[Wired]

You should probably be focus-and-recomposing, rather than just selecting one edge focus point, the since the eyes are going to be close to outside of the AF frame. Pic crudely related:

I re cropped it for a 8x10.


The problem with focus and recompose is that when you realign your camera you may move it that one or two inches you needed to get the shot. I use Back button auto-focus set to AF-C

---------- Post added 03-06-2015 at 11:44 AM ----------

Live View AF would probably work better than the optical viewfinder for photo elements at the very periphery of the frame. You could also then, of course, zoom in right on the eyes to confirm focus before you take the shot.

Great idea! I'll try this. thank you.

[JinDesu]

Doesn't sound like the AF module is misaligned, so unless the module is curved for some reason, it sounds like either:

a) the outer sensors are all improperly programmed/misreading

b) all lenses distort sufficiently to mislead the phase detect sensors on the outer portions of the AF module.

If it's b), it should be testable on other cameras of other brands to see if they do the same. It should also affect wider lenses more than telephoto lenses. If it's a) it should affect wide and telephoto lenses the same.

If we want to test this, I'd suggest using a 31mm, the 55, the 77, and the 50-135mm at the long end and see if the back focusing values change accordingly (if it's b, it should decrease in backfocus).

All the above is my theory only...

[Wired]

Doesn't sound like the AF module is misaligned, so unless the module is curved for some reason, it sounds like either:

a) the outer sensors are all improperly programmed/misreading

b) all lenses distort sufficiently to mislead the phase detect sensors on the outer portions of the AF module.

If it's b), it should be testable on other cameras of other brands to see if they do the same. It should also affect wider lenses more than telephoto lenses. If it's a) it should affect wide and telephoto lenses the same.

If we want to test this, I'd suggest using a 31mm, the 55, the 77, and the 50-135mm at the long end and see if the back focusing values change accordingly (if it's b, it should decrease in backfocus).

All the above is my theory only...

interesting... I'll play with this over the weekend.

I don't see this issue when shooting with my D800, which is why I'm frustrated.


another theory brought up by accident in another thread is that maybe I accidentally turned on moire reduction. who knows what thats doing when coupled with the strobes...

[JinDesu]

Does the D800 have focus points that far out as well? I'm not familiar with it, but if it doesn't occur on the D800 at all, then it might just be your k-3 (at which point I have no idea why it is doing that).

Moire reduction should affect the whole image though (as far as softening). But if you did it with strobes and the strobes are firing faster than 1/1000s.. maybe it's only blurring part of the frame..

[Wired]

Does the D800 have focus points that far out as well? I'm not familiar with it, but if it doesn't occur on the D800 at all, then it might just be your k-3 (at which point I have no idea why it is doing that).

Moire reduction should affect the whole image though (as far as softening). But if you did it with strobes and the strobes are firing faster than 1/1000s.. maybe it's only blurring part of the frame..

thats the theory I'm thinking. as I only notice it on vertical images...

[Wired]

Does the D800 have focus points that far out as well? I'm not familiar with it, but if it doesn't occur on the D800 at all, then it might just be your k-3 (at which point I have no idea why it is doing that).

Moire reduction should affect the whole image though (as far as softening). But if you did it with strobes and the strobes are firing faster than 1/1000s.. maybe it's only blurring part of the frame..

So, I did have Moire reduction on high. But I haven't had a chance to test it yet using strobe. Using natural light it was bang on using the outer points. My D800 does have focus points pretty close to that far out in the view finder, and they are tack sharp when locked on.

[Prakticant]

Lucky one as I'm virtually fighting with my K-3 to get sharp what I'd like to be sharp (the eyes, or the front eye, DA* 16-50mm & DA* 200mm).
Do you use K-3 AF Fine Adjustment micro settings for your K-3 body (Apply All) or/and for your lenses (Apply One) or combined?
I'm playing with the above micro adjustments, which makes things better, but not dead accurate (some back-focussing, some randomness).
So I miss focus-bracketing with K-3
Answering your question:
AFAIK side focusing points may be calibrated in K-3 service-menu, so you may want your camera to be checked in a service center.

-----

anytime I use the center focal points the images are tack sharp.

[Wired]

Lucky one as I'm virtually fighting with my K-3 to get sharp what I'd like to be sharp (the eyes, or the front eye, DA* 16-50mm & DA* 200mm).
Do you use K-3 AF Fine Adjustment micro settings for your K-3 body (Apply All) or/and for your lenses (Apply One) or combined?
I'm playing with the above micro adjustments, which makes things better, but not dead accurate (some back-focussing, some randomness).
So I miss focus-bracketing with K-3
Answering your question:
AFAIK side focusing points may be calibrated in K-3 service-menu, so you may want your camera to be checked in a service center.

