Sanity check and opinions please - Soft focus issue on K1000 ...

[dmr]

I'm gonna post this both here and on the "other" system, as I know there's an incredible pool of expertise both here and there.

Here's the issue, to make a long story long ...

My "main" camera for decades has been a Pentax K1000. This is a first generation 1980s vintage Japanese model, all metal, well broken in and until a couple of years ago performing flawlessly. It's the one I feel very comfortable with and I can shoot it instinctively. It's never been physically damaged or even dropped onto a hard surface that I know of.

A few years ago some of the slides I shot on a trip to some of the National Parks came back disappointingly soft. This was shooting with an Ozunon 28-135 zoom of 1990s vintage which until then had been tack sharp. I did a quick test roll at the time which was really inconclusive.

To be sure, I picked up a Takumar 28-80 zoom at a local camera show, shot with that for a while with what I considered good results, and blamed the issue on the Ozunon zoom.

Fast forward to a couple of months ago. A roll I shot with the Pentax zoom came back with some disappointingly soft images, mainly those with images off in the distance where I would have focused on infinity.

I decided to do some more rigorous testing and I would like to hear what the gang here thinks.

First of all, two "control" images are at the very bottom of the post. These were taken of the same subjects but not at the same time. I'm including them to show what I consider "normal" and what I'm expecting. The control images were shot with the Mamiya SD rangefinder but well before the test roll on the Pentax.

All of these images were (re)scanned with the KM SD IV at full 3200 resolution, absolutely no sharpening or photoshopping at all. These are all 500x500 full size crops and all show the grain very well. For the real-world shots on all but the control, I centered on what I and the neighbors call "the big honking nest" in the trees in the woods to the rear. The nest was center-left in the control image.

K1000 with 50mm f2.0 prime lens f16:

This one above was taken with the lens set hard to infinity. When comparing with the control image below it appears a wee bit softer to me, but when viewed as a whole image on the screen, there's really no difference. The image appears normal when viewed normally.

Test target 50mm f2.0 prime lens f16:

This is the test target at maybe 6' or so, f16, daylight, manually focused (I did wear my glasses!) . It looks very normal to me.
. . . . . . . . . .

K1000 with Takumar-A 28-80mm zoom:

28mm f16:

The shot above was at 28mm with the focus set hard to infinity. Comparing it to the control photo and the prime lens photo above, it appears to be to be somewhat softer. When viewed at a normal size on the monitor it still looks quite normal.

80mm f16:

This shot above was zoomed out to the max with the focus set hard to infinity, f16. To me, this looks unacceptably soft. When viewed on the monitor it's normal or near-normal, but a 8.5x11 print does appear soft to me.

80mm f5.6:

This is overexposed. I wanted to see what it would do in the middle of the f stop range so it was f5.6 at 1/1000. The softness is REALLY showing here!
. . . . .

K1000 Ozunon 28-135 zoom:

I'm not posting the 28 or 50mm Ozunon shots since they look near normal but when I zoom it way out to 135, the softness is unacceptable at f16 and UGLY at f5.6!

135mm f16:

135mm f5.6:

Test target:

The shot immediately above is with the Ozunon zoomed out to 135 at about 10' from the target. It looks normal to me. It was carefully focused while wearing glasses.
. . . . . . . . . .

I'm not really sure what my conclusion is, and I would appreciate other opinions. I'm leaning toward thinking I might have a bad body but I'm not sure. The prime kit lens, which I really don't shoot with that much, appears to be near-normal at worst but both zooms are producing unacceptably soft images.

. . . . . . . . . .

"Control" photos:

Not the same roll, obviously, not the same light, and not the same day. These were taken with the Mamiya SD rangefinder, match-needle exposure on the outside shot, exact exposure not remembered but something like 1/500 between f11 and f8 or so. The test target was done indoor with fllash some time before.

Thanks for any advice or opinions, gang!

 

[shutterfinger]

Zooms: I always recheck the focus on them. Zoom out to the highest magnification, focus, reset to desired zoom level, make exposure provided the zoom tracks focus at all settings.
I use a Nikon Nikkor 35-135 f3.5-f4.5 zoom. Focusing at the wide to medium settings, focusing, then zooming out to 135 the focus is always off, zooming to 135, focusing, resetting to a wider setting the focus holds and results in a sharper image than if focused at the wide setting.

