[tvetter]

I just got a K20D body to use with my collection of old manual-focus primes and some of the lenses are showing severe backfocus of the type which has been discussed in other threads (viewfinder and AF assist agree on target being in focus; actual images show focus point well behind target). Worst offenders are 50mm f1.4 M, 28mm f3.5 K. Others, like a 50mm f4 macro K, and a 135mm f3.5 M, are fairly close to focusing correctly. Has anyone dealt with this problem successfully, say by replacing or repairing the body, or making mechanical adjustments to the lenses, or shimming the viewfinder, or ...? The camera is relatively useless to me if I can't reliably focus half my stock of legacy primes. Help!

[Lowell Goudge]

you might want to post a photo or two.

Also describe your approach to focusing, etc.

[Wheatfield]

If what you are describing is actually the case, then you need to get the focusing screen shimmed in, and then the AF adjusted to match the screen.

[Lowell Goudge]

In re-reading your post, and also looking at the response from wheatfield, I think wheatfield is describing something different, as shimming the focusing screen is something to make the screen match the AF indicator, and you are stating these both agree.

I also got to thinking about another post

https://www.pentaxforums.com/forums/pentax-dslr-discussion/41713-focus-shifti...e-setting.html

It might be that your 50mm is a little soft at full apature and hence difficult to get the most accurate AF indicator, or MF actually right. I have not thought about this much and might be able to see how my K10D and *istD perform with my K50mm F1.4

I don't own a 28, but have had no problem with my *istD with my Tamron 24mm F 2.5 or my M42 35mm F2.

I have not tried them on my K10D but since depth of field is quite high on these lenses, what is considered in focus may not be exctly in focus.

Also, Have you considered trying to focus from infinity and also from min focus, to see if the behavior changes.

You note better performance with your longer slower lenses, which should be sharper , especially the 50mm macro at F4 vs a 50mm F1.4 wide open,, and also the 135mm which will have much tighter DOF and therefore easier to focus precicely.

Just some thoughts, before you spend forever playing with focus adjustments on the body.

p.s. I am also considering at some point to go for a split image finder, probably for my K10D when I upgrade next. I think this is probably the best solution, because just using the screen is not as accurate, and I fing the focus confirmation light too far off the field of view when concentrating on shooting.

[Wheatfield]

In re-reading your post, and also looking at the response from wheatfield, I think wheatfield is describing something different, as shimming the focusing screen is something to make the screen match the AF indicator, and you are stating these both agree.

I also got to thinking about another post

https://www.pentaxforums.com/forums/pentax-dslr-discussion/41713-focus-shifti...e-setting.html

It might be that your 50mm is a little soft at full apature and hence difficult to get the most accurate AF indicator, or MF actually right. I have not thought about this much and might be able to see how my K10D and *istD perform with my K50mm F1.4

I don't own a 28, but have had no problem with my *istD with my Tamron 24mm F 2.5 or my M42 35mm F2.

I have not tried them on my K10D but since depth of field is quite high on these lenses, what is considered in focus may not be exctly in focus.

Also, Have you considered trying to focus from infinity and also from min focus, to see if the behavior changes.

You note better performance with your longer slower lenses, which should be sharper , especially the 50mm macro at F4 vs a 50mm F1.4 wide open,, and also the 135mm which will have much tighter DOF and therefore easier to focus precicely.

Just some thoughts, before you spend forever playing with focus adjustments on the body.

p.s. I am also considering at some point to go for a split image finder, probably for my K10D when I upgrade next. I think this is probably the best solution, because just using the screen is not as accurate, and I fing the focus confirmation light too far off the field of view when concentrating on shooting.

Lowell, my understanding is that when his viewfinder is indicating focus, the image isn't in focus. This indicates either a focusing screen registration problem or else the sensor isn't sitting exactly on the plane it is supposed to be on.

Truthfully, having given it some thought, if he is one of the fortunate people whose AF and viewfinder actually agree, then it is entirely possible that his sensor is a bit misplaced.
I don't know if sensor positioning is possible, but if it isn't, then both viewfinder and AF need to be brought into alignment with the sensor.
This is presuming that it isn't a simple problem like inaccurate focus with wide aperture lenses or some such.

