on the film/sensor plane that will show on the normal print at normal viewing distance as a distinct ring (circle in circle of confusion) [as opposed to a dot].

Then there is some math to relate that size to two distances; one in front and one in back of the normal focusing distance-the DoF.

Of course you can factor all manner of 'extras' into the situation as J. Sachs has done especially for digital. Ending, ultimately with the size of two pixels as the 'spot defect size'; he calls this 'resolution'.

Of course that 'resolution' is not the same 'resolution' found in tests like Imatest by N. Koran. ...

It's a good thing that the photo's are much clearer than the explanation.

[Akcelik]

You need a better understanding of Depth of Field. Google 'fCalc'; there is a small applet (free) that gives a whole lot of photo information, much of it related to your photography.

For grins and chuckles I built a large spreadsheet that does a similar set of tasks. Here is a small piece of the output showing the DoF for 28mm lens. Aperture across top, distance to subject down the left side (I work in inches and feet, so the strange distance numbers are cm converted from inches--starting at 8 inches to subject).

The pink area is your total DoF in cm-very small; it's NOT constant!

thank you very much jfdavis!

it seems i might need that katzeye just to prove to myself if the camera has any issues.
the AF assist is "Never" accurate neither is AF lens.

i also notice, when using AF lens, the focus in the VF slightly strays to a blur (but only slightly) when focusing and all this adding to my worries. and my success rate is pretty low with taking sharp focused pics.

if there is any way i can find out without buying a Split Prism focus screen please be kind enough to let me know as i wold be dreading the installation process that anything might go wrong.

[Akcelik]

It's a good thing that the photo's are much clearer than the explanation.

yes. imagine if i had to say a thousand words.

[Jonas B]

Hmm. Tricky. From the pictures it seems as you have a back focusing problem. But, both of the pictures are up close the the DOF is shallow. You may have moved a bit when pressing the trigger, or there is something wrong with your focusing technique or the camera is defect. Who knows?

The DOF has allready been discussed above. Some other things that come to my mind:

* Unless you use a tripod you don't know if you have moved a bit. I am sometimes swaying when holding the camera and concentrate on focusing. So, you need to do the tests with the camera and subject both fixed.

* The focusing indicator is not trustable. Ignore it as soon as you got the first confirmation. You have to carefully focus on the matte screen and trust your eyes watching the transition of your subject going in and out of focus.

* A Katz (or any other) screen won't help you with this. The only type of screen that is better for manual focusing is a screen made for manual focusing (they are darker with a more coarse etching on them) that are better for showing the small differences when the focus plane moves. I don't know if there is any such screen made for Pentax.

* A magnifier will help you. With a 2-2.5 magnifier attached to the viewfinder you can work more conveniant with the camera on a tripod and get a much improved view of the exact part you are focusing at. Check this thread:
https://www.pentaxforums.com/forums/wanted-items/3956-searching-k10d-2x-magnifier.html
for some options. In my opinion a magnifier is a good thing to have, it will come to use from time to time and it can be cheaper than a new screen.

* A split prism screen is good for manual focusing when it comes to daily use, for normal pictures and snapshot situations. When really critical focusing is necessary a magnifier and a matte screen is better. The split prism is seldom perfectly made.

* Sometimes focus bracketing is the last resort.

I hope the above can help somewhat and that you get this sorted - it's annoying not being sure about the focusing point.

[jfdavis58]

I'm in the group of people who believes that most auto-focus issues are 'user related'; i.e. user error.

The camera and lens are manufactured devices. They don't have some specific single point of perfect quality; they have a range of acceptable quality--a tolerance or in statistical terms they are manufactured to some confidence interval or confidence limit.

Do back or front focus issues exist? Of course! But not to the degree encountered by the users of several popular forums. And the testing procedures they use are totally inadequate to determine front or back focus with any precision. When ever you operate a mechanical device near it's limits, you stray into the realm of statistics and most importantly, the area of operation with rapidly diminishing quality returns.

The short of this is that there is only a very small chance that something is wrong with your camera; not enough to worry about or to try to adjust.

The single most difficult aspect of macro photography is light. And light is key in photography. You need to raise the intensity of good quality light (wide spectrum) to a very bright level. This in turn, gives you a better DoF as it allows flexibility with the aperture selection.

Next to quality bright light you need rigid support: a high quality tripod and head. And your subject must be held stationary as well. Considering the DoF numbers in the chart I extracted, millimeter or better accuracy is typically required.

Finally, forget about auto-focus. Relative to your subject the coverage of the auto-focus sensor is just to broad. On that microphone it covers left to right the full distance of it's circular diameter--and the subject to camera distance at the edge is significantly different from the subject to camera distance at the center--perhaps as much as 1 or 1.5 cm. That's the Dof in the wide open aperture situation!

Final recommendations: quality bright light, rigid support, manual focus and lots of practice!

[Lowell Goudge]

* Sometimes focus bracketing is the last resort.

Interesting comment.

I think this was on someone's wish list for software improvements, using perhaps selectible autofocus sensors to bracket between, or front to back of the scene focus bracketing

[jfdavis58]

In a perfect 'auto-focus' world; perhaps.

Considering the very low comprehension exhibited when it comes to auto-focus limitations and parameters, compounding the situation with more sensors and more procedures seems almost recklessly foolish.