[mgvh]

I'm using a Pentax A50 1.7 and trying to figure out the best way to get pictures in sharp focus on my K-x.
I'm trying 3 different methods, but I just can't seem to get consistent results. The pic below is an example, but things would likely change if I retried it all again.
To try to standardize things, I'm using a tripod with the shutter delay, so shake should not be an issue. I set ISO to 400 on each so that isn't a factor. My focus is on an object about 5 feet away. Using F4, that should mean that I have about 5.9 inches of depth of field: about 1.7 inches in front and 4.2 behind. The picture is a 100% center group, but it's been resized down. I love using Catch-in-Focus (CIF), and I try to adjust focus very slowly, and there is plenty of focus adjustment available on the A50. In each instance I focused on the "Sharpie."

Left: Using CIF where I focus lens to nearest and then move out until shutter activated.
Center: Using CIF where I focus lens to infinity and them move in until shutter activated.
Right: Using LiveView and magnifying view to 10x to get best view of object.

As you hopefully can see in this set, the first one is the least sharp and, as would be expected considering the method I used, is front focused.
The second is actually the sharpest.
The one using LiveView is in the middle in terms of sharpness, and I seem to have the least success using this method.
When I've done other tests, sometimes the near>far CIF does the best job.

So, my questions:

1. I would have thought that I have plenty DOF (there's only an inch between the Sharpie and the York County behind it), so I'm wondering why there is such a difference in sharpness.
2. I would have thought that using LiveView and magnifying the view up to its maximum 10x would provide the best results. My eyes aren't that bad...
3. Hmmm.... I just thought of something else. I had to microadjust the focus on the camera to get the best pics with my 18-135. (I had to crank it up to +150, but this really is the best setting for the 18-135 which is most often on the camera.) Would that microadjustment affect the CIF function when used with a manual lens?
4. Also, why would the image be so much more overexposed using LiveView? I had to use -.7 EV to make it comparable to the pics using CIF. The lighting did not change.

I know that some people have installed a different (split-prism) viewfinder, but I'm not too keen about spending money on that.


Is there something else I should be trying to improve my skills at focusing?
Thanks!

---

Last edited by mgvh; 10-01-2012 at 11:07 AM.

[JinDesu]

If I may suggest, use a slightly less curvy object for this test. I use a flat paper and a ruler on the side (45 degree angle, aligned the center with the paper) to check for front or backfocus. The reason why I would suggest using a flat paper is that, in CIF, you are still depending on your center cross-type AF sensor to register focus. So if you are coming from near focus or infinity focus, the center sensor is going to see two different things (and it might assume some part of your pen is already in focus because it is so curved).

Live view is weird - but that may just be user error for this test. Normally I can easily see where the focus is in live view (even without focus peaking), but with a curved object like the highlighter, your eyes may get fooled.

For my DA 35 F2.4, I have a setting of -130um to get correct focus, so it can happen. I am of the opinion that anything over 100 or -100um is too much, but I'm too lazy to do anything about that single lens.

I can't imagine why it's overexposed in live-view. Are you using manual mode? Aperture mode? I don't remember if my k-x has different exposure metering for live view vs viewfinder.

[johnyates]

Hmmm.... I just thought of something else. I had to microadjust the focus on the camera to get the best pics with my 18-135. (I had to crank it up to +150, but this really is the best setting for the 18-135 which is most often on the camera.) Would that microadjustment affect the CIF function when used with a manual lens

Yes it would.

Check out the diagram on this page: Focusing Screen--How to adjust focusing screen--

In order to focus properly, the distances L1, L2 and L3 must be equal. To tune L1, you may need to adjust the shims. Lots of info on these forums about that.

[elliott]

I know that some people have installed a different (split-prism) viewfinder, but I'm not too keen about spending money on that.

You can get a cheap diagonal split from jinfinance on eBay for under $30 shipped, if you use a lot of manual focus glass it is definitely worth a shot. If you hate it and didn't damage it, you could probably resell it here on the forum and get about half of your money back. I have the dual diagonal split and it make a huge difference over the stock screen, I don't have a problem even at f1.4.

Catch in focus is a very nice feature to have, but don't expect it to be perfect. It is most useful when you have a moving object and deep DOF so you have room for error.

[Pontax]

I have converted three Contax lenses with the leitax mount, so MF is a big thing for me. The CIF feature is great and I use it all the time, but found that that it didn't focus right, so I made some AF-microadjustment.

