[Deiberson]

If the dot means nothing...what should I be using to focus? If I'm at 3.2 and everything is on an equal plane I'd expect at least the head to be in focus especially if I'm placing the dot on the head. I'm referring to my OP.

The question now becomes...."How do I tell the camera to focus on that spot" ? On my neighbors 1DX...of which I have zero experience using, I can nail a microscopic flower-head out of a garden every time. The camera is spot on. With the k5 sometimes I get it....sometimes I can't hit the side of a barn.

[rawr]

On my neighbors 1DX...of which I have zero experience using, I can nail a microscopic flower-head out of a garden every time. The camera is spot on. With the k5 sometimes I get it....sometimes I can't hit the side of a barn.

Are you kidding?

The EOS 1DX is:
(a) a top of the line, brand-new, expensive pro DSLR, and
(b) has 61 AF points, compared to the K-5's 11.

Naturally, you will get more granular AF control with a 1DX.

How do I tell the camera to focus on that spot

You indicate to the camera that you want it to work on a specific area, not a laser-dot.

Test it out. Learn the sensitivity area boundaries of the AF points. Practice.

[arnold]

If the dot means nothing...what should I be using to focus? If I'm at 3.2 and everything is on an equal plane I'd expect at least the head to be in focus especially if I'm placing the dot on the head. I'm referring to my OP.

The question now becomes...."How do I tell the camera to focus on that spot" ? On my neighbors 1DX...of which I have zero experience using, I can nail a microscopic flower-head out of a garden every time. The camera is spot on. With the k5 sometimes I get it....sometimes I can't hit the side of a barn.

That is why I like to keep it simple. Centre focus, half press the release to hold it, then re-compose. This way you tell the camera what distance to use.

[adwb]

And which is why I suggested using Photo Tips and Sips: The Poor Man's (or Woman's) Focus test

so you can learn how to use your focus system.

if nothing else the examples shown will prove if there is a fault with the camera /lens or not

[Deiberson]

My issues are that my results aren't consistent. Like I said.....I have A LOT invested with Pentax. Thousands. I want to make this work before jumping ship. If I had to spend 2000 more for a body that can nail this every time, why not? Over the course of 2-3 years that would be a few bucks a month. Hardly worth fighting poor a AF system over. The 5diii has the same 61 point system with 41 cross points. My K5 has 11. Who are we kidding? I'm not saying I don't get GREAT shots with my camera. I do. But they are not consistent. And if I'm trying to get candid poses with burst shots, the center point/recompose technique isn't going to work.

ADWB...I do appreciate the link, its what I asked for when I started the thread. I will give it a shot.

But, I need to be able to use Af select for composition of portraits. So, what's the best way to tell the camera to work in a specific area so I can continually work that area without recomposing every shot? That in itself is worth an upgrade.

[adwb]

Deiberson,
it occurs to me that using the same crib with perhaps a doll or something other than a child [unless sleeping] that won't move and in the same light as your original post and using the user select and single point as you did, what if you try the shot on and off a tripod?

Now remember we are told to use the self timer or a remote on a tripod or turn off the anti shake, personally I never bother and never have a problem with that, but hey ho who am I to argue?

I want to try to eliminate a mechanical fault with body/lens. since you say it is not consistent in being wrongly focused we need to see when it happens.

If it never, ever, happens on your tripod and very shot is sharp at the point you select then we can then try to see what else is causing this.

We can sort something out with a bit of patience and trial and error.

I shoot lots of weddings and for fun street photography in between the mundane commercial cars and suchlike.

I also get f ups as well but I only use centre and recompose, single shot, I mostly know my problems are me being shaky and mostly not watching shutter speed vs focal length.

When I look at the shot data in LR it nearly always is a shutter speed thing. some people say that it it not so important with the Pentax anti shake any more to watch the speed/focal length but for me it certainly is.

try some test and come on back as they say!! pm if you want or email me look on my website for details in the tag below.

[GibbyTheMole]

You've gotten a bazillion different answers so far, so here's your bazillion and oneth... ;-)

What you have to do first is figure out if it's your technique or the camera. Here's what I'd do:

1) Set the camera to center focus point. This is what I always use & I've never had a problem. Put the thing you want to focus on in the center of the frame, press down halfway to focus, then while holding the shutter release halfway down to retain the focus point, recompose & shoot.

2) Test your camera. Grab the fastest AF lens you have, put your camera on a tripod, and stick a newspaper flat on a wall. Aim the camera straight at the newspaper and shoot with the lens wide open. Do it 3 or 4 times from different distances . If you can clearly and consistently see the fibers in the paper & microdots of ink with no blurring, etc., your camera is ok.

3) If it's not, you can recalibrate the focus according to the camera manual. (I have a K-x, so it's likely a different procedure with mine. I can't tell you the exact method with the K-5.) Write down the setting before you change it so you can go back to it if you get it out of whack.

We have 5 Pentax DSLRs in my house. 3 of them needed a little tweaking to be dead-on-the-money all of the time accurate focusing. I think if you approach it methodically and logically, you'll get it sorted out.

Cheers,
Bobbo :-)

---

Last edited by GibbyTheMole; 09-11-2012 at 05:46 AM.

