Stop-down metering on the K-5

Finally got around to putting the K-5 on a tripod and writing down the metering under reasonably consistent lighting with a couple of lenses.

In summary, the K-5 stop-down metering is much, much better than the K10D's.

It doesn't quite follow the arithmetically flawless reciprocity you get in A mode, setting the aperture from the body, but it's not far off.

If you're interested in the numbers, here they are:

[IMG]http://i162.photobucket.com/albums/t244/chris5gd/Technical/StopDownMetering-K5-2011-10-01_1439.png[/IMG]

At first glance it looks like the stop-down results for the 28-200 are about a stop wrong all the way down, but I think this is probably due to the fact that the marked apertures are really about a stop slower as you approach the long end. So the right hand column needs to move down one place really - which makes the exposures much more consistent with A mode.

The camera was on a tripod, and I had my eye jammed up against the viewfinder (I didn't bother with the viewfinder cover). I ran through them as quickly as possible, and there's no cloud today so I think the lighting was pretty consistent.

It looks to me as if stop-down mode is good to about half a stop, on this showing.

K and M lenses are usable again!

I think the more recent DSLRs are better at this. I know I don't have any problems with K or M lenses on the K-m down to f11 or so, but others have nothing but trouble on older bodies.

Weirdly the same people advocate using M42 lenses instead with the camera set to Av and using the lens like a preset one. If I try that I get truly hopeless underexposure! I think older bodies were better with the adapter but newer ones are better with native K lenses.

It's a lot easier Chris just to photograph a white or grey card
and read off the RGB value in any editor.

A piece of tracing paper over a filter will also work.

Generally your meter should read right at f/5.6 and then you
will need a +/- constant applying to the exposure.

I've just checked a Pentax-A 50mm f/1.7 lens and apart from the
1st reading it's constant right through:

All 3 RGB values were the same so I'm just listing one value on each line:

ExpoDisc shots at sky on Av
f/1.7 - 105
f/2 - 94 About 1/3rd stop too dark
f/2.8 - 92
f/4 - 94
f/5.6 - 93
f/8 - 93
f/11 - 92
f/16 - 93

I don't really see what you're driving at I'm afraid.

I'm not concerned with the aperture priority results in A mode at all - just whether stop-down metering (in manual mode) gives exposures that are close to what you'd get by metering wide open, and then doubling the resulting exposure for each stop away from that wide open aperture you want to shoot at.

On the K10D it doesn't, by a country mile. On the K-5 it does, more or less.

Given that half a stop either way is very rarely enormously important (unless that half stop happens to burn out the brights), it's clearly good enough for K and M lenses - set the lens at the aperture you want, leave the camera in manual, and just before you take the picture, press the green button to set the exposure.

I forgot to say, by the way, in my OP, I took all sets of readings at ISO 200. I didn't try it at higher sensitivities.

These are the sort of posts that must attract the casual passer bye to consider using Pentax cameras. Why buy another brand that does not have these issues when you could by a Pentax body and spend time working out how to get correct exposures with old lenses on various recent bodies.

I have a shelf full of older glass. Do I use it often? No I don't.

For the same reasons that people don't try to get an Intel 286 to run MS Office 2010. Today you can take photographs with Pentax lenses that do meter and work well, tomorrow you may be dead.

Just a point of view. Otherwise, keep up the good R&D.

Unless I'm misunderstanding you, this seems rather unnecessarily sarcastic.

Optically, the old K and M lenses are excellent, and incredibly cheap.

The 286/Office 2010 analogy is misleading, since new computers are cheap, unlike new lenses, which are expensive.

In a sense I agree with you - I sold all my M lenses when I discovered that the metering didn't work right on the K10D. And A-series lenses are fine, but they're quite few and far between for that very reason, as well as being much more expensive than the M ones.

The fact that K and M lenses are now convenient to use with the K-5 and other recent bodies is surely only a good thing.

I'd consider getting one again now, whereas before I certainly wouldn't.

The ISO shouldn't matter within reason. 200 should be OK.

Sorry about the confusion I just posted the A lens results to show
that very accurate exposure in terms of the result
is possible with the K-5

Speeds for some reason do vary from what they should be, possibly
some correction in the camera. I noticed on the K20D that it quoted
an Effective LV which was different to the LV metered in each segment.
The K-5 doesn't seem to give the Segment LV's in PhotoMe

When I look at the figs for the 'A' Lens posted above the speeds are correct at
f/4 and f/11 1/800th and 1/100th (3 stops) and they are
correct at f/2.0, f/2.8 and f/16

The speeds given for in between aren't right, but the exposure was
correct going off the RGB values? Speeds are rounded off but not
by this much?

