[teoprome]

Hi, I read a lot about use manual lenses on Pentax cameras, so I have 'little' problem with my K3 and M lenses

When I use a manual lens (example a K mount lens with a lever on the back and no contact pins) in AV on my K3, it remains wide open and the measurement of exposure is correct, as expected: all ok.

In manual mode, I use this procedure:
set aperture on lens, measure exposure with green button (I can also see proper DOF preview), then release shutter

what I get is this:
with aperture 2.8, metering is correct, shot exposed fine
with other apertures, it seems that aperture values will not be considered on shot; exposure seems to be correct based on setted aperture, DOF preview seems to be correct, so when I take a shot the photo is overexposed...

example of time/f when stop down

measured time 1/15 at f/2.8
measured time (stopdown) 0.3'' at f/8

the value measured at f/8 is correct, 3 stops of difference... so shot is very overexposed

It's possible that the lens is stopped down correctly on DOF preview and the desidered aperture not read when the shutter is released?
I'm a little confused...

thank'you for all of your tips and experiences

[Ole]

You must enable the use of the aperture ring in the custom menu.

Refer this article for more detail on how to use M and K lenses.

[teoprome]

You must enable the use of the aperture ring in the custom menu.

Refer this article for more detail on how to use M and K lenses.

Thank's Ole for your response, yes, I've enabled that setting in the custom menu...
I have read the article so I continue to have problems
the lens seems to work properly when not inserted in the camera, and when I change aperture values




I press the green button and do the (right) measurement, release the green button then press the shutter and get shot overexposed

What other tests can I do?

[Ole]

That's weird.

With the lens unmounted, check that the aperture actually stops down when you push the lever, and verify that the lever isn't bent out of shape so that it wouldn't engage with the mechanism in the camera.

When you hit the green button in "M" exposure mode, does the lens actually stop down?

[teoprome]

With the lens unmounted the aperture stops down when I push the lever, the lever is not bent
When I hit the green button in "M" exposure mode the lens stop down

other possible thing would be the size of the lever incorrect...? the lens is a REVUENON 28mm

as soon as I have time I'll try with another lens; I have a Pentax KA 50mm, I'll try with the lens set in 'non A' mode.

Thank's,
I'll let you know soon about this test

[stevebrot]

Welcome to the Pentax Forums!

I am with Ole. What you are describing is not what is expected. I have been able to use my lenses without the "A" contacts with good success on my K-3 and have not had the problem you describe. I would also suggest that you confirm that the lens actually stops down during the exposure. Set the aperture ring at its narrowest (largest f-number) and the shutter speed at a long value (several seconds) and confirm that the lens fully stops down by looking into the lens while you press the shutter release.


Steve

---------- Post added 06-04-14 at 08:00 AM ----------

other possible thing would be the size of the lever incorrect...? the lens is a REVUENON 28mm

as soon as I have time I'll try with another lens; I have a Pentax KA 50mm, I'll try with the lens set in 'non A' mode.

Ummmm...any chance that your Revuenon had been subjected to surgery by a previous owner? FF Canon users regularly shorten the actuator levers on adapted K-mount lenses to remove mirror interference on those cameras. Of course, doing so ruins the lens for use on K-mount cameras.

Steve

[Ole]

OK, not a Pentax lens. Is the rear of the mount conductive (like Pentax lenses) or painted black? Not that I know that would matter - since the lens stops down when you push the green button one would think it should work.

[OregonJim]

Another thing to check is the snappiness. Do the aperture blades instantly "snap" from wide open to minimum? A little oil on the blades can slow them down enough so that at the time of exposure, they haven't fully closed yet.

I would expect it to affect green button exposure as well, but the timimg may be different from "actual" exposure.

[ChristianRock]

You didn't tell us what lens you are using.

I used to have a Sigma UC 70-210 lens and whenever I used the green button with it, it would severely overexpose the pictures - by at least two stops - wide open a bit less, it got worse as I stopped the lens down. There were no iris problems. My other manual lenses also overexpose with my K20D sometimes, but it's usually between half step and a step. I've read about other people complaining that old Sigma lenses do this.

Also, if you are using spot metering, it's easier to over or under expose, as I found out - it just depends on the color of what's in the center when the meter is activated...

[teoprome]

thanks to all for the answers and the tests that you have proposed,
Stevebrot, I looked into the lens while I press the shutter release with aperture ring f/22 and t=3'' and the lens fully stops down
Ole, the rear mount is conductive
OregonJim, the aperture blades instantly "snap" from wide open to minimum

ChristianRock, after other 'home made' tests what you say seems to be the problem....
I don't use spot metering except in special situations, yes it's easier under or over expose
the problem seems to be the center weighted metering, that with K lenses, overexpose a bit compared to KA lenses with lens set to 'A' mode

here are the 'home made' test, I shot a pile of books with diffuse lighting and a little spot light on the right (to better see overexposed areas)

1. with lens SMC Pentax-A 50mm f/2 in 'A' mode

Multi-segment measurement (green) t=1/40 at f/2.8 - t=1/5 at f/8 exposure good little area overexposed (as expected)
center weighted measurement t=1/30 at f/2.8 - t=1/4 at f/8 exposure good (not like the previous, but good)


2. with lens SMC Pentax-A 50mm f/2 NOT in 'A' mode
center weighted measurement (the only avalaible in this mode) t=1/30 at f/2.8 - t=0.4'' at f/8 overexpose a lot


3. with lens Auto Revuenon K 28mm f/2.8
center weighted measurement (the only avalaible in this mode) t=1/30 at f/2.8 - t=0.4'' at f/8 overexpose a lot

I know that this is only a 'home day' test so...

the best measurement is with a Ka lens setted in 'A' mode and multi-segment measurement
with a KA lens setted in 'A' mode and center weighted measurement, the measurement is not bad, only a little overexposure, about 1/3EV
with K lenses, or KA lenses not in 'A' mode, the measurement visibly degrades as you close the aperture, 2/3EV-1EV at f/8, more at f/16

This says that multi-segment measurement in K3 do a very good job, due to the sensor with 86K-pixel
This also says that with old 'A' lenses also the center weighted measurement do a good job (overexposed areas may occur with this measurement and lights in the scene, I think it's normal)

With K lenses on K3 we have only the center weighted measurement and spot, I don't know why multi-segment measure is not settable (do you know why?);
I try only the center weighted measurement, and it does not do a great job: I don't think it's only a lens problem, but also a problem of the calibration of the measurement in the camera (what do you think about this?), so in the future I'll try with other lenses

I think that the best option in this mode is to measure exposure at the largest aperture, record the time proposed by the camera, then close the aperture and calculate mentally the time to set; alternatively set desidered aperture press green button and then compensate the time down for a minimum of 1EV at f/8 and more at f/11, f/16...

[Ole]

Your tests reminds me of a discsussion some years back here on PF.

As I recall the culprit is the modern bright focusing screen which the light meter meteres off of. The screen doesn't get darker by exactly two stops from an F1.4 to an F2.8 lens for example. It is non-linear.

So the difference in brightness of the screen between an F1.4 lens and a F2.8 lens is not two stops. Hence the light meter gets fooled, but only with M and K lenses. For A and newer lenses, the camera "knows" the aperture of the lens and compensates.

It is thus not possible to achieve consistent exposure across the aperture range with M and K lenses. You need to dial in EV compensation based on experience.

In live view this inaccuracy should not be present I'd assume. Live view wasn't around when we had the above discussion years back.

[teoprome]

ok it might be a correct explanation, thank you

[stevebrot]

Your tests reminds me of a discsussion some years back here on PF.

As I recall the culprit is the modern bright focusing screen which the light meter meteres off of. The screen doesn't get darker by exactly two stops from an F1.4 to an F2.8 lens for example. It is non-linear.

So the difference in brightness of the screen between an F1.4 lens and a F2.8 lens is not two stops. Hence the light meter gets fooled, but only with M and K lenses. For A and newer lenses, the camera "knows" the aperture of the lens and compensates.

It is thus not possible to achieve consistent exposure across the aperture range with M and K lenses. You need to dial in EV compensation based on experience.

In live view this inaccuracy should not be present I'd assume. Live view wasn't around when we had the above discussion years back.

Ole, while wildly inconsistent stop-down metering (generally 1-2 stops underexposure at wider apertures) was a huge problem on the K10D, K20D, and other Pentax bodies of that time, it is no longer an issue on the newer bodies. My K-3 works quite happily with my K and M series lenses as well as with adapted M42 glass.

One thing we forgot to ask the OP is whether the stock focus screen is being used. I recently had to stop using a focusingscreen.com screen on my K-3 due to metering problems. If an aftermarket screen is installed, that may be the cause of your problems. As Ole suggests, a comparison with live view might be helpful.


Steve

[teoprome]

I use the original focus screen, the problem exist and is well visible and measurable on my K3; it's not a big problem I've only to compensate changing exposure time, so in landscapes, photos of architecture, in general static shots when we have time is not a big problem.
when I do portraits of my son, who is continually moving, is not easy to take shot with this type of objectives and this problem, however is a good way to improve skills

Honestly I was expecting something better with this camera regarding the center weighted measurement with K M lenses
When I have time I'll make other tests, also with live view, shooting modes that do not particularly like

Teo

[stevebrot]

Honestly I was expecting something better with this camera regarding the center weighted measurement with K M lenses
When I have time I'll make other tests, also with live view, shooting modes that do not particularly like

Sorry you are having these problems. I have found that the stop-down metering on my K-3 with legacy glass* is quite acceptable and agrees well with open aperture readings of the same scene. Most of my photography is done with older lenses and after years using the K10D (legendary for flaky stop-down metering) it is so very refreshing to have an accurate metering system on my K-3.

My advice is to carefully consider your technique and whether the aperture mechanism on your M-series lens is fully operational. If you decide to do some comparison testing, I would suggest a evenly-lit blank wall as the metering target with an "A" contact lens as the metering standard. First record the meter settings for various apertures with the "A" contact lens. Then record the stop-down settings with your M-series lens (same focal length) at the same apertures. Use center-weighted metering for bot lenses. The variance of shutter speed should give you a good idea of comparative accuracy.


Steve

* I shoot a mix of K, M, and M42 lenses