[Lititz]

(Pardon me if this has been discussed before; I've looked at other threads and didn't find anything quite like this question.)

Exposure inaccuracies when using K and M lenses on at least some DSLRs are well documented. From what I've read, these exposure/metering problems arise because between the M and A series lenses Pentax changed the way f-stops were calculated (something like going from linear to non-linear measures, or using diameter of the opening vs. area of the opening--sorry, I've forgotten those details at this point). Here's my question: do the same kind of exposure inaccuracy problems arise going the other way, that is, does using A-series or later (F, FA, DFA) lenses on early K-mount film bodies (KX, K2, ME, etc.) throw off the camera's metering? I tend to assume there's no problem, given that in the late 1980s Pentax was selling A-series lenses on the ME and MX bodies. But does anyone have any first hand experience with this? (Why am I interested? I use M and K lenses on my K film bodies, and was considering purchasing an A lens.) Thanks for any insight.

[Lowell Goudge]

No.

The issue is not going the other way because before the A series cameras the aperture lever was a switch. It always stopped down the lens to the selected aperture. The aperture coupling was not removed until I think the *ist series cameras. Much later

Edit note.

The issue on DSLRs is with respect to the metering and the scatter of light off the focusing screen. from testing I did when the issue was first reported, it is a function of the native aperture of the lens, or the aperture when doing actual stopped down metering on KSeries lenses.

The camera bodies in A series lenses know the maximum aperture and compensate for this issue, when metering in a mode. This was best exhibited by the error introduced when adding a teleconverter with A contacts to an a series lens. Exposure was perfectly flat, just shifted by a constant error because the TC modified the native aperture of the lens without changing the A contact information.

[vonBaloney]

No.

The issue is not going the other way because before the A series cameras the aperture lever was a switch. It always stopped down the lens to the selected aperture. The aperture coupling was not removed until I think the *ist series cameras. Much later

That's what I would figure, but why would it be a problem the other way around on a modern camera if you are using it the same way? If I am using a K or M lens on a DSLR, that means I'll be in M mode metering with the green button, which actually stops down to take a reading (so that should be accurate). It is actually using 'A' lenses normally where it is guesstimating exposures that I'd expect to see some inaccuracies, or using K/M lenses on older cameras with non-crippled mounts where it was essentially doing the same thing mechanically instead of electronically.

[Lowell Goudge]

That's what I would figure, but why would it be a problem the other way around on a modern camera if you are using it the same way? If I am using a K or M lens on a DSLR, that means I'll be in M mode metering with the green button, which actually stops down to take a reading (so that should be accurate). It is actually using 'A' lenses normally where it is guesstimating exposures that I'd expect to see some inaccuracies, or using K/M lenses on older cameras with non-crippled mounts where it was essentially doing the same thing mechanically instead of electronically.

See my edit note to my post. I spent a long time looking into this, and found for example, the K10 screen, which is identical size to then*istD had horrible non linearity in K lens metering, but when I put the *istD screen into the K10 it was almost perfect (like the *istD)

I think it is a function of the way light hits the screen, and the impact of the scatter off the screen, as it hits the metering segments, that causes the issue.

[Lititz]

Thanks for all the very helpful information!

[dakight]

Issues like this are exactly why I almost always use a 3 exposure bracket. Yeah, it uses more storage and takes a little longer to process but I almost always have at least one of the three that has a useable exposure. It's not only inaccuracies in the metering but also different light conditions can and often do confuse the meter. So a set of bracketed exposures with + or - 1 or 2 EV almost always yields something at least close to correct.

[Lowell Goudge]

I test all my lens and body combos for exposure. Most follow reasonable rules , sick as my Tamron 28-75-2.8 which is spot on wide open but for some reason, the aperture lever is not calibrated correctly and exposure drifts up gradually as you stop down, t being +1 at F32. Knowing this makes me conscious of the errors so I check thenhistogram and set EV comp when using this lens

[wombat2go]

I only have one -A lens among my collection of -M.
I rarely use the SMC Pentax-A 1:1.4 50mm , but took it to Korea with the Pentax MX one time, and did not notice any exposure problems.
This film was Kodak iso 800 home processed in C41 Digibase
Note that the Pentax MX has a limited meter coupling range with faster and slower films and users should be aware of that, as explained in page 22 of the manual.
As I recall the second photo had the lens at f/2 or so and was metered with vf centre on the bright part of the road before framing.
https://app.box.com/s/omlyj78yl9m884dq7kdv
https://app.box.com/s/oo6slnkpofcjoh4zsouf

[From1980]

See my edit note to my post. I spent a long time looking into this, and found for example, the K10 screen, which is identical size to then*istD had horrible non linearity in K lens metering, but when I put the *istD screen into the K10 it was almost perfect (like the *istD)

I think it is a function of the way light hits the screen, and the impact of the scatter off the screen, as it hits the metering segments, that causes the issue.

Great work. Possibly know if K20 is similar? Recently got one and haven`t yet tried older lenses on it. Anybody know of any exposure trends to watch for?

[Kerrowdown]

Possibly know if K20 is similar? Recently got one and haven`t yet tried older lenses on it

I'm still using K20D's with exclusively A series primes... pleased to report no real issues of problems with exposures at all.

[From1980]

Thanks for the experience.

[Lowell Goudge]

Great work. Possibly know if K20 is similar? Recently got one and haven`t yet tried older lenses on it. Anybody know of any exposure trends to watch for?

K20 meters just like K10. Badly on non A lenses. Somewhere I posted how all the bodies performed from *istD through K5. It is a chart showing grey scale vs f stop. Note 45 grey scale =1 stop between 25 and 230 greyscale. The ends are non linear but the middle of the range has about 5 uniform stops

[Canada_Rockies]

K20 meters just like K10. Badly on non A lenses. Somewhere I posted how all the bodies performed from *istD through K5. It is a chart showing grey scale vs f stop. Note 45 grey scale =1 stop between 25 and 230 greyscale. The ends are non linear but the middle of the range has about 5 uniform stops

When I was looking into this discussion in the dark ages, I solved two of my problems at the same time with my K10D by installing an LL-60 screen from the *ist D series. The other problem is leaning buildings and trees. I seem to lean to the right all the time. I'm not sure if it is because I've been married almost 50 years and She Who Must Be Obeyed usually is on the left where I don't hit her with the K10D on its wrist strap or some other reason, but I do lean. The 1/3 lines in the screen help me tell which way is up

[vonBaloney]

When I was looking into this discussion in the dark ages, I solved two of my problems at the same time with my K10D by installing an LL-60 screen from the *ist D series. The other problem is leaning buildings and trees. I seem to lean to the right all the time. I'm not sure if it is because I've been married almost 50 years and She Who Must Be Obeyed usually is on the left where I don't hit her with the K10D on its wrist strap or some other reason, but I do lean. The 1/3 lines in the screen help me tell which way is up

Some of the K10Ds actually had a unlevel horizon problem, might not be you.

[Canada_Rockies]

Never thought of that! The grid lines help quite a bit, so I think I'm a little bit out of warranty to claim the tilt. I bought the K10D in November 2007.