They are:

f/5.6 - 1/320th (should be 1/400th)
f/8.0 - 1/125th (should be 1/200th)

I think if you just meter at f/4 or f/5.6 and adjust the
compensation +/- so that it looks right you won't be far
out.

If there's any white in the scene the burnt out indicator
will show indicating that you need to reduce the exposure.

I've just realised This is all wrong
The exposure is being metered for each shot and
the Speed is a Variable

So if the lighting changes the speed will vary... which
is what must have happened with two of my exposures,
these were done outside

The one's that Chris has done will also vary unless
they were taken with Constant Daylight Lighting equipment.

For me, the stop down metering of M lenses works well enough on my K7 - and I can't recall it being terrible on my k200d either. I expect to have some input into the exposure anyway, even on a DA lens. It's interesting that people do these tests, but they only confirm what I have already found in practise - stop down metering works well enough, and M lenses are fun to use. I doubt I could have afforded a new version of the f2.5 135mm.

Yes, my memory of the discussions at the time is that stop-down was a lot better on the Kx00D than the Kx0D.

But it was rubbish on the K10D - completely unusable, and I stopped following the debate after I got rid of my M lenses.

In a sense, Gary's right - if you have newish lenses, it's all a bit of a non-issue, since you'll just stick the aperture ring on A if it has one and forget about it.

As soon as you're really up against a budget, though, it all becomes very relevant again, and a body that can use the old lenses easily has a huge advantage.

Indeed. This is the whole point. The lighting is constant, or as near constant as it needs to be; the aperture is the independent variable, and the exposure is the dependent one.

The two series of measurements are respectively the exposure that the camera (a) calculates with the lens set to A and (b) measures when stopped down with the green button.

You really do not need the ultimate in precision lighting to establish whether exposures follow the standard reciprocity law as you stop down.

If the wide open exposure is the same with the lens set to A mode and the camera to Av, as it is using the aperture ring and manual, you've already verified that the lighting is constant, to plenty enough precision for this experiment.

A milligram balance is not needed for making cakes, and nor is a micrometer needed to check your height.

I've contributed to a similar thread to this before, so apologies if I'm repeating myself. I fully agree with Chris with regard to the old lenses. I may be unusual, but I find myself using the old M42 and manual K-mount lenses more than my modern ones. I love the IQ they provide, and the general 'quality feel'. I also enjoy focussing and setting the exposure manually.

My K-x (and my son's K-m) have no trouble with stop-down metering on any of these old lenses and the exposures rarely need any post processing adjustment. I regularly use M42 Takumar 135 f3.5, Pentax-M 28 mm f3.5, Tamron Adaptamatic 300mm f5.6 and all of these meter accurately. All of these lenses were under £30, and in general produce better results than my newer glass.

I would thoroughly recommend the old manual lenses, subject to your camera metering with them as well as the K-x and K-m do.

Regards
David

Not sarcasm.

I did try to word the post correctly. Basically trying to bring to attention that it may not inspire some people to take up the brand.

I have used M series glass on my istD and K10D. I've not had issues with my A series lenses.

Best regards

Well, fair enough, if you say so. Perhaps you can see how I interpreted your comment:
... as such, if what you actually meant was the opposite:
The thing is, I've never said that M lenses cannot be used. Of course they can. Even with the limitations of the K10D, Pentax's old lens compatibility is probably better than any other brand - even with the limitations, so why would a discussion about it put people off Pentax?

And if the limitations have effectively gone away with the newer bodies, that's even more reason for someone to go to Pentax and experiment with all kinds of optically excellent, cheap lenses.

My feeling is that honest discussion of equipment and its limitations is more likely to have a positive effect than a negative one - especially if the discussion shows that the limitations can be overcome with the older bodies, and aren't there at all with the newer ones.

Other brands' mounts have changed over the years, as I understand it, so the older lenses can't even be physically attached to their cameras in some cases.

Seriously, if a non-Pentax user has taken the trouble to come here and read the discussions, in an attempt to find out about the brand, they are likely to be a lot more careful than most to research things properly. They deserve the truth

It's strange that people seem to think that Pentax have improved compatibility with old lenses with the K-5. I can't believe that they would even bother considering doing that

It is a FACT that the K-5 no longer recognises 'A' Series lenses and you have to manually input the focal length for SR in the same way as you do for M42 and 'M' lenses PhotoMe doesn't seem to be able to find the name for the lens either.