I suspect your camera and zooms need servicing. A thorough cleaning of the pentaprism, focus screen, mirror, and eyepeice should be done then verify the mirror is returning to the correct position.
The lens elements should be cleaned then if that does not improve performance completely disassemble the lens mechanics and apply fresh lubrication if wear is not excessive.

 

[flavio81]

Great scans, dmr.
On cameras that present focus issues, usually the little thingy that keeps the mirror at a 45° angle miscalibrates, and this throws the focus off (on the viewfinder).

If you rely on the infinity stop for tests you'll run into a lot of problems since they are not so reliable.

 

[Dan Fromm]

Hmm. Your problem is that you don't know where the plane of best focus is.

The standard way of finding where the plane of best focus lies is to set the camera on a tripod in front of a brick wall at an angle to it. The angle most often used is 45 degrees. Make a prominent vertical mark on the wall (it you don't own the wall, use masking tape and remove it when you're done). Position the camera so that the tape's image is in the middle of the frame. Put lens one on the camera, focus, shoot, change aperture, shoot, ... Do the same with lens two ... If one of the lenses is a zoom, do the exercise at shortest focal length, longest focal length, and midway between.

After the film is processed, examine it with a strong magnifier to see where the wall is in best focus.

I had a similar problem with a Nikkor zoom some years ago, eventually determined that its internals had worn and that it was no longer parfocal.

 

[dmr]

Hmm. Your problem is that you don't know where the plane of best focus is.

Thanks. This makes sense to me as a good next step. An off-line reply also opined that the infinity stops on BOTH zooms may be off.

Tomorrow I'll do some kind of a focal plane test. I don't know of a good long brick wall (my office building is block and neutral in color) but I do know a railroad track where I can focus right on a signal light and then count the track tie planks in front or in back of it where the focus is sharpest. If we have good light I'll try that tomorrow afternoon.

 

[Prof_Pixel]

I have no idea of how old you are, but in my mid-50s, I started having problems focusing my Nikon F-3; it took a +1 diopter correction lens to correct for my eye changes. (I also put one on my Pentax K1000 when I got it.) So your problem may not be your camera - it may be changes in your eyes.

 

[John Koehrer]

The infinity stops may allow the lens to overshoot intentionally to allow for expansion in the heat.

 

[Ces1um]

The infinity stops may allow the lens to overshoot intentionally to allow for expansion in the heat.

I'm pretty sure John is correct here.

Over at pentax forums a guy named Ron Boggs has quoted the following:

From the manual for the FA*600/4

"The distance scale on the lens stops slightly past the infinity mark. This allows the lens to focus on a point past the infinity scale in the case where a temperature change causes the focusing point to shift. Even when shooting at infinity confirm that proper focus is achieved through the viewfinder before releasing the shutter."


From the Pentax 67 Interchangeable Lenses Operating Manual

"Note on Infinity Indication: With a powerful telephoto lens, there may be a focus shift between the distance indicated on the lens and the actual camera to subject distance, influenced by temperature conditions. For this reason, the focusing ring is specially designed to rotate slightly past the infinity indication. So, even when you take pictures at infinity, be sure to confirm good focus through the viewfinder."

 

[dmr]

I have no idea of how old you are, but in my mid-50s, I started having problems focusing my Nikon F-3; it took a +1 diopter correction lens to correct for my eye changes. (I also put one on my Pentax K1000 when I got it.) So your problem may not be your camera - it may be changes in your eyes.

I know I don't act it, but I'm coming up on my (why am I admitting this?) 10th. anniversary of being in my mid 50s.

I've worn glasses since my teens, nearsighted, and as I went into middle age I did not have the presbyopic change to farsightedness. My nearsightededness actually increased a bit and I'm now almost up to a -3.0 correction, spherical, which has not changed in the past three exams or so. For the test shots above I did wear my glasses and was careful to focus on the center ground glass closely. I admit that I do go without glasses quite a bit and I'm probably a bit overconfident in saying that I can focus well with or without glasses. I do know that I can usually focus a RF quite well without.

In thinking critically about this, yes, human focus errors may be one factor.

I'm pretty sure John is correct here.

Over at pentax forums a guy named Ron Boggs has quoted the following:

"The distance scale on the lens stops slightly past the infinity mark."
...
So, even when you take pictures at infinity, be sure to confirm good focus through the viewfinder."