[ogl]

My issues are that my results aren't consistent.

https://www.pentaxforums.com/forums/pentax-k-30/197362-most-important-questio...ike-k-5-a.html

[atekant]

Well, here is another theory, and it makes sense to me:
Reading one of the threads on this forum about the K-5 II, I read about the new AF sensor, and how it was an f/2.8 sensor, as opposed to the old K-5 at f/5.6
I didn't know that the AF had an "aperture" but it makes sense now I think about it that it does.
More so, it kinda explains the focus issues, and I believe it makes the AF useless if you are using a fast lens.

I did a quick calculation and test based on my K 50/1.4, and found out that shooting at a distance of 3 feet and at f/5.6, the "in focus zone" (going by the focus markers on the lens) is about 4-5 inches. So, the +/- 2 inch error in the focal plane would obviously matter with a fast lens.

It kinda makes me glad that I invested in a focusing screen, and not spent a dime on fast AF lenses.

[cali92rs]

Well, here is another theory, and it makes sense to me:
Reading one of the threads on this forum about the K-5 II, I read about the new AF sensor, and how it was an f/2.8 sensor, as opposed to the old K-5 at f/5.6
I didn't know that the AF had an "aperture" but it makes sense now I think about it that it does.
More so, it kinda explains the focus issues, and I believe it makes the AF useless if you are using a fast lens.

I did a quick calculation and test based on my K 50/1.4, and found out that shooting at a distance of 3 feet and at f/5.6, the "in focus zone" (going by the focus markers on the lens) is about 4-5 inches. So, the +/- 2 inch error in the focal plane would obviously matter with a fast lens.

It kinda makes me glad that I invested in a focusing screen, and not spent a dime on fast AF lenses.

That is a very interesting point, and something to keep in mind.

But just looking at the original picture, I am thinking there may be an actual hardware or adjustment problem here. If the OP selected the upper left AF point, and put it right over the baby's eye, that is a long way from the front rim of the bowl, where the focus actually locked onto. That seems to me outside the scope of normal tolerance of the AF system.

Edited: From the eye to where the camera actually focused looks about 8-12" to the right. I really doubt adjustment is the issue here, and forwhetever reason the camera grabbed the center point (well not really point, but area) to focus.

[rawr]

I believe it makes the AF useless if you are using a fast lens.

AF is at it's best with fast lenses. I think you are mis-interpreting what has been said about f2.8, f5.6 and AF sensors.

[cali92rs]

AF is at it's best with fast lenses. I think you are mis-interpreting what has been said about f2.8, f5.6 and AF sensors.

Weird, I never said that...I think you are quoting someone else

[rawr]

Weird, I never said that...I think you are quoting someone else

Weird indeed. Didn't mean to attribute it to you.

That was a quote from atekant. I must have picked that up from where you quoted him above.

Gremlins.

[adwb]

OK here is a test I have just carried out . I was going to post the pictures but if you are really that interested I suggest you go try it out your self.

I took a plain white sheet of A3 size paper as that in landscape is the same ratio as a K5 and therefore fills the frame. I very carefully measured the centre and drew with a thin black marker pen a vertical line 1/4 of a inch each side of the centre, and stuck the sheet to a door. target area is therefore 1/2 inch high by about 1/8 " wide

K5 on a tripod set to single focus centre point with 50mm f1.7 fitted. and placed so the sheet of paper just filled the viewfinder.

if the black line in on or in side the right hand ) in the centre of the viewfinder the focus lock green hex comes on.

if the black line is inside but NOT on the left hand ( in the centre of the view finder then focus lock is achieved.

if you imagine the ( ) creating a circle the focus lock comes on if the top or bottom edge of the black line touches the imaginary completed circle or any where inside that circle.

nothing new here we knew that any thing inside that ( ) will be found no matter what the dot is on.

but what is very interesting is to try this same exercise with any of the user selectable points.

you will find that in fact the distance either side of the black line where the red dot is placed is not at all equal. if you take the far left one you will see it is much larger and locks a bit left of the line but not at all to the right and also quite some distance above and below the line.

the same applies to the other smaller usable selected points, the distance that they lock on is is not equal either, they are not crosses so much as we have been told.

this is quite hard to explain and either needs images with the red dot added or you you need to test for your self.

what it all therefore indicates is that even if you have put the centre dot on a object say a face, if there is something more contrast in the ( ) area it can lock on that and not the face. So your chosen subject should fill the ( ) area.

What is more important is the distance away from the line that the user selectable chosen point will still lock up can lead to the lock being on completely the wrong thing.

whats worse is that in a small dslr viewfinder with no split screen focus can be very hard to spot.

if you go back to page one of this thread and look at the original image you can possible see why the edge of the crib is in focus and not the face.

it also make no difference what f stop you are using as far as I can see.

as the op said [see below]" I use my arrow keys to move the red dot over what I want to be in focus." if you do that you could easily get the crib edge in focus and not the face.
Try it out you will soon see what I mean.

Perhaps I don't understand how it works then. I assumed that if I'm on SEL on the back of the camera and AF-S, I use my arrow keys to move the red dot over what I want to be in focus. I'm "selecting the point" for the camera so it doesn't have to read my mind.

[MPrince]

Unfortunately, the northlight-images link is not working now. A Google search on "AF adjust moire" finds other sites that explain the technique.

Well, that sucks. It was working at the time of my post, but now 4 days later still isn't working. Wonder what happened. The northlight sight had a downloadable moire pattern that can be used to utilize the technique.