-----

I've calibrated every single lens separate from each other. They do loose calibration over time, and if you mess up it causes more issues. So it's a matter of getting it right.

[rodeknyt]

The problem with focus and recompose is that when you realign your camera you may move it that one or two inches you needed to get the shot. I use Back button auto-focus set to AF-C
---------- Post added 03-06-2015 at 11:44 AM ----------

Curious why you would use continuous AF for shooting this type of work. You can't really "lock" focus when using AF-C.

[Wired]

Curious why you would use continuous AF for shooting this type of work. You can't really "lock" focus when using AF-C.

quite the opposite actually.

While running continuous AF using back-button auto focus setup I'm constantly locked. So if the distance from the model to the camera changes for whatever reason, no matter how minute, the camera will correct for it. Works brilliantly. Unless your using the side focal point and have morie reduction on for some bizarre reason when using studio strobes.

https://www.pentaxforums.com/forums/38-photographic-technique/236970-perfect-...everytime.html

Back Button Focus: What it is, why you want it & setup guide

[Adam]

Plus one for using live view for this particular application.

[Wired]

I should update my weekend findings...

I did a calibration test using FA 43, FA 77, DA*55. The center group of points are dead on sharp. The two outermost points (that show up as lines are actually out! The right side is back focused, the left side is front focused. Now while that is consistent, the findings across the lenses (in terms of ratio) is not. Shooting at f2.8 (for consistency) the difference when adjusting the AF fine tune is really wack. The DA*55 was actually pretty close across the board. the FA77 didn't have as much front focus on the left, but the right was severely back focused. The FA43 was like a teeter-totter with how far out it is from side to side. Just silly (-8 /+9 or something like that)

Moire reduction was turned on, and test firing did show major softness issues that look like backfocusing across all focus points using strobes. When using natural light there is still a little bit of softness if you 1:1 on the eyes, but only if your being a whiny lil brat about it. I'd say it's sharp enough.

Live view, moire reduction off, solved everything.

[stevebrot]

The two outermost points (that show up as lines are actually out! The right side is back focused, the left side is front focused.

Time for a service call?


Steve

[emalvick]

I should update my weekend findings...

I did a calibration test using FA 43, FA 77, DA*55. The center group of points are dead on sharp. The two outermost points (that show up as lines are actually out! The right side is back focused, the left side is front focused. Now while that is consistent, the findings across the lenses (in terms of ratio) is not. Shooting at f2.8 (for consistency) the difference when adjusting the AF fine tune is really wack. The DA*55 was actually pretty close across the board. the FA77 didn't have as much front focus on the left, but the right was severely back focused. The FA43 was like a teeter-totter with how far out it is from side to side. Just silly (-8 /+9 or something like that)

Moire reduction was turned on, and test firing did show major softness issues that look like backfocusing across all focus points using strobes. When using natural light there is still a little bit of softness if you 1:1 on the eyes, but only if your being a whiny lil brat about it. I'd say it's sharp enough.

Live view, moire reduction off, solved everything.

It sounds like your lenses might be decentered if they are showing the type of variations you are seeing (except the *55).

I do notice with all my lenses that the extreme focus points don't work as well as the center points, but they work better for my shots when I really need them rather than focus and recompose as I used to do. However, I do take it as a given that few lenses can be perfect across the frame. I might expect a true macro lens to perform well, but I expect limitations with most other lenses. BUT, I would expect the lens to be off by a similar amount on each side of the frame. I would not be happy if the right side was FF and the left side was BF or something similar as you reported.

Photography isn't so much about having perfect equipment but rather knowing your equipment and being able to work within its limits. But the limitations of lenses should be fairly consistent for that to really work in practice.

[rodeknyt]

I should update my weekend findings...

I did a calibration test using FA 43, FA 77, DA*55. The center group of points are dead on sharp. The two outermost points (that show up as lines are actually out! The right side is back focused, the left side is front focused. Now while that is consistent, the findings across the lenses (in terms of ratio) is not. Shooting at f2.8 (for consistency) the difference when adjusting the AF fine tune is really wack. The DA*55 was actually pretty close across the board. the FA77 didn't have as much front focus on the left, but the right was severely back focused. The FA43 was like a teeter-totter with how far out it is from side to side. Just silly (-8 /+9 or something like that)

That looks like the AF module is not properly installed in the camera body. That probably can only be fixed with a trip to the shop.