I just did a sanity check and this may be part of it. Looking at a lit porch down the block, the Ozunon zoom does NOT appear to focus through infinity, but when I really look closely, it seems to hit the hard stop before that perfect focus is reached.

Thanks again everyone!

 

[Prof_Pixel]

I've worn glasses since my teens, nearsighted, and as I went into middle age I did not have the presbyopic change to farsightedness. My nearsightededness actually increased a bit and I'm now almost up to a -3.0 correction, spherical, which has not changed in the past three exams or so. For the test shots above I did wear my glasses and was careful to focus on the center ground glass closely. I admit that I do go without glasses quite a bit and I'm probably a bit overconfident in saying that I can focus well with or without glasses. I do know that I can usually focus a RF quite well without.

I'm was about as near-sighted as you are, but needed bi-focals to read, thus the need for a +1 diopter lens over the camera viewfinder window even though I was using my glasses.

 

[dmr]

I'm was about as near-sighted as you are, but needed bi-focals to read, thus the need for a +1 diopter lens over the camera viewfinder window even though I was using my glasses.

I admit I'm too vain to do bifocals, and really I don't need them. I'm very fortunate in that my near vision is very good at reading and CRT viewing distance.

For the latest test roll for this issue, I wore my new glasses, as my prescription sunglasses are technically an older prescription, and I could very easily see the best focus point (or lack of same with one of the lenses) in that center ground glass circle.

 

[John Koehrer]

Varifocals are nice. If you want to see what's available, take a look at Zenni.com.
All you need is your prescription and PD measurements. They're not expensive either.

Not related to them, used them in the past. www.zennioptical.com

 

[Dan Fromm]

Varifocals are evil. Because their power varies continuously across the lens, the magnification they give varies continuously. They make circles look like ovals. If you don't care whether you know exactly what you're seeing, they're for you. But if you want to see reality as it is, bi- or tri-focals are probably the better solution.

 

[David Lyga]

What few people understand, even the more seasoned ones, is that if your SLR mirror is off by only a tiny amount that focus you see in the viewfinder will not match with the focus on the film plane.

On the K1000 that mirror’s front edge is held, (underneath), by merely a metal prong. That prong can easily bend either way with even relatively slight force. Alternatively, some SLRs have a set screw (Zenit, Fuji ST series) and some have a stationary disk with a screw (Spotmatic) that is purposely not precisely centered so that turning that disk raises or lowers that mirror. Either way, that mirror is held to a precise angle by that (front, underneath) anchor.

The first thing that I would do is to see by how much the focus is off. With camera securely on a tripod, and aperture fully open (for maximum limitation of depth of field) shoot, at a 45 degree angle, a 'picket fence' with the focus on the dead center. (NOTE: does not have to be an actual picket fence, but, rather, something with a 45 degree displacement. I use the reverse of an LP record jacket and focus upon the type that is half-way across. After developing the negative you will see what is ACTUALLY in focus, versus what you THOUGHT was in focus (through the viewfinder). Obviously, if there is a difference, that mirror's angle is not correct.

If the REAL (film plane) focus on that negative is FURTHER from what the viewfinder told you, that mirror's front edge is too low and needs to be very slightly raised. If the REAL (film plane) focus on that negative is CLOSER that what the viewfinder told you, that mirror's front edge is too high and needs to be very slightly lowered.

By how much? With the camera on that tripod in the same precise location, focus upon the part of the picket fence that was REALLY in focus. (In your viewfinder that picket fence area will, of course not seem to be in precise focus, because of the error.) You need to adjust that mirror’s angle so that when you set your lens at that precise location on that fence, the viewfinder also agrees with that setting. Then, and ONLY then, will VF focus match film plane focus. – David Lyga

 

[flavio81]

What few people understand, even the more seasoned ones, is that if your SLR mirror is off by only a tiny amount that focus you see in the viewfinder will not match with the focus on the film plane.

+1

I told this to the OP twelve days ago on post #3, but it seems my advice was ignored.

 

[CMoore]

+1

I told this to the OP twelve days ago on post #3, but it seems my advice was ignored.