[Lowell Goudge]

Lowell, my understanding is that when his viewfinder is indicating focus, the image isn't in focus. This indicates either a focusing screen registration problem or else the sensor isn't sitting exactly on the plane it is supposed to be on.

Truthfully, having given it some thought, if he is one of the fortunate people whose AF and viewfinder actually agree, then it is entirely possible that his sensor is a bit misplaced.
I don't know if sensor positioning is possible, but if it isn't, then both viewfinder and AF need to be brought into alignment with the sensor.
This is presuming that it isn't a simple problem like inaccurate focus with wide aperture lenses or some such.

I don't disagree, and we need the OP to revisit, but he stated "(viewfinder and AF assist agree on target being in focus; actual images show focus point well behind target)"

Edit note, for the OP's benefit. Just checked my K50mm F1.4 with my K10D and focus indication was consistent with viewfinder focus, and sharpest focus in the image was where the lens was pointed. in this respect, I don't see why your body would be any different, unless as wheatfield suggests the sensor is not where it should be.

[WheresWaldo]

I don't disagree, and we need the OP to revisit, but he stated "(viewfinder and AF assist agree on target being in focus; actual images show focus point well behind target)"

Edit note, for the OP's benefit. Just checked my K50mm F1.4 with my K10D and focus indication was consistent with viewfinder focus, and sharpest focus in the image was where the lens was pointed. in this respect, I don't see why your body would be any different, unless as wheatfield suggests the sensor is not where it should be.

Maybe I am not understanding this but why this phenomenon only on Manual focus lenses, wouldn't the results be the same on Autofocus lenses. If it truly is a body issue then all lenses without using the in body focus adjustments should back focus. Is anyone suspecting a bad lens?

[Lowell Goudge]

Maybe I am not understanding this but why this phenomenon only on Manual focus lenses, wouldn't the results be the same on Autofocus lenses. If it truly is a body issue then all lenses without using the in body focus adjustments should back focus. Is anyone suspecting a bad lens?

I agree, which is why I said we need a revisit from the OP.

for me, I am at a loss on the whole issue of FF and BF as (for what ever reason, I don't seem to have any issues with either my *istD or K10D.) so maybe I'm not a good reference on these issues

I do have issues with human error from time to time but we won't discuss that here, every one makes mistakes, and I do, from time to time snap photos before AF is finished, or while I am turning my focusing collar, but thes are my fault, not the equipments/

[stevebrot]

Before anyone digs out a shim kit, may I suggest that the OP check the diopter adjust on the viewfinder. If the screen itself is not in focus, then attaining focus will be very difficult. I was having focus problems for a while until I realized that the diopter had shifted out of adjustment.

As for the focus-assist, that is a sticky issue. It is hard to say just what the focus confirmation is confirming. The best way to get to the bottom of all this is to actually do test images with a scale (ruler) and a set focus subject.

Steve

[Marc Sabatella]

If we accept at face value that the viewfinder and AF confirmation agree but the picture does not, that suggests the path from lens to focus screen and from lens to AF sensor is the same length, but ut doesn't match the path from lens to camera sensor. I see three possible explanations: either the reflex mirror is misaligned, or the focus screen and AF sensors are misaligned by the same amount, or else the actual camera sensor is misaligned.

But before jumping to that conclusion, I'd want to see some sample pictures. As i've pointed out many times before, it is a little-known fact that focus screen *lies* at large apertures, showing you more DOF than your picture will actually have. This is not a defect but a perfectly normal side effect of the way the focus screen is designed. It is easy to demonstrate this by shooting a newspaper wide open at an angle and observing how much text appears to be in focus in the viewfinder versus in the actual picture. At wide apertures, you will always find more text readable in the viewfinder than on the picture. So it is completely normal to have things be in focus in the viewfinder but not in the picture. As long as in focus area in the picture is also in focus in the viewfinder, there is no problem. If there are in focus in the picture actually appeared out of focus in the viewfinder would there be a problem.

So that alone can explain the issue from a viewfinder perspective. As for the claim that the AF sensor agrees, I would also observe the AF sensor is not a "point" at all but a fairly broad area - *much* larger than the red dot that appears in the viewfinder. So it might be appearing to confirm focus in one place but actually detecting focus elsewhere.