Here is how I did. I focused with LV and magnified the view, it is somehow difficult to see focus in LV, so examining the picture works better for me. (question 2)
Enable CIF and then AF microadjust.
At -7 no shutter release, at -8 the shutter went of after the second try, at -9 and -10 the shutter went of on the first, so -9 was chosen.
All tests were made using tripod and remote. So now the CIF is "calibrated" using microadjustment.

If I want to recompose I can press AF on the back (K5) to disable CIF.

Ok to "use all" in microadjustment, works for your A50/1,7 and the rest of your MF lenses. "Use one" works for your 18-135 lens and only this lens.

Question 1 and 4 I don't know.

I think the focussingscreen everybody is talking about is the ee-S, see thread https://www.pentaxforums.com/forums/pentax-k-5/142241-best-screen-ever-manual...anon-ee-s.html

[mgvh]

Thank you, all!
I see that the focusing screens are not terribly expensive, but I really don't like the idea of messing around inside the camera.
@Pontax: I would be happy to do as you suggested, but unfortunately the K-x only allows for a single microadjustment. (Being able to set microadjustments for more than one lens is one of the things that makes the K-30 certainly more attractive.)
@JinDesu: Good point on focusing on a flat rather than curved surface using LiveView. I still think I would have enough DOF for that not to make a difference. I was AV on all the shots. I tried a number of other settings, and LV consistently overexposes just a bit as compared to using CIF. I can't find anything specific in the K-x manual about metering in LV, but it does say on page 143: "The image in Live View may differ from the captured image if the brightness of the subject is low or high." Another vote for the focus peaking in the K-30.

[JinDesu]

It's true, you should have a large enough DOF, however it's very odd for the Live view shot to be out of focus, if it's accurately done in Live view. In fact, it should not be possible. So I can only imagine there was something in error there...

[Paleo Pete]

I've shot almost nothing but manual since I got my K-x, and I tried different methods. Using film I usually did back and forth, but after getting some practice with the K-x I found I got my best results by focusing in front of the subject and moving the focal point back until I got the focus I was looking for. The reason is I shoot mostly birds and wildlife, and I don't usually have a very large subject in the viewfinder, so it's hard to tell if I'm focusing on the eye or the body or the tail...so I bring it in from in front of the subject and pay attention to the focus confirmation beep to help out. I've made no adjustments on the camera and found that the focus confirmation was very accurate with my particular camera, with just about every lens I have.

Aside from that, the best actual advice I can offer is the most common for me...

Always remember the 3 P's...

Practice, Practice and...Practice.

[mgvh]

@PaleoPete: I think the key is practice...
I'm finding that LiveView works best when the screen is easy to see (i.e., does not work best outdoors) and when there is something clear enough and big enough on which to focus with the screen magnification zoomed.
CIF works best when I don't have a lot of time to focus or if the subject is moving.
Otherwise, looking through the viewfinder and using the focus indicator is my best bet. It does make a difference whether I focus down or up using CIF, and what I found is that best focus is really somewhere in the middle. That means just focusing as best as I can ... and keep practicing it.
I shot some pics where I had 8 feet of DOF, and I still see a difference in center sharpness of shots. I'm not understanding why everything in that DOF isn't the same sharpness.
And I still haven't figured out why the LiveView overexposes. Cutting back on the EV does help apparent sharpness a bit since fewer details are getting blown out.

[alamo5000]

I haven't read any of the other responses, but I experimented with this a lot myself. The DOF calculations you are using are for 'acceptable sharpness' not 'absolute sharpness' and please note in my experiments it worked somewhat like this...

out of focus... fuzzy...less fuzzy....perfectly sharp...less fuzzy...fuzzy....out of focus...

The real DOF for you to get a real 'sharp' picture is much less than what your 'rough guide' calculator told you...I figured it out and its roughly 6 inches total of 'acceptability'...but that is not to be translated as absolute sharpness...the closer you get to the subject the harder and more fine tuned you will need to be.

Simply put the real 'sweet spot might be 1-2 inches at best and probably less... so either A. focus better....or B. Stop down a few stops and give yourself some more space to work with. Either that or try setting your camera and moving your subject forward or backward in 1/4 inch increments...

If you really want to see what that lens can do I would suggest setting your camera up to where you can focus down the plane of something....such as a table top. Unfold a newspaper and lay it on the table top as flat as possible and then snap a few shots so you can see where the 'real sharp' starts and the kinda ok acceptable starts and ends... you will be able to actually see it a lot better if you do the newspaper thing...