I saw it, and i have been wondering about that since.
How does a camera tech check this.....not with a tripod and brick wall i assume.?
Thank You

 

[flavio81]

I saw it, and i have been wondering about that since.
How does a camera tech check this.....not with a tripod and brick wall i assume.?
Thank You

David Lyga has given a method that works. I will give another method. I hope Daria Morgendorffer reads this. To you, dmr, from Lima to Omaha, my advice:

Get a KNOWN GOOD Pentax K-mount lens. Don't use a zoom, don't use an off-brand lens. Get a Pentax lens that you are positive that the infinity stop is well calibrated. This is simple, just fit the lens on a known good K-mount camera, aim for the infinity (real infinity, like the mountains) and check that the focus screen shows you correct focus.

Now fit that lens onto your troubled K1000 and do the same test, you will see the focus screen does not indicate focus.

Now correct the mirror positioning stop described by David Lyga above:

On the K1000 that mirror’s front edge is held, (underneath), by merely a metal prong. That prong can easily bend either way with even relatively slight force. Alternatively, some SLRs have a set screw (Zenit, Fuji ST series) and some have a stationary disk with a screw (Spotmatic) that is purposely not precisely centered so that turning that disk raises or lowers that mirror. Either way, that mirror is held to a precise angle by that (front, underneath) anchor.

move the mirror stop a little bit and check again focus. Mess with it until your K1000 shows correct focus with the good lens.

Afterwards the problem should be solved. If your Ozunon zoom or your Takumar zoom does not show correct infinity focus afterwards, blame the lens' infinity stop.

 

[David Lyga]

Admittedly, using a lens that has a 'known' infinity that is 'correct' can be a bit misleading. I go further and want to see what THAT camera actually delivers. As much film travels along that rail (beside the film aperture) in the back of the camera, that rail actually wears down a tiny bit through the years. This might make lenses that are 'correct' for infinity actually very slightly incorrect with THAT PARTICULAR CAMERA!!!

What to do? Again, I shoot the 'picket' fence and focus (lens WIDE OPEN FOR MINIMIZATION OF DEPTH OF FIELD) very, very accurately upon the center of that (well marked) 'fence' (it could be a record jacket 45 degrees from the camera) and shoot (slightly underexpose so you can clearly see the focus point). NOW you have (upon negative inspection) the TRUE focus point on that fence, but now indicated on the actual film plane. If that determination does not agree with what you THOUGHT was in focus in the viewfinder, you now know that a slight correction must be made. Again, what you see in the VF to be in STRICT focus has to also be in STRICT focus on the film plane. Too many simply state that their respective lenses are not all that sharp and do not realize that it is the focus that is off. Don't become a slave to what the camera says!!!!!!! - (Did I add enough exclamation points?)

Once you align VF focus with film plane focus, ANY lens you mount on your SLR will deliver a matched 'VF/film plane' correct focus. - David Lyga

 

[flavio81]

Admittedly, using a lens that has a 'known' infinity that is 'correct' can be a bit misleading. I go further and want to see what THAT camera actually delivers.

You are correct...

As much film travels along that rail (beside the film aperture) in the back of the camera, that rail actually wears down a tiny bit through the years. This might make lenses that are 'correct' for infinity actually very slightly incorrect with THAT PARTICULAR CAMERA!!!

... but not for that reason.
The film gate, on 35mm SLR, has a rather wide space in which the film sits. The film doesn't sit flat, it curls, one part of the film sits against the "pressure" plate, but it does not seat fully. The film edges "float" between the space created between the "Pressure" plate and the innermost film rails, the center part of the film sitting against the plate, hopefully. This i recall from memory from a book written on the subject of the subtle details of camera technologies.

BTW the "pressure" plate name is a bit misleading name since it does not press the film; it presses against the outermost film rails. That is, the top and bottom film rails.

What appears as "wear", from when i've seen on my cameras, is just corrosion. The cameras (in my collection) that don't show this corrosion have really nice shiny film rails.

You are 100% correct on that the best test is with actual film, since this takes also into account the curl.

I do have seen misaligned cameras, as you mention, but in my case it was due to the camera falling down and then the lens mount being knocked out of alignment. I've seen this on not one but two Spotmatics. This was a BIG, significant misalignment. But you don't see the defect until you take a careful look.

The other cause is when a techician disassembles the machine but does not put the lens mount shims in the correct place (damn peruvian technicians!).