A picture of a well-designed focus test chart like the one at Home would be most helpful. And when focusing manually, try to focus not only using the center line, but also taking into account the rules at the side. Focus so that there appears to be about the same amount in focus in from of the 0 and behind it, or perhaps a bit more behind than in front. And make mental note of exact what the in-focus zone appear to be - "-10 behind, +15 in front", of whatever. Then see how the picture actually taken matches up with this. If the in focus area is within that same range, then all is normal, and you just have to learn how to take this into account when focusing.

[stevebrot]

If we accept at face value that the viewfinder and AF confirmation agree but the picture does not, that suggests the path from lens to focus screen and from lens to AF sensor is the same length, but ut doesn't match the path from lens to camera sensor. I see three possible explanations: either the reflex mirror is misaligned, or the focus screen and AF sensors are misaligned by the same amount, or else the actual camera sensor is misaligned.

But before jumping to that conclusion, I'd want to see some sample pictures. As i've pointed out many times before, it is a little-known fact that focus screen *lies* at large apertures, showing you more DOF than your picture will actually have. This is not a defect but a perfectly normal side effect of the way the focus screen is designed. It is easy to demonstrate this by shooting a newspaper wide open at an angle and observing how much text appears to be in focus in the viewfinder versus in the actual picture. At wide apertures, you will always find more text readable in the viewfinder than on the picture. So it is completely normal to have things be in focus in the viewfinder but not in the picture. As long as in focus area in the picture is also in focus in the viewfinder, there is no problem. If there are in focus in the picture actually appeared out of focus in the viewfinder would there be a problem.

So that alone can explain the issue from a viewfinder perspective. As for the claim that the AF sensor agrees, I would also observe the AF sensor is not a "point" at all but a fairly broad area - *much* larger than the red dot that appears in the viewfinder. So it might be appearing to confirm focus in one place but actually detecting focus elsewhere.

A picture of a well-designed focus test chart like the one at Home would be most helpful. And when focusing manually, try to focus not only using the center line, but also taking into account the rules at the side. Focus so that there appears to be about the same amount in focus in from of the 0 and behind it, or perhaps a bit more behind than in front. And make mental note of exact what the in-focus zone appear to be - "-10 behind, +15 in front", of whatever. Then see how the picture actually taken matches up with this. If the in focus area is within that same range, then all is normal, and you just have to learn how to take this into account when focusing.

Thank you for the excellent and complete discussion. I especially appreciate the comments about the image in the viewfinder having somewhat greater DOF than the image we view on our monitor. The smaller the apparent viewfinder image, the harder it is to attain fine focus.

Steve

[Marc Sabatella]

Thank you for the excellent and complete discussion. I especially appreciate the comments about the image in the viewfinder having somewhat greater DOF than the image we view on our monitor. The smaller the apparent viewfinder image, the harder it is to attain fine focus.

Thanks! I should point out, though, that it's not just the fact that the image in the viewfinder is small that makes the DOF large (although as we know, that contributes). The phenomenon I am talking about is specific to the micro-lens design of modern DSLR focusing screens (as opposed to old-style ground glass, for example). I'm not quite versed enough on the physics to explain what is happening, but see this link for more more thorough (while still not be heavily technical) discussion:

Manual Focus with AF DSLRs

[stevebrot]

Thanks! I should point out, though, that it's not just the fact that the image in the viewfinder is small that makes the DOF large (although as we know, that contributes). The phenomenon I am talking about is specific to the micro-lens design of modern DSLR focusing screens (as opposed to old-style ground glass, for example). I'm not quite versed enough on the physics to explain what is happening, but see this link for more more thorough (while still not be heavily technical) discussion:

Manual Focus with AF DSLRs

You are very welcome. I have been shooting a little lately with my Ricoh XR7 film SLR and have really come to re-appreciate the large/bright viewfinder with split-image/microprism focus aids. I am very, very close to springing for a Katz-eye!

Steve

BTW...The link you included is a real eye-opener! It makes you wonder about a whole lot of other issues such as the manual lens metering foolishness...