[Isaac314]

1. The camera has started frontfocusing while using manual lenses and shake reduction. It doesnt seem to affect digital lenses, but my 50mm suffers from just a smidge of disagreement between the sensor and viewfinder which seems irreparable. Very annoying, just as i got good at fast focusing i can no longer trust my viewfinder (i cant just compensate, the amount of front focusing is random, sometimes non at all if i point the camera down or something)
To me it seems like the sensor is moving forwards or backwards because the shake reduction has done something bad.

2. No temporary raw file in live view mode. You know how when you take shots and then go into white balance, the last shot is there for a reference? Not in live view mode it seems... more evidence of an ill thought out function

[audiobomber]

1. The camera has started frontfocusing while using manual lenses and shake reduction. It doesnt seem to affect digital lenses, but my 50mm suffers from just a smidge of disagreement between the sensor and viewfinder which seems irreparable. Very annoying, just as i got good at fast focusing i can no longer trust my viewfinder (i cant just compensate, the amount of front focusing is random, sometimes non at all if i point the camera down or something)
To me it seems like the sensor is moving forwards or backwards because the shake reduction has done something bad.
(

I suspect I'm having the same issue. I just took a bunch of photos with a couple of lenses, some shot in MF, some in AF-S, some in AF-C. All of the AF shots are in pefect focus. None of the MF shots hit. This was on a tripod with 2s delay, which means shake reduction was disabled. I have excellent distance vision, I could not have misfocussed all those photos.

[kmccanta]

check the diopter setting for your eyepiece?

[bdery]

Oe thing I notice with my A-50 f1,4 is that the camera takes a few miliseconds to confirm focus, and sommetimes when turning the focus ring I'm past the actual focus point when I get confirmation. So what I do now is prefocus visually, then move the focus ring slowly, and in steps, near that point.

So far it has helped.

[audiobomber]

check the diopter setting for your eyepiece?

I can see the target is in focus in AF, so that can't be it. I'll try some more shots.

[audiobomber]

I just tried some more photos with MF and it looks like (surprise!) operator error. Some hit, some didn't. The camera and I agree, it beeps when I see focus. I suspect I've been nudging the focus ring when I take my hand away.

[k100d]

I just tried some more photos with MF and it looks like (surprise!) operator error. Some hit, some didn't. The camera and I agree, it beeps when I see focus. I suspect I've been nudging the focus ring when I take my hand away.

that's the most important. when MF'ing, the AF indicator is insufficient to guarantee focus. you might notice that it is on for a small range of focus values. make sure that you and the camera agree

[Marc Sabatella]

It's also important to remember that with fast lenses (better the f/2.8), the viewfinder lies to you about what's in focus. It's incapable of showing DOF as thin as what you get at f/2.0, much less f/1.4. So there will *always* be things that appear in focus in the viewfinder that aren't when you take the picture (when using wide apertures). Tht's because of the limitations of the focus screen. So you need to learn to anticipate this and figure out which things that appear in focus won't be. On my camera, it's the rear portion of the one of acceptable focus. So if I;m looing at an angle at a newspaper and 20 lines appear in focus in the viewfinder, only the *front* 10 of those will be in the picture. SO I focus by making sure my subject is toward the *front* side of being out of focus, not toward the back. Stated another way, I make sure nothing in front of my subject is in focus, but something behind it is. Your camera's screen might be adjusted slightly differently, so do the test with a newspaper yourself to see.

[creampuff]

I find the very idea of front or back focusing when using a manual focus lens to be absurd. 100% of the time it has to be user error. PLBV.

[Isaac314]

I am so sure that there is something wrong. If i have a pin highlight, such as a piece of glitter on a cats collar, i can see the exact point at which i am in focus because the light is a pin and not a circle. Then the final image is not in focus. I have tried a lot of times focussing as well as i can (i am very good, lots of practice), and then switching to live view and it being backwards by a nudge.

The manual lens has a very stiff focus ring, i'm not accidentally pushing it...

If anything makes me buy the 55mm SDM this will be it...

[Marc Sabatella]

I find the very idea of front or back focusing when using a manual focus lens to be absurd. 100% of the time it has to be user error. PLBV.

Not true - if the focus screen is mispositioned, it will produce exactly those results. Too close to the mirror = FF, too far = BF. That's why third party screen include shims.

[Marc Sabatella]

I am so sure that there is something wrong. If i have a pin highlight, such as a piece of glitter on a cats collar, i can see the exact point at which i am in focus because the light is a pin and not a circle. Then the final image is not in focus.

OK, when dealing with an absolute point source, it may be possible to nail it, but but I still say that you should do the test I mentioned, shooting text at an angle to quantify which part of the apparent focus range is in focus, and tht after doing this, you'll be much better equipped to focus on "regular" objects (eg, not pin points).

Better yet, try a good focus test chart like the one at Home. Designed for AF, of course, but it works fine for MF too.

[morfic]

I find the very idea of front or back focusing when using a manual focus lens to be absurd. 100% of the time it has to be user error. PLBV.

You assume AF dslr screens are calibrated the way screens were when they were the only way of reaching focus. now a viewfinder mostly just serves composition, else we would get calibrated screens that are siutable for manual focusing.

[creampuff]

Not true - if the focus screen is mispositioned, it will produce exactly those results. Too close to the mirror = FF, too far = BF. That's why third party screen include shims.

C'mon, the OP did not indicate if he had changed his focusing screen from the original one that came with the camera, so where can the error arise, when we're talking about manual focusing? Either the diopter wasn't adjusted properly first or it's simply user error.

If you're using 3rd party screens, that's a different issue altogether. Any focus errors arise mainly because of the varying thickness of the 3rd party screens, necessitating the use of shims to match the thickness of the original Pentax focusing screen. Seen a Katzeye installed K20D whose AF confirmation doesn't even match with what the split image or microprism indicates.

[creampuff]

You assume AF dslr screens are calibrated the way screens were when they were the only way of reaching focus. now a viewfinder mostly just serves composition, else we would get calibrated screens that are siutable for manual focusing.

That may be true but, I've talked with the friendly and experienced service techs at the Pentax authorised agent where I am (who happen to service Hasselblad and previously Nikon too) and I can assure you the stock screens are accurate for manual focusing and AF.

There are any number of reason why we don't focus very well manually:
Diopter not set correctly
Differences in magnification and size in DSLR viewfinders compared to film cameras.
No focus aids like split image rangefinder/microprism that film cameras had.
Focusing with lenses that do not have a fast aperture, especially wide angles where everything appears sharp
As we age, our eyesight deteriorates.
The lenses we use could have been opened up at some point and not collimated for proper focus. I have seen this plenty of times with old lenses that have been serviced by people without proper equipment.