Then again, I always issue my normal disclosures... I am new to this stuff so take it for what its worth.

[alamo5000]

https://www.pentaxforums.com/forums/photographic-technique/191692-depth-field...discovery.html

[Just1MoreDave]

I haven't read any of the other responses, but I experimented with this a lot myself. The DOF calculations you are using are for 'acceptable sharpness' not 'absolute sharpness' and please note in my experiments it worked somewhat like this...

out of focus... fuzzy...less fuzzy....perfectly sharp...less fuzzy...fuzzy....out of focus...

That's what I would have said too.

When I try to compare lenses, I have learned to repeat the focusing at least three times. No matter how I focus or how careful I am, I often find one set that focused on a slightly different point, enough to throw out those photos. For actual photos, it's rarely necessary or possible to focus that well. "Acceptable sharpness" should work for real-world images, fortunately. A 100% crop is a tougher test, and the scene probably introduced some complications for all the techniques.

I have used magnified live view with great results but the K-7 has a better LCD than the K-x. I find that a split-prism is not absolutely a requirement for me, but I am faster focusing with one and being confident in my results. The Katz-Eye price is tough when you think about it in terms of the cost of the camera. I reasoned that it was something that I would look through for every shot, and I'd only have to install it once. Plus my wife bought it for me. I might not buy one for an older body. My Katz-Eye for the *ist DS fits other cameras but is not usable in the K-7.

It's too late to check tonight but does the K-x have CDAF and does that even work with manual lenses? I'll look into that tomorrow.

[Paleo Pete]

https://www.pentaxforums.com/forums/photographic-technique/191978-depth-field-dummies.html

Depth of field pictures that might give you an idea what to expect at different apertures. Generally, and I didn't catch this when poking around in my various books, you have more area nominally in focus behind the subject than in front.

I've also noticed what Dave mentioned. I shoot mostly birds and nature shots, and I usually try to get as many pictures as I can of any particular subject. Even if a bird is sitting still on a limb, I'll see focusing at different spots in any 3 or 4 pictures taken from the same location within 30 seconds of each other. Depth of field is usually not a serious issue, f8 at 50 feet gives me loads of depth so I don't have to worry about it too much as far as backgrounds go. But for closer shots it can be noticeable.

What happens though, is not everything in the depth of field is exactly sharp focus, it gradually gets more and more out of focus both in front of and behind your subject. (to explain it a different way than above..same thing, different method of explanation) So if you grab a shot at 4 feet at f8 you can see in my link that objects a foot away will already be starting to blur while at f16 the out of focus is barely noticeable, while the outside balls are well out of focus. This gives you a visual representation of both explanations. The balls are a foot apart. You'll also notice in the f8 shot that the ball behind the focal point is slightly more in focus than the one in front of it. That's an example of having more depth to the rear than in front. So in reality, the only area really in perfect focus is the exact focal point, but depending on aperture, which determines depth of field, objects in front of and behind that focal point may or may not be nominally in focus. Notice I said nominally, not perfectly...

[Just1MoreDave]

It's too late to check tonight but does the K-x have CDAF and does that even work with manual lenses? I'll look into that tomorrow.

Yes it does (p. 144-148 of the manual). CDAF is contrast-detect autofocus, using sensor data directly. I don't think it works with manual lenses like catch-in-focus does with PDAF (the ordinary method with separate sensors). That is, it doesn't say directly in the manual and I can't get it to work on my K-7, but there's always a possibility I'm not doing it right.

[Lowell Goudge]

In reading the latter posts here there are a few things not really well explained, although the concepts are touched upon.

Dept of field is actually based upon what is called acceptable sharpness, and it all comes om a point of light still appearing as a point of light on an 8x10 print. DOF is, in fact the product of total enlargement of the lens, through to the print, the bigger you print the less DOF you have

It is also true that there is usually more acceptably sharp behind than in front of the focus plane. It is 2/3 behind 1/3 in front at thenhyperfocal point and beyond, apwith this ratio moving to 50/50 when you get really close

The biggest issue we have today, and I am equally guilty of this, is blowing things up huge. I review my photos on a 22 inch 16x9 format monitor. (it measures 13x19 inches) and significantly bigg than the 8x10 print. I also crop in a lot too. so when I look at shots the normal rules of DOF are useless.

Let's not forget that in discussions about blurry images.