[Mark Castleman]

I'm new here and I keep seeing people talking about "front focusing" and while I can see that it appears to be a Bad Thing, no one ever defines what it means.

So what exactly is "front focusing" and how worried should I be about it?

[volosong]

Anything that I would write would just be a re-hash of what is presented on Tim jackson's web page (link below). Be sure to download the PDF file at the bottom of the page so you can run the tests for yourself, if you desire. (Even though Tim is talking about the Nikon D70, what he writes is applicable to any dSLR camera.)

Nikon D70 Focus Chart

[jfdavis58]

Many of the so called 'problems' found in new cameras are at the extreme edges of the cameras operating parameter's envelope--extreme low light AND high ISO, maximum wide aperture AND very near focus, extreme environmental conditions of heat and cold.

Front or back focus is in the middle class problem above: where the camera auto focuses is not where it's aimed--it focuses in front or behind the target point.

There are problems with using any device in such a near out of specification manner, there are problems with the amateur testing methods used to find said problems, there are problems with the testors own biases.

Save your worry for important things like if you will have saved enough money to buy your next lens when it next becomes available.

Edit: specific to the K10D: BEFORE DOING THE JACKSON TESTING read the manufacturers warnings on page 66 of the owners manual; NOTE ESPECIALLY a, b, d, e,e,e,e,e,f,f,f,f,f!!!!!

[Ivan Glisin]

Take a series of photographs of various subjects at various distances and if you can not see obvious AND consistent problems with auto focus everything is just fine. There may be some visible variances (as AF precision depends on lighting, contrast, subject type, pattern and similar) but as long as they are exceptions they ARE NOT an indication of camera or lens malfunction.

I recommend you to AVOID using test charts!!! The problem is twofold: First, using test chart does not represent typical situation AF system had been designed for -- AF target is too close, DOF is too shallow, contrast is too high and pattern is too simple. Second, remember that AF sensor area is MUCH BIGGER than that little red dot indicating AF lock. This is important because when testing you can not precisely control what AF sensor will include in evaluation. In the focus test chart in question you will see text, numbers, ruler lines, target area frame and similar "distractions" AF system may catch and properly focus as designed (maximizing contrast in objects covered by AF sensor) which may not render area under the red AF light as sharpest area. This is not a malfunction. Be careful with this even with real world subjects.

[Ivan Glisin]

Excellent post!!!

[roscot]

Edit: specific to the K10D: BEFORE DOING THE JACKSON TESTING read the manufacturers warnings on page 66 of the owners manual; NOTE ESPECIALLY a, b, d, e,e,e,e,e,f,f,f,f,f!!!!!

Actually, that same message is on page 46 in my DS manual as well. Another reason that I love the Katz Eye focus screens.

[volosong]

specific to the K10D: BEFORE DOING THE JACKSON TESTING read the manufacturers warnings on page 66 of the owners manual; NOTE ESPECIALLY a, b, d, e,e,e,e,e,f,f,f,f,f!!!!!

How small (or large) is the focus area when using center spot focusing? Are you saying that even though a camera is set on center spot focusing, it's focus area is much larger, to the point of including those distance bars on the sides?

[jfdavis58]

If you're referring to the "()" bars, YES the center focusing spot is two orthogonal lines-one vertical, one horizontal , and they extend a small distance outside the circle defined by the "()". None of the center nine are 'spots' all are crosses. The 10th and 11th lines, outlying left and right, are vertical lines. The little red square or dot is related only to the extent that it indicates the active/activating sensor; it doesn't indicate sensor size.

[Ivan Glisin]

Take a blank sheet of paper and make a small but bold dot in the center. Switch your camera to selectable focus points and try focusing aiming at and near that dot. You will be surprised by the result, especially by the coverage of eight AF sensors surrounding the center AF point. (Some AF points seem to be overlapping as well!)

Then point at a real world subject and focus. Once red dot confirms AF lock look carefully and try to imagine what part of the scene got actually used for contrast evaluation by the AF system. You will quickly find situations that may be confusing even to you and are certainly to effectively blind AF system.

BTW, how and why I know all this? I performed focus test using that chart confirming my DS with FA35/2.0 has AF issues. I opened my DS to adjust AF sensor plate, confirming with additional tests chart shots everything looks good. But in real world situations I was getting more AF errors which forced me to open my DS again and move AF sensor plate to the original position. At that point I figured out that something must be wrong with the test and this is what I have discovered: AF point is not an AF point! It is in fact an AF AREA that may cover much more then what we would expect from a "point".

And this also explains why some people complain about FF/BF issues with some but not all lenses. Two reasons: (1) lenses with larger apertures have shallow depth of field and make focusing errors more visible and (2) shooting a test chart with all lenses from approximately the same distance would force AF area with wide angle lenses to evaluate greater percentage of the image using selected AF area resulting in apparently reduced AF precision.

[Ray Pulley]

Hi all,

My new K10D Front Focuses noticeably with every lens I put on it (FA 50mm F1.4, F 50mm f1.7, DA 40mm pancake, DA 16-45, FA 31mm litd, FA 43mm ltd, FA 77mm ltd, etc...).

I noticed this not in some measurbation test, but while processing about 100 RAW images taken with the K10D on Christmas eve. I had maybe 3 or 4 that I considered sharp. I was shooting indoors with the 16-45, wide open aperture or nearly so (and yes, I know that this lens will not be razor sharp wide open).

Lest anyone think this is pilot error, I own 2 *istds bodies, on K100D, 2 MZ-s bodies, a ZX-L and a PZ-1 and none exhibit this sort of focus error.

It is more pronounced in incandescent lighting, but still noticeable in full daylight.

After discovering so many soft images, I did set up the measurbation chart (tripod, set the lens to OOF and the AF, etc.) and ran the lenses and the K10D as well as the K100D through some tests. While it is clear that the test method is far from 100% scientific, I can reproduce FF on the K10D on every lens I own, and the K100D is as close to spot on as possible in the same setup.

In addition, I can carefully manually focus the K10D on the test chart center bar and get equal DOF front and back and the focus bar is sharp and the focus confirmation light is NOT lit when I manually achieve best focus.

So, the K10D is off to Pentax today, and hopefully they can simply adjust it for a better result.

I do not know the cause of this problem, but I know that it is real because I have tested all sorts of lenses and more than one body.

The good news is that it is very consistent, so if they can get it calibrated closer, I will be able to trust it as it rarely, if ever, locked in different places from shot to shot.

Ray

[jfdavis58]

Ray,

I invite you and others to actually open your mind and ....

Google this set of terms: "front or back focus auto focus".

Read carefully THE TEXT FOUND IN several of the top-most search results. Notice exactly what the 'testing' is designed to show

HINT: IT AIN'T FRONT OR BACK FOCUS!!!

[Dana G]

It's all nonsense.

Every autofocus system is subject to irregularities. If it's so critical to have the exact focus point you want, use the center point, open the lens up wide, and make sure you have good light.

If you want accurate focusing, get the Katz Eye screen with he split prism and microprism collar around it. And focus manually.

[Ray Pulley]

Ray,

I invite you and others to actually open your mind and ....

Google this set of terms: "front or back focus auto focus".

Read carefully THE TEXT FOUND IN several of the top-most search results. Notice exactly what the 'testing' is designed to show

HINT: IT AIN'T FRONT OR BACK FOCUS!!!

Your childish ranting aside:

HINT: almost 100 real world images taken in the same room I have been taking Christmas eve photos for almost 30 years (using many different cameras during that time, all Pentax, film and digital) were all unacceptably soft with the K10D. This prompted more testing on my part.

Images from 2005 in the same exact lighting, using the same lens (the 16-45) on my DS are all properly focused and sharp. Hmmmmmm....

Every lens I own focuses towards the front on this body on the test chart, and strangely enough they do not do so on my other 3 Pentax DSLR's. Perhaps you can capitalize and bold some words that explain that?

Properly set up, the focus chart will fairly accurately tell you the depth of field of a given lens at a given distance and aperture, and it is set up very similar to expensive depth of field wedges designed and sold for the same purpose.

Pay attention now: if you carefully autofocus the camera on the center stripe and the 2mm stripe to the rear is out of focus and the main stripe is barely in focus while the 14mm mark in the front is just out of focus, something is wrong as the camera did not focus on the stripe it was pointed at. It does this consistently, with every lens I tested it with, and at every one of the focus points in the camera. The depth of field would be the same either way, all else being equal. That does not mean that the camera focused where you pointed it.

When you carefully manually focus the same lens for the sharpest image of the center stripe you can achieve, set at the same aperture, on the same body, and the depth of field is almost exactly equal front and rear, then it is quite obvious that the AF system did not adjust the lens to stop at the proper focus point when autofocusing in the earlier test (it was obvious before as the chosen focus point was clearly not real sharp just as my images were not).

If I carefully place the cross-hairs on my scoped .300 Remington Ultra mag at a the center of a target 200 yards away and the projectile hits the dirt 5 yards in front of the target, it likely needs re-zeroing. Seems pretty obvious.

Regardless of all of this measurbation, the camera only produced 3 or 4 reasonably sharp images of people, not charts, while mainly using flash, out of nearly 100. Perhaps you can bold some more words that will explain this as well?

My K100D autofocuses perfectly on the center stripe and delivers equal depth of field front and back with the same setup. Perhaps you can bold some words and explain that as well?

I have been a Pentax user since 1977, and I have never had even one failure of any kind with any of my cameras going back to the screw mount days and including my 645 gear (well, I had to have one 645 fixed once, but I bought it a bit used and abused), so it is not like I am looking to bash Pentax or do not know what a sharp well focused image looks like.

I currently own:

2 *istds bodies, 1 K100D, 1 K10D, 2 MZ-s bodies, 1 ZX-L body, 1 Pz-1 body, and I had the MZ-5N for a while. I have shot tens of thousands of images with these bodies over the years.

Pentax does not have the best AF system on earth, but I have never had almost everything I shot in a given period look as soft and poorly focused as these did and the comparison tests between bodies explained why.

What I fail to understand is why some cannot accept the fact that there might be a problem with a piece of equipment, either by design, or through less than adequate testing and quality assurance.

Maybe you could go to DPReview and do a search in the various forums on AF errors and problems. Pentax is hardly alone in this sort of problem, but at least my body is consistent, so adjustment should be possible if Pentax can reproduce the error on the bench.

I'll be happy to post the results after the camera is returned.

Ray

[Ray Pulley]

It's all nonsense.

Every autofocus system is subject to irregularities. If it's so critical to have the exact focus point you want, use the center point, open the lens up wide, and make sure you have good light.

If you want accurate focusing, get the Katz Eye screen with he split prism and microprism collar around it. And focus manually.

I do not expect perfection, and frankly, the K10D is extremely consistent, so irregularities would not be the correct term as the error was hardly random or irregular.

BTW, I use the center point almost exclusively, and speaking of nonsense, the camera always autofocuses at the widest aperture available on the lens by design, assuming it is not an old non-A lens.....

Ray

[bdavis]

the camera always autofocuses at the widest aperture available on the lens by design, assuming it is not an old non-A lens.....

Can you tell me where you got this information, Ray? I've never heard that &am intrigued.