Exposure question

Yes, I have tried it

Yup, I can confirm that is true, and this technique works on my K100D, K20D and my M42 lenses. I should have remembered it doesn't work on PK M lenses, but it's been a while since I used my one and only PK M lens. With an M42 lens the aperture definitely changes size if you turn the aperture ring while the lens is mounted on the camera. On my three Takumars you need the auto/manual switch on the lens to be set to manual for this to happen.

Chris and Mannesty you're both right, and YES, i was wrong, and talking out of my ass...

My mistake is an honest one though, as I usually use the M100 macro in Manual mode, and always get a correct exposure by pressing the green button, or just by arranging the shutter speed myself (Which i'm quite used to doing). When I wrote the post I had just been shooting a few macros in Av with the same lens, and indeed, it was metering perfectly - but only wide open. Same subject in the same light lead me to think it was automatically metering, whilst it obviously wasn't. THerefore my mistake SORRY!!!


The meter does work in manual mode though, and this is where the problem still hasn't been solved: There might be physical (:twisted limitations to the lens' performance in bright light. In the end, any lens will need to be stopped down whilst shooting in those conditions. WHat I also wrote was that the two lenses are different, and unless there is a problem with the 135, its very possible that there are few stops difference between the 2.

Question is, how overexposed is your shot?
And also, does the problem appear in any other condition but bright sunlight?

CHeers

You know, I always find great respect for someone that makes an honest mistake, and 'fesses up!!

No worries - and we all live in glass houses, so no stone throwing from me

As it happens, I have the SMC-A 100mm F4 macro, and as it happens, it's overcast today with very consistent, even lighting. So a direct comparison of A-mode, and manual with a lens similar to yours (is yours the F4 or the F2.8?) was possible.

I've just done a very quick check with my K10D... I picked a reasonably well lit, not particularly contrasty scene, lit naturally from the window, tripod, covered viewfinder...

... and wrote down the metered exposures at ISO 200, using the lens first set to A, varying the aperture from the camera in Av, then using the aperture ring with the camera in manual, using the green button for stop-down metering:

F-----A-------M
4-----1/80--1/60
5.6---1/40--1/30
8-----1/20--1/10
11----1/10--1/5
16----1/5----0.4
22----0.4----0.8
32----0.8----1.3

Note the exact reciprocity in 'A'*, and the overexposure using stop down metering. I repeated the 'A' metering to ensure that the light hadn't changed, and it hadn't.

This trend of progressive overexposure with decreasing aperture exactly corresponds with what I found with several lenses when I did all this with the K10 when I first got it.


I am intrigued that you find the stop-down metering accurate - was this with the K-7? Either way, I'd urge you to have a bit more of a careful look at this.

Maybe the accuracy of stop-down metering has been improved with the K-7 which would be good news indeed.

_______

* this reciprocity shouldn't be a surprise of course - since the camera meters wide open, it's calculating the exposure required for the smaller apertures, not measuring it.

To be fair to Pentax, they say quite openly that the metering with older lenses may not be accurate, in the instruction book.

At least older lenses can be used and the assumption I would make is that those of us who are savvy enough to want to use these lenses will be able to sort out the slightly more inconvenient working methods.

In contrast, users of the Olympus 4/3 cameras will find a very expensive adapter for their OM lenses and one that can only be used at all at a limited number of apertures.

Personally I've totally moved on to A lenses and later, which makes the whole image making process much easier.

Ohh Absolutely CHris!

All the 3 K10's i've had gave me consistently overexposed exposures exactly the same as yours. You're absolutely correct.

I have been doing all my shooting with the K-7 recently, and YEs the metering is greatly improved even with manual lenses. (Mine is the M100 f4)

By using the green button to meter at each exposure these are the results. (Same lighting conditions as you... )
Note - although the lens shows 7 f-stops on the barrel, there are indeed 10. I therefore don't know what exact apertures are in between the stated ones.
f-stop
4 1/1000
5.6 1/400
8 1/250
11 1/100
16 1/50
22 1/30
I then get 1/20 and 1/15 (somewhere over the aperture range there was an f-stop not mentionned)
f32 1/8

Just to make sure, all on K-7, the shots are PERFECTLY metered and the results absolutely consistent.

(another reason to buy the new toy Chris? )

Nah, the lens is fine in all other respects. It's just in bright noonday sunlight that I've had this overexposure. Closing down the aperture a couple of stops solved the problem.

I will have to do some more tests I think. Perhaps I cocked things up somehow (wouldn't be the first time )

Edit: Hmm start writing this half an hour ago, then distracted by the phone. This was a reply to Nathan's 10:13 a.m. post.

Actually John, although it appears they do, I'd argue that they don't, at least not with stop-down metering with the K10D.

I expect the reference you're thinking of is p210 in the K10D manual, where it refers to exposure errors with wide open metering...

"The shutter speed changes in relation to the open aperture but an exposure error may occur"

However, the reference in the manual to stop-down metering is on p158, in reference to the green button:

"Shutter speed is adjusted to appropriate exposure according to lens aperture when lens aperture is not set to A position."

I think my F4, 1/60 exposure reading (wide open) above is probably an example of the former error, which they do mention explicitly.

The overexposure while stopping down isn't consistent with their use of the word "appropriate", in my view.

Indeed, although wouldn't it be nice with an old M lens set to manual, to poke the green button and have it get it right!!

Me too!!

That really is great news.
And it's my birthday soon!!

Why is it just not plausible that the lighting conditions are too bright for that lens?

I was getting 1/5000 on f4, (ISO100) by simply spot-metering the very over-cast sky over London, just now. I can safely say that a bright sunlit sky is going to be various stops brighter than that, and I would therefore need to stop the lens down to have the right exposure.

If I shoot with the 50-135 (@100mm) on f4, in the same situation, I get 1/3200. And compared to the scene described in the previous post, I get between 1/640 and 1/500 where the M lens was 1/1000.

Yes it is plausible I was just making an assumption that a longer lens @ f2.5 would be darker than than a shorter lens @ f1.8. I don't know why I assumed that, it just seemed to make sense.

I wonder if light absorption by the glass may have some impact as well, since the 85/1.8 is a 6 element lens (diagram here, it's the 6th variation 1972-74) and the 135/2.5 is a 5 element lens (diagram here).

The other exposure error that has always been the case is with the SMC Pentax 50mm f1.2 lenses, both plain and A versions.

With this lens we need to dial in 1 stop of under-exposure.

Presumably the mater recognises the f1.2 aperture as one stop adrift for some technical reason.