[iDon]

Hey,
my K-3 and also my old K10d show an odd behaviour when I use them with K or M lenses.
Mode dial is set to M, use of aperture ring is allowed und green button is used to set shutter speed.
Both cameras tend to overexpose with smaller apertures. The more I step down, the brighter are the results.
Oddly enough metering works perfectly fine, when I do the same procedure in live view. I can reproduce this behaviour with all my manual lenses and also A/F/FA lenses set to another position than A. All those lenses work completely fine with my film bodys.
Do you have an explanation or even a solution for this issue? I'd really like to use my K lenses more often...

iDon

[jatrax]

Do you have an explanation or even a solution for this issue? I'd really like to use my K lenses more often...

No solution really but that is typical. Theory is that lenses with ID chips allow the camera to be more precise with the metering but all manual lenses do not have that. Several threads discussing this over the years, mostly with Takumars but the same applies to K and M. F and FA lens should not be showing much of an issue though, they seem fairly consistent across apertures. Some require exposure compensation but I've not noticed it was inconsistent like the all manual lenses.

[seventhdr]

One reason is that Pentax changed how the aperture lever moved from the K and M Lenses to the A and later lenses.

From the K-mount article on the home page:

It looks the same as on the "K" mount but it is now calibrated differently and can stop the lens down very accurately by fractions of F-stops as determined by the camera's meter.

Read more at: The Evolution of the Pentax K-mount - Articles and Tips | PentaxForums.com

Also from Bojidar Dimitrov's K-Mount page:

The least-noticeable change is in the movement of the diaphragm actuator. In the original K-mount its displacement is proportional to the diameter of the diaphragm opening. In the KA mount, it is proportional to the area of the diaphragm opening, and thus to the selected f-stop. This leads to an identical stop-displacement between any two consecutive f-stops, and thus greatly simplifies the operation of the body in the Tv and P operating modes. All KA lenses have the same stop-displacement.

The diaphragm actuator's change in operation does not lead to any incompatibilities with older camera bodies. This is because the diaphragm release of these bodies always moves all the way up, and allows unrestricted travel of the lens' diaphragm actuator.


Hope this helps.

Regards

Chris

[Ole]

The focusing screen is the main culprit when it comes to metering inaccuracies when using the viewfinder and M/K lenses. Today's focusing screens are designed to be bright when the lens has a large aperture (fully open). But they get relatively too dark as the lens is stopped down and fools the meter, making the camera overexpose. Note that in view finder mode the light meter meters off of the focusing screen.

This is not an issue with "A" (and newer) lenses with the aperture ring set to "A" (if a ring is present) because the camera meters with these lenses wide open no matter what aperture is set. M and K lenses are metered "stop down", i.e. at the actual aperture set which fools the meter on today's cameras.

In live view there are no issues, since the focusing screen is out of the picture so to say.

[iDon]

I've read about this issue some time ago in another forum (can't find it anymore), but I haven't found a proper explanation.

In my understanding stop-down-metering (=using the green button) should be a simple and reliable solution. It stops the lens down to the set aperture and measures the light coming through.
Isn't that completely indipendent of lens transmission, diaphragm actuator control curve, algorithms etc.?
Shouldn't this method be more accurate than an aperture simulator found "decrippled" film bodys and todays Nikon cameras?

@seventhdr: this is an explanation why you can't just modify a K lens to an A-lens, but I'm afraid it doesn't explain this issue, because the camera doesn't need to know which aperture is set at the lens. It just needs to know how much light hits the sensor.

You can try to use a fairly new FA lens and test it with aperture ring set to A at every aperture value versus setting the aperture value using the aperture ring + green button. I'm sure the results will differ significantly and will show the overexposure issue.

edit:
@Ole:
thank you very much! That should be the reason...

[UncleVanya]

At low ISO with long exposures in see significant underexposure in viewfinder based vs live view based exposures.

[cbope]

One reason is that Pentax changed how the aperture lever moved from the K and M Lenses to the A and later lenses.

I don't believe this is valid for Pentax DSLR's, since they all have a "crippled" lens mount that is missing the lever to read the aperture position set on the lens. This is why you must use the green button to perform stop-down metering with manual lenses that do not have an A(uto) setting on the aperture ring.

I haven't really noticed this to be a significant problem with my manual lenses, but I rarely use them with small apertures. I have seen some small variations at larger apertures though, but nothing that can't be dealt with in LR since I shoot raw.

[iDon]

I haven't really noticed this to be a significant problem with my manual lenses, but I rarely use them with small apertures. I have seen some small variations at larger apertures though, but nothing that can't be dealt with in LR since I shoot raw.

if you use smaller apertures (let's say starting at 5.6) and don't correct your exposure after pressing the green button, the results will be too bright, even for RAW processing

So I've come to the conclusion that I'm going to use my manual lenses, but with live view only... focus peaking is also a big help.

[Stagnant]

At first I tried using manual lenses by focusing through the viewfinder, but I quickly figured out that it is near to impossible to nail the focus correctly. I switched to LV mode and I use both focus peaking and magnification, especially if I shoot portraits. Even then the exposure can sometimes be not quite as precise as with modern automatic lenses.

[stevebrot]

All those lenses work completely fine with my film bodys.
Do you have an explanation or even a solution for this issue?

No solution really but that is typical.

jatrax is essentially correct. It is typical for stop-down metering across brands (e.g. Nikon, Canon, etc). Conventional wisdom is that the cause may be traced as follows:

• PDAF systems use a half-silvered main mirror which results in less light to the viewfinder
• The stock focus screen and open-aperture metering system are optimized* to work together to provide a bright viewfinder image and compensate for the half-silvered mirror
• Stop-down metering creates an additional challenge since the screen brightness varies according to both the light and the size of the aperture pupil. In simplest language, the brightness of the screen is not a consistent representation of the amount of light to the sensor.

The result is that stop-down metering may vary as much as 2 EV through the aperture range depending on camera model and lens. The problem was more acute on the older Pentax dSLR cameras (K10D, K20D particularly) and is much less evident on current generation Pentax cameras.** On current cameras, the meter will apply different processing in M-mode when the A contacts are not detected. Stop-down metering should be essentially linear. Av mode will still work for manual aperture lenses, but results will be less consistent.

Now...having said that, here are the notes:

• Don't expect to get the same EV from stop-down metering as for open-aperture. Dispel the notion that one is "right" and the other less so. Neither is "right". Both are approximations for acceptable exposure.
• Do expect that stop-down metering will be essentially linear for the range of lens aperture in adequate light
• Potential still exists for significantly inaccurate metered EV in dim light at narrower apertures. The meter still has a lower range for linearity and it is based on the amount of light reaching the sensor. If the ambient light would normally meter at EV 4 (f/2.0 and 1/4s @ ISO 100), the amount of light to the meter sensor would only be EV 0 at f/8. This is at the raw edge for meter linearity for SPD meters. The RGB sensor on the K-3 is supposed to be good for an additional -3 EV, though I would not push it. This is true regardless of the ISO setting.

Steve

* The focus screens are designed for optimum brightness. This has been described to me as "scavenging" off-axis light to increase the apparent brightness of the frame. How well this works varies by the aperture size. The open aperture metering system compensates for the false-brightness by factoring in the lens maximum aperture when calculating the appropriate EV.
** The "fix" was first implemented on the K-7 and migrated to the rest of the line-up in the years since that